crr.z: 6r/)( * *<* t.*' O'c, % \ 4^ ^-«^s>^ A The Global Weathe Experiment — Final Report of US Operations Rockvllle, Md. April 1981 U. S. DEPARTMENT OF COMMERCE National Oceanic and Atmospheric Administration Office of Research and Development -^ OVERVIEW OF THE GLOBAL IaTEATHER EXPERIMENT The Global Weather Experiment was the largest international scientific experiment ever attempted. WHEN Build-up Year 1 December 1977 to 30 November 1978 Operational Year 1 December 1978 to 30 November 1979 Special Observing Periods 5 January to 5 March 1979 1 May to 30 June 1979 The research phase actively began in 1979 and will continue for a decade. WHO Over 140 countries contributed to the Global Weather Experiment through the World Weather Watch. Of these, 70 countries and 5 international organizations made special contributions to the Global Weather Experiment and the associated regional experiments. Total international participation included over 5000 individuals. HOW OBSERVING SYSTEMS WORLD WEATHER WATCH: Surface network, upper air network, special stations, temporarily upgraded island stations. GEOSTATIONARY SATELLITES: GMS (Japan) 140°E; GOES-West (USA) 135°W; GOES-East (USA) 75°W; METEOSAT (ESA) 0°; GOES Indian Ocean (USA) 58°E. POLAR ORBITING SATELLITES: NOAA-5; TIROS-N; NOAA-6; NIMBUS-7. DRIFTING BUOYS: 319 in the Southern Hemisphere; 28 oceanographic drifters in the Tropics; 27 drifting ice buoys in the Polar Regions. TROPICAL CONSTANT LEVEL BALLOONS: 313 platforms and approximately 50,000 wind and temperature observations. AIRCRAFT DROPWINDSONDE SYSTEMS: U.S. Air Force C-141's in the Pacific and the Atlantic; U.S. Air Force C-135's in the Atlantic; NOAA P-3's and C-130 in the Indian Ocean. Nine aircraft and 5,000 wind and thermodynamic soundings. TROPICAL WIND OBSERVING SHIPS: SOP-I: 40 ships; SOP-II: 43 ships. AUTOMATED AIRCRAFT FLIGHT LEVEL SYSTEMS: 80 Aircraft Integrated Data Systems (AIDS); 17 Aircraft to Satellite Data Relay Systems (ASDAR); 11 ACARS/MOAT systems (ARINC Communications Addressing and Reporting System/Meteorologi- cal Optional Auxiliary Terminal). OTHER OBSERVING SYSTEMS: The regional experiments associated with the Global Weather Experiment are the Winter. and Summer Monsoon Experiments (MONEX); the West African Monsoon Experiment (WAMEX) ; and the Polar Experiment (POLEX). There were also related oceanographic experiments being carried out in the Tropical Oceans. DATA MANAGEMENT: 26 Data Management Centers worldwide COST The estimated international financial resources for the Global Weather Experi- ment are between $300 and $500 million, depending upon how one allocates the satellite costs between FGGE operations and normal operations. Similar uncer- tainty exists for the U.S. costs, but the total is approximately $200 million or approximately $1.00 for every U.S. citizen. "^^^'■-WElTofC.O^ ..ss.. The Global Weather Experiment — Final Report of U.S. Operations Collected and edited by the U.S.FGGE Project Office Rockville, Md. April 1981 U. S. DEPARTMENT OF COMMERCE g^ Malcolm Baldrige, Secretary o B National Oceanic and Atmospheric Administration g^ James P Walsh, Acting Administrator (U o &> Office of Research and Develooment ^ Ferris Webster, Assistant Administrator Mention of a commercial company or product does not constitute an endorsement by the NOAA Office of Research and Development. Use for publicity or advertising purposes of informa- tion from this publication concerning proprie- tary products or the tests of such products is not authorized. n PREFACE ,■■'''■ From the beginning of December 1978 through November 1979, the highest concentration of scientific resources ever assembled was brought to bear on the challenge of observing the Earth's atmosphere and oceans. This activity involved the efforts of over 140 countries and was called the Global Weather Experiment, the largest international scientific experiment yet attempted. The origin of this international effort occurred 18 years earlier when President John F. Kennedy gave his 1961 address to the United Nations General Assembly calling for international cooperation in the environment. This speech ignited a sequence of events leading to the unique 1967 agreement between an intergovernmental body, the World Meteorological Organization of the United Nations, and a non-governmental body, the International Council of Scientific Unions, to sponsor jointly the Global Atmospheric Research Program (GARP). From the beginning of GARP, it was understood that an extended global observing program was needed, and after more than 10 years of planning, the First GARP Global Experiment (FGGE), also called the Global Weather Experiment, became a real ity. It was originally thought that a composite observing system of satellites, aircrafts, ships, drifting buoys, balloons and other special systems augmenting the conventional surface and upper-air observations would monitor the entire global atmosphere for a full year. However, primarily due to the high operating costs of aircraft and ships, the final design of the Global Weather Experiment confined operations of some systems to two Special Observing Periods within the Operational Year. Following the field phase and data management phase, a multi-year evaluation and research program began. It will continue until the late 1980s. An explicit statement of the goals of the Global Weather Experiment can be found in several GARP publications (a complete bibliography of related documents can be found at the end of this volume). These goals can be briefly summarized in layman's terms: (i) to improve our understanding of atmospheric dynamics and the general circulation of the atmosphere, and hence improve our ability to model those mechanisms responsible for that circulation; (ii) to determine the theoretical and practical limits of atmospheric predictability; (iii) to design an optimal, affordable observing system for the future; (iv) to improve models of climate change by fully and accurately simulating the annual cycle as observed over FGGE's Operational Year. Achieving these goals will require the zealous efforts and creative talents of the research community. Early indications are that the seeds sown in this international effort will reap a rich harvest of results -- meeting the aforementioned goals serving as a fountainhead for fresh ideas, and stimulating new thrusts across a spectrum of atmospheric research topics and applications. These early results have already appeared in published reports and scientific periodicals and are not repeated in this volume. This brings me to the purpose of the unique collection of papers which comprise this report. The manuscripts in this set describe the field events during the Experiment and review the complex operations through the eyes of those who were present. They are meant to complement those many planning ill documents prepared before the Experiment and the scientific results which have been or will be published after the Experiment. This volume, with the above mentioned literature, thus attempts to complete a proper historical perspective of the Experiment as a whole. Another purpose of this document is to aid in the planning and implementation of future large scale experiments. The authored papers contained herein describe operational experiences with the various United States observing systems. In view of the fact that the United States contributed substantially to each of the major types of observing systems, this provides a reasonably comprehensive guide to the international experience as a whole. In describing the observing system operations, authors were asked to provide a brief review of system objectives, data extraction techniques, notable successes, difficulties encountered and fixes required, along with recommendations for future experiments. A wealth of material and detail has been provided. We hope it will serve as a useful reference. Chapter 1 is a discussion of the United States FG6E Coordinating Center, a critical component to an experiment of this magnitude. Chapters 2 and 3 describe the United States polar orbiting and geostationary satellite operations. The Experiment, including the associated Regional Experiments (the Monsoon Experiment, the West African Monsoon Experiment and the Polar Experiment), was fortunate to have all of the originally planned five geostationary satellites operating for the FGGE year. This was the first and only year of such coverage. There are currently only three of these satellites operational; the European Space Agency satellite failed just four days before the year-long Experiment ended and the spare United States satellite temporarily placed over the Indian Ocean for the Experiment has since been returned to its pre-Experiment location. A special note of praise is due those agencies in the United States (NASA, NOAA and NSF) and in Europe (European Space Agency) for their rapid response in filling the unexpected eleventh hour gap in the Indian Ocean satellite coverage. Chapters 4 and 5 describe United States buoy and constant level balloon operations. These systems communicated their data through the new United States third generation polar orbiting satellite system, TIROS-N. These three systems together made substantial contributions to the Experiment. The drifting buoys gave us a new perception of the number and intensity of Southern Hemisphere storm systems; the constant level balloons provided over 50,000 upper tropospheric wind vectors in the tropics; TIROS-N supplied improved temperature soundings globally (see Chapter 2), and the French-built communication system on-board was indispensable to the success of the Experiment. Consequently, the last minute launch of the satellite just prior to the Experiment provided an extra level of excitement. Chapters 6 and 7 discuss the Aircraft Dropwindsonde System deployed in the tropics. The United States had agreed to assume the lion's share of the difficult challenge of observing the three dimensional structure of the tropical wind field. United States Air Force and NOAA research aircraft flew missions over all three tropical oceans covering a combined area of approximately 10 million square miles each day. Complementing the aircraft were the Tropical Wind Observing Ships. The United States contribution to this system is discussed in Chapter 8. iv Chapters 9 and 10 describe United States efforts to augment the World Weather Watch during the Experiment. These include: the implementation of a new real-time automated observing system for wide-bodied commercial aircraft (ASDAR), a similar automated record-only system (AIDS), and special efforts to both upgrade and add new land-based upper-air observations in the tropics. Chapter 11 reveals the United States Data Management operations, which were an integral part of the complex international data management plan. In addition to operational and special data collection efforts, several important data sets were provided by NASA's NIMBUS-7 research satellite. While these were planned as part of the Experiment, the operation of the satellite system was primarily independent and specific discussion of it is not included here. United States contributions to the Monsoon Experiment (MONEX) and the Polar Experiment (POLEX), which were regional experiments associated with the Global Weather Experiment, are also not included here. The planning for these activities was fully coordinated with the United States FGGE Project Office, but the successful operations of their respective observing systems v/ere for the most part independent (e.g., there was a separate MONEX Project Office responsible for the United States contributions to MONEX, and the operation of the United States drifting ice buoys program in the Arctic was conducted by the University of Washington). FGGE oceanographic activities were primarily planned and implemented by individual scientists and are not reported here. Finally, there was another real-time aircraft system operated by American Airlines, producing data similar to ASDAR and AIDS, which provided relevant error statistics of winds and temperatures over the continental United States. This program was not an official part of the Experiment and is not discussed in this report. This Experiment represents only a brief episode in the' evolution of atmospheric research. However, it was carefully constructed on the foundation of our previous knowledge, and I believe that it will be viewed as a memorable milestone in atmospheric science. Let us hope that history also looks favorably upon the actions of the nations of the world immediately after the Global Weather Experiment -- that they used the Experiment as a signal for ever greater cooperation and as a pivotal milestone which launched a truly global operational atmospheric observing system. A great many scientists, engineers and administrators from many countries contributed to the Global Weather Experiment planning and implementation through the years. Those agencies and organizations which funded and/or participated in the United States implementation are listed in the appendix. I do wish to thank the authors who contributed to this volume. Thanks are also due to Mr. William Murray for coordinating its preparation and for his help in editing; Ms. Tina Loughran for final assembly and graphics work; and Mrs. Betty Sonnefeld and Ms. Noreen Prather for their splendid typing. Rex J. Fleming Director U.S. FGGE Project Office CONTENTS Overview of the Global Weather Experiment cover 2 Preface, R. Fleming iii Chapter 1. U.S. FGGE Coordinating Center, T. Kaneshige 1 Chapter 2. Operational U.S. Satellites, NESS, R. Green 13 Chapter 3. GOES, Indian Ocean, F. Kahwajy and F. Mosher 27 Chapter 4. Southern Hemisphere Drifting Buoys, E. Kerut 41 Chapter 5. Tropical Constant Level Balloons, E. W. Lichfield 61 Chapter 6. Aircraft Dropwindsonde Program, 0. Scribner and J. Smalley 71 Chapter 7. Dropwindsonde Operations, Part 1 The Pacific, E. Tiernan 87 Chapter 7. Dropwindsonde Operations, Part 2 The Atlantic, J. Smalley 103 Chapter 7. Dropwindsonde Operations, Part 3 The Indian Ocean, J. McFadden Ill Chapter 7. Dropwindsonde Operations, Part 4 Flight Track and Mission Summary, 0. Thomas 123 Chapter 8. Tropical Wind Observing Ships, W. Keenan 153 Chapter 9. Aircraft to Satellite Data Relay, J. Sparkman, Jr., J. Giraytys, and G. Smidt 173 Chapter 10. Augmentation of the World Weather Watch (U.S. Participation), T. Bryan 185 Chapter 11. Data Management Overview, J. Harrison 195 References 207 Bi bl iography 210 Appendix 213 vn ,jn:-,j ««J6.- .' . ' i: ••'}.' "-ir? OC:,, „ 'I "} (If V "■• v;^^^r, htir/i? ni-:li. *! if'; -■^;i^^ri'! ..J :■■"!■ V '.('^ ■*';;/.; ^ ^ "I CJ .1' •." 6 T' * i";0' w'- I'V . '/'.? Jiv r, r ... •dJi^' '. t^ CHAPTER 1 U. S. FGGE COORDINATING CENTER By Mr. T. Kaneshige (U. S. FGGE Project Office) 1. INTRODUCTION The First GARP Global Experiment (FGGE) - the Global Weather Experiment, employed a complex array of observing systems and data management activities to produce the meteorological and oceanographic observations and initial state parameters required for FGGE research. Because of the global nature of the experiment, these activities were widely dispersed among a large number of nations and international organizations. To ensure the success of the experi- ment, an international FGGE Operations Center was established within the WMO Secretariat in Geneva, Switzerland, during the FGGE Operational Year. The FGGE Operations Center relied heavily on the periodic status reports received from the various observing system and data management components to monitor the day-to-day status of the FGGE operations, and to assess (in a preliminary way) whether or not the FGGE objectives for observational coverage were being met. The U.S. FGGE Coordinating Center (US-FCC) served as the focus for all U.S. participation in the Observational Phase of the experiment and provided the necessary U.S. status reports to the international FGGE Operations Center. In addition, the US-FCC was also involved in some of the day-to-day planning and execution of U.S. field activities during the Special Observing Periods. Descriptions of the FGGE observing systems and progress reports on the successes and failures of these systems during the first seven months of the Operational Year (December 1978-June 1979) were described in the articles by Fleming et al (1979a, 1979b). This report describes the US-FCC and its planned activities, summarizes the results of the departures from the planned operations, highlights the difficulties encountered and the corresponding actions taken, and concludes with some statements on the overall success of the operations and on recommen- . dations for coordinating future operational activities. 2. PURPOSE OF THE CENTER The U.S. FGGE Coordinating Center was established to serve: as the national focus for all U.S. FGGE operational activities (tasks included monitoring the status of implementation and operations of all U.S. components; providing U.S. field units with operational, logistical, and administrative support as needed; and providing the Director of the U.S. FGGE Project Office with up-to-date information on the progress of U.S. operational activities.) as the international focus for all U.S. FGGE activities (tasks included providing the international FGGE Operations Center with periodic status reports on the operations of U.S. FGGE special observing systems and data management centers; serving as the focal point for ensuring that requests from the FGGE Operations Center for changes to U.S. FGGE operations plans were evaluated and implemented, if possible, and providing other international centers with pertinent information concerning the deployment and operations of certain U.S. platforms.) as the source of operational guidance for certain U.S field units during the FGGE Special Observing Periods (tasks included evaluating the status of dropwindsonde aircraft and equipment; providing the anticipated wea- ther along the flight tracks; transmitting messages to each aircraft operating location containing recommended tracks to be flown the following day; evalua- ting all pertinent information related to the spatial distribution and move- ment of tropical constant level balloon platforms; and transmitting advisories, as needed, to the balloon launch sites concerning balloon launch operations.) 3. DESCRIPTION OF THE CENTER AND ITS PLANNED ACTIVITIES 3.1 Organization of the Center The US-FCC was organized under the U.S. FGGE Project Office, which pro- vided or arranged resources needed to implement and operate the center. The center had direct access to international and national centers and activities, as shown in Figure 1. 3.2 Description of the Center During the FGGE Operational Year (1 December 1978 to 30 November 1979), US-FCC operations were conducted for the most part during normal working hours (Monday-Friday, 0800-1630 Eastern Time) at the U.S. FGGE Project Office (FPO) in Rockville, Maryland. Staff members of the FPO and other NOAA headquarters components participated in the work of the center, since only a few individuals were directly assigned to the US-FCC. During the two FGGE Special Observing Periods (SOPs), which extended from January 5 through March 5, 1979, and from May 1 through June 30, 1979, a special center was activated in the World Weather Building in Marlow Heights, Maryland, to enhance operational support of U.S. field activities and inter- national monitoring and management activities. The center was located in a work area provided by the National Meteorological Center's Forecast Division, and US-FCC personnel were within easy reach of the National Meteorological Center, the National Environmental Satellite Service's (NESS) Satellite Analysis Branch, and the National Weather Service's (NWS) Communications Operating Branch. The center was manned 24 hours a day during the intensive periods of the SOPs, and during most of the daylight hours of the nonintensive periods of the SOPs. In addition to NOAA personnel, the National Center for Atmospheric Research provided a scientist, and the USAF Military Airlift Com- mand and Air Weather Service provided aircraft, weather, and communications specialists to assist in the work of the US-FCC during the intensive periods of the SOPs. The center had multiple communications links with the centers and ac- tivities shown in figure 1. These included commercial and U.S. military voice systems, the World Weather Watch Global Telecommunications System (GTS), commercial and U.S. military teletype networks, and a special National Aero- nautics and Space Administration satellite link between the United States and Europe. 3.3 Planned Activities • During the intensive periods of the SOPs, primary consideration was given to supporting the daily aircraft dropwindsonde operations out of the four operating locations (OLs): FGGE OPERATIONS CENTER (GENEVA) BUOY DEPLOYMENT CENTER (CANADA) [AIRCRAFT) OL-1 CANAL ZONE (I), MEXICO (II) OL-2 HAWAII OL-3 DIEGO GARCIA OL-4 ASCENSK U.S. FGGE COORDINATING CENTER (WASHINGTON) (TWOS) (TCLBS) WILKES (I, II) RESEARCHER (I, II) DISCOVERER (I), OCEANO- GRAPHER (II) CROMWELL (I), KNORR (Ii; GYRE (I. II) JORDAN (I), I5ELIN (II) ASCENSION LAUNCH CANTON LAUNCH GUAM LAUNCH (II ONLY) SERVICE ARGOS FRANCE [BUOYS] (LAND STATIONS) (SATELLITES) ORCADAS MAUMEE BLAND ACUSHNET CANTON ENEWETAK WOLEAI TIROS-N, GOES NIMBUS-G POLAR STAR KAPINGA- MARANGI SERVICE ARGOS FRANCE Figure 1 .--Interfaces for U.S. FGGE Coordinating Center. OLl: 0L2: 0L3: 0L4: Howard AFB, Canal Zone (SOP-1); Acapulco, Mexico {SOP-2) Hickam AFB, Hawaii Diego Garcia Ascension Island This meant that both duty scientists and aircraft specialists were needed during the times of the day when the status of aircraft, equipment, expen- dables, and crew and the forecast weather conditions over the flight track areas had to be carefully evaluated prior to decisions on the flight tracks for the subsequent day's missions. Although the aircraft operated out of bases spread throughout most of the tropical belt, flight track selections for more than one operating location could be made at the same time because of the schedule of availability of certain guidance products. For this system for communicating status reports from the island sites to the US-FCC was to use an HF radio link from the island sites to an appropriate Trust Territory District Center, where a teletype message was expected to be pre- pared and disseminated via teletype to the US-FCC. A few status reports were received at the US-FCC during the beginning of SOP-1, but severe problems with maintaining working generators at the observing sites resulted in the loss of some of the status reports. Since the 6th Weather Squadron (6WS) headquarters at Tinker AFB, Oklahoma was in frequent radio contact with its mobile upper air teams on Kapingamarangi and Woleai, the US-FCC decided to make an informal arrangement with the 6WS to obtain status information di- rectly from the 6WS. This modified procedure provided most of the needed information. 4.2.2 Meteorological satellites The coordination of satellite operations was very successful. Desig- nated representatives of the National Environmental Satellite Service pro- vided timely information by telephone on changes to the status of opera- tional polar-orbiting and geostationary satellites. 4.2.3 Aircraft dropwindsonde program ' The planned use of teletype and voice communications between the US- FCC and the aircraft operating locations to coordinate the FGGE dropwindsonde missions worked \/ery well. Both incoming and outgoing messages (teletype and voice) were disseminated/received in a timely manner. When unusual problems occurred, the problems were discussed and solved over the telephone. Weather forecast support from the NESS satellite meteorologists was good. Timely satellite photographs and film loops for the Pacific and Atlantic Ocean areas were available for the track selection decision making process. Available polar-orbiting satellite photographic products for the Indian Ocean area tended to be late or old, but the satellite meteorologists gave their best efforts to provide US-FCC scientists with "long-range" forecasts for the Indian Ocean flight track areas. FGGE TWOS status information from the international FGGE Operations Center for the most part was available to the US-FCC. Some messages were un- explainably lost. Daily US-FCC outgoing aircraft dropwindsonde status reports to the FGGE Operations Center were disseminated routinely over the GTS without any difficulties. 4.2.4 Tropical constant level balloon system - The coordination of balloon platform launches was outstanding. The launch schedules were pre-planned but adjustments could be made if either the launch site or the US-FCC believed them to be necessary. On a few occa- sions when the weather at and around the Canton Island launch site was not favorable, launches were temporarily halted or slowed down until more favor- able conditions occurred. In addition, when the circulation patterns at platform altitude were unfavorable, a similar halt or slowing down of the Table 1.--US-FCC personnel (World Weather Building operations) 1. Director : Thomas M. Kaneshige 2. Officer in Charge , Aircraft Specialists : Onial A. Thomas X X X X X X X X X X X X X X X X X , X X X X X X X SOP-1 SOP-2 3. Duty Scientists : Paul R. Julian (NCAR) John Pavone, Major, USAF (AWS) James K. Sparkman (FPO) Wayne E. McGovern (FPO) . 4. Aircraft Specialists : . Donald A. Thompson (N0AA/0A3) Roger Sorenson, Capt., USAF (MAC) Samuel Trunzo, Capt., USAF (AWS) David Gurkin, Col., USAF (AWS) ,,. Ronald Barrick, Capt.. USAF (MAC) Delbert Simmons, Maj., USAF (MAC) Charles Steverson, Capt., USAF (MAC) 5. Duty Coordinators : Donald J. Florwick, CMDR, NC (FPO) ^ Karen L. Cox, LTJG, NC (FPO) David K. Howard, LTJG, NC (FPO) Michael J. Kretsch, LTJG, NC (FPO) Earl Snipes, CAPT, USAF (AWS) 6. Other FPO Members Who Assisted : •<' Kenneth W. Foulke ,, -■ Warren H. Keenan „ mainly to a diminishing number of available personnel, were necessary during the remainder of SOP-1. Work schedules for SOP-2 were quite stable, and the three categories of personnel (duty scientist, aircraft specialist, duty coordinator) were scheduled so as to provide for maximum work accomplish- ment by a minimum number of personnel. During the nonintensive periods of the SOPs, the center was manned during most of the daylight hours by duty coordinators. Table 1 lists the personnel who participated in the work of the US-FCC in the World Weather Building. 4.2 Observing Systems Support The results of the coordination and operational support of U.S. components of the FGGE composite observing system are summarized below. The effectiveness of communications support, which was arranged prior to the start of field operations, is also indicated. 4.2.1 Special land-based upper air stations The coordination of the upper air observing programs on Enewetak and Canton Islands worked very smoothly. All scheduled messages and a few special operational problem reports were received in a timely manner at the US-FCC. The success of this coordinating effort was due in part to the excellent cooperation of the special U.S. Army mobile upper air teams and also to the excellent (existing) military telecommunications system. The coordination of the programs on Kapingamarangi and Woleai did not fare as well because of severe telecommunications difficulties. The pla'nned system for communicating status reports from the island sites to the US-FCC was to use an HF radio link from the island sites to an appropriate Trust Territory District Center, where a teletype message was expected to be prepared and disseminated via teletype to the US-FCC. A few status reports were received at the US-FCC during the beginning of SOP-1, but severe prob- lems with maintaining working generators at the observing sites resulted in the loss of some of the status reports. Since the 6th Weather Squadron (6WS) headquarters at Tinker AFB, Oklahoma, was in frequent radio contact with its mobile upper air tems on Kapingamarangi and Woleai, the US-FCC decided to make an informal arrangement with the 6WS to obtain status information directly from the 6WS. This modified procedure provided most of the needed information. 4.2.2 Meteorological satellites The coordination of satellite operations was very successful. Des- ignated representatives of the National Environmental Satellite Service provided timely information by telephone on changes to the status of opera- tional polar-orbiting and geostationary satellites. 4.2.3 Aircraft dropwindsonde program ; The planned use of teletype and voice communications betv/een the US- FCC and the aircraft operating locations to coordinate the FGGE dropwindsonde missions worked very well. Both incoming and outgoing messages (teletype and voice) were disseminated/received in a timely manner. When unusual problems occurred, the problems were discussed and solved over the telephone. Weather forecast support from the NESS satellite meteorologists was good. Timely satellite photographs and film loops for the Pacific and Atlan- tic Ocean areas were available for the track selection decisionmaking process. Available polar-orbiting satellite photographic products for the Indian Ocean area tended to be late or old, but the satellite meteorologists gave their best efforts to provide US-FCC scientists with "long-range" forecasts for the Indian Ocean flight track areas. ; FGGE TWOS status information from the international FGGE Operations Center for the most part was available to the US-FCC. Some messages were unexplainably lost. Daily US-FCC outgoing aircraft dropwindsonde status reports to the FGGE Operations Center were disseminated routinely over the GTS without any difficulties. - " :, .i ' -.'i ' 4.2.4 Tropical constant level balloon system The coordination of balloon platform launches was outstanding. The launch schedules were preplanned, but adjustments could be made if either the launch site or the US-FCC believed them to be necessary. On a few occasions when the weather at and around the Canton Island launch site was not favorable, launches were temporarily halted or slov/ed down until more favorable conditions prevailed. In addition, when the circulation patterns at platform altitude were unfavorable, a similar halt or slowing down of the platform launches took place. The military teletype system provided excellent telecommunica- tions for the necessary coordination. For the determination of balloon distribution and platform level circulation patterns, the US-FCC relied solely upon the teletype printouts of balloon reports (COLBA) received via the GTS and the NMC tropical 150-mb analysis displays with data plots. There were significant problems with the COLBA reports received during SOP-1, which made it difficult to assess the distribution of platforms. The probable reason for the problems was the limited amount of time France had to check out its real-time proc- essing and dissemination programs before declaring them operational. (The overall delays in the TIROS-N launch reduced this checkout time.) The situation improved markedly toward the end of SOP-1, and few problems were experienced during the remainder of the SOPs. Arrangements for obtaining copies of the NMC 150-mb analyses worked very well. 4.2.5 Tropical wind observing ship program The success of the coordination of U.S. participation in the inter- national TWOS operations varied from ship to ship and from ocean to ocean. The planned coordination effort called for each U.S. ship to transmit to the US-FCC a daily status report in a prescribed format. Ships operating in the Pacific and Atlantic Oceans were expected to telecommunicate their reports via HF radio to the nearest U.S. Coast Guard radio station for relay to the US-FCC. Ships in the Indian Ocean were expected to telecommunicate their reports to one of the U.S. Navy communications stations in the area for relay to the US-FCC. For outgoing messages from the US-FCC to the ships, plans called for the dissemination of messages via the reverse of the arrangements specified above. In view of the lateness in finalizing the ship operations plans, it appears likely that the planning arrange- ments with some of the ships' operators were never fully coordinated. Overall, the reception of status reports from the NOAA ships (RESEARCHER, DISCOVERER, OCEANOGRAPHER, TOWNSEND CROMWELL and DAVID STARR JORDAN) was excellent, regardless of the ocean in which the ships were operating. This could be attributed in part to the more complete coordina- tion effort achieved by the US-FCC prior to the start of ship operations. Reception of status reports from the USNS WILKES was spotty. Additionally, the ship was diverted from FGGE work to higher priority Department of Defense missions during part of SOP-2. The university ships (KNORR, GYRE, COLUMBUS ISELIN) did not communicate their status reports according to plans. As a result, the US-FCC in telephone coordination with the research institutions responsible for the operations of the ships was able to make arrangements to have each institution obtain the needed status information. The US-FCC then obtained the information by telephone from the institutions. Unfortunately, no such arrangements could be made over the weekends and on holidays, so status reports for these were obtained on the first working day following the weekend or holiday. US-FCC personnel encountered significant difficulties in the dis- semination of messages to the ships. During SOP-1, it was not always clear which coastal radio station was monitoring the radio transmission of each ship, so there was an element of guess-work involved in identifying the radio stations to which the message should be addressed for relay to the ship. On a few occasions, messages were returned "un-sent" to the US-FCC because the addressed radio station did not know the whereabouts of the ship. In some instances, it was not possible to get messages to a ship. Overall, there was some improvement during SOP-2. The US-FCC agreed to send twice-weekly status reports about the oper- ations of each U.S. ship to the FGGE Operations Center. However, because the US-FCC Operations Plan and the international TWOS Operations Plan were not finalized until quite late, there were some mix-ups in the coordination be- tween the US-FCC and the FGGE Operations Center. The format for the status reports prepared by the US-FCC were not exactly in agreement with the format prescribed by the FGGE Operations Center. This problem was resolved during the period between the two SOPs. On a few occasions throughout the SOPs, status reports were not sent to the FGGE Operations Center. This was due to temporary breakdowns in the internal tracking system of scheduled incoming/ outgoing messages. 4.2.6 Southern Hemisphere drifting buoy system The deployment of U.S. drifting buoys for SOP-1 was planned in con- siderable detail and there was only a minimum amount of involvement of the US-FCC in this activity. During the deployment phase, each of the deployment ships regularly transmitted buoy checkout and deployment reports to the US-FCC. The US-FCC in turn summarized the information and transmitted weekly status reports of U.S. buoy deployments to the FGGE Buoy Logistics and Deployment Center in Vancouver, Canada. > . The United States also participated in the international re-seeding effort for SOP-2 by air-deploying 18 buoys from two USAF C-141 aircraft oper- ating out of Argentina and Australia/New Zealand. Some of the expected buoy deployment reports from the aircraft staging locations were not received at the US-FCC until after the aircraft and personnel returned to the United States. However, since there was no great urgency for the deployment infor- mation, the "apparent" loss of the near-real-time information did not have any impact on operations. 4.2.7 Automated AIREPS US-FCC involvement in monitoring the status of operations of wide- bodied jet aircraft participating in the Aircraft to Satellite Data Relay (ASDAR) program was very limited, since personnel responsible for the ASDAR program were maintaining a close watch over the program. The US-FCC did monitor the receipt of ASDAR reports via the GTS by collecting all avail- able ASDAR bulletins and delivering them to ASDAR program personnel. 4.3 Data Management Support The US-FCC began monitoring the activities of U.S. FGGE data manage- ment centers in July 1978. From monthly information received by mail, the US-FCC summarized the activities at each center and prepared monthly status 10 .; reports of U.S. FGGE Data Management Activities. These reports were mailed to the international FGGE Operations Center from July 1978 through December 1979, when the US-FCC terminated operations. (Data management status reports will continue to be prepared through June 1980 by the U.S. GARP Office.) This activity was \/ery successful. 5. DIFFICULTIES ENCOUNTERED AND ACTIONS TAKEN As with most field experiments, no amount of planning can take into account all that eventually occurs and the success or failure of the opera- tions is somewhat dependent on the creativity and flexibility of the indi- viduals participating in the experiment. The success of the US-FCC operations is due in part to the planning effort, but also to the inventiveness, adapt- ability, and hard work of the individuals who participated in the work of the center. There were some difficulties encountered in conducting the operations of the US-FCC in the World Weather Building. Some of the operational diffi- culties have already been discussed in section 4 and need not be repeated here. The following are difficulties associated with personnel and physical arrangements. 5.1 Adverse weather conditions . The Washington, D. C. area was hit by a record 24-inch snowfall on 18-19 February 1979, which completely para- lyzed the city and its surrounding areas. The unfortunate aspect of this storm was that the heaviest snowfall occurred during the 4-hour period when the US-FCC was not manned on the evening of 18 February and the individual going off duty was able to leave the US-FCC, but the individual who was sup- posed to report for duty early on the 19th was completely snowbound. As a result, it was not until 1100 EST on 20 February 1979 that the first US-FCC individual was able to make his way into the US-FCC. This meant that the US-FCC was not manned for nearly 40 consecutive hours during the last days of the extended SOP-1 operations. 5.2 Building entry difficulties . Entry into the World Weather Building, where US-FCC operations were conducted during the SOPs, was limi- ted to individuals with building passes. The IIS-FCC made special arrange- ments with the Building Services Officer to ensure that US-FCC duty personnel would be able to gain access to the building during all hours of the day. On a few occasions, the building guards made it difficult for duty personnel to gain entrance, but in every case, the affected individual was able to gain entry. (All incidents were subsequently reported to the Building Ser- vices Officer.) 5.3 GTS teletype printer . The US-FCC was unable to obtain a GTS teletype printer on its premises during SOP-1. The printer was needed to monitor the receipt of near-real-time aircraft dropwindsonde, constant level balloon, drifting buoy, and conventional upper air reports. As a result, US- FCC duty personnel had to pick up the information at the Federal Office Building No. 4 in Suitland, Maryland. This made it difficult to monitor n the operations of these programs. The situation was resolved with the in- stallation of a printer prior to the start of SOP-2. 5.4 Tracking of incoming/outgoing messages, charts, satellite photos, etc. There was a considerable amount of information flowing in and out of the US-FCC, and some information was lost or misplaced. Also, some of the out- going messages were mistyped by the communications center or, worse yet, not drafted for delivery to the communications center. Where necessary, a check was made with the communications center to locate missing (scheduled) incoming reports and copies of the reports were usually accounted for. In the case of erroneous or missing outgoing messages, US-FCC usually followed up with cor- rected messages. 6. CONCLUSIONS AND RECOMMENDATIONS Overall, US-FCC operations were yery successful. The individuals who participated in the work of the center did an excellent job in resolving the numerous difficulties which arose and in maintaining the schedule of planned operations. The support from the NWS Communications Operating Branch in handling all incoming and outgoing teletype traffic was outstanding, as was the support from the NESS Satellite Analysis Branch in providing satellite weather support for FGGE aircraft dropwindsonde missions. The NMC made all the necessary arrangements for the physical location and furniture for the center, and provided several types of guidance products used in conducting the day-to-day operations. Regarding recommendations for coordinating future experiments, two important ones come to mind. These are: 6.1 Operations plans for the Operations Center or Coordinating Center should be developed, coordinated.^ and finalized in detail well enough in advance of the start of operations, so that a thorough checkout can be conducted before the start of actual operations . In the case of international experiments, and to a certain extent national experiments, this is always a problem. Final systems checkouts always seem to occur during the early phases of the actual operation. Nevertheless^ this should be a goal for any future experiment. 6.2 Procedures for tracking the flow of incoming and outgoing infor- mation should be developed in detail prior to the start of operations and adhered to during the operations . This was a problem during FGGE, because the procedures were not developed in as much detail as was needed. This resulted in an occasional loss or misplacement of incoming information and an occasional failure to disseminate scheduled messages. A good followup system is also needed to ensure that errors introduced in the preparation of messages are corrected before the messages are disseminated. A system for numbering all outgoing messages should be used to provide for better control over outgoing information. . . , 12 CHAPTER 2 OPERATIONAL U.S. SATELLITES NESS By Robert N. Green (NESS) L 1. INTRODUCTION The National Environmental Satellite Service (NESS) has played a major role in the First GARP Global Experiment (FGGE) - the Global Weather Experiment. Sea surface temperature values, cloud motion vectors, and vertical temperature soundings were derived for the FGGE Level Il-b Research Data Set from data ob- tained from the NESS fleet of polar-orbiting and geostationary operational en- vironmental satellites. In addition, the operational satellites provided image products for the archive at the National Climatic Center and served as a host to a Data Collection and Platform Location System for terrestrial and atmospheric instrumented platforms. Through the use of the operational satellites, thou- sands of meteorological measurements have been added daily to the total number of observations obtained through more conventional methods (e.g., surface and upper air observations). 2. U.S. OPERATIONAL SATELLITE SYSTEM NESS operated two environmental satellite systems during the Global Weather Experiment. These were the polar-orbiting, sun-synchronous satellites (NOAA-5, TIROS-N, and NOAA-6), which provided daily global coverage and the Geostationary Operational Environmental Satellites (GOES). The polar-orbiting data provided mainly global quantitative products such as vertical temperature soundings of the atmosphere and sea surface temperatures. The GOES satellites provided environmental data for the Earth's disk facing each satellite at per- iodic intervals, usually every 30 minutes. The data from both satellite systems were routinely processed by NESS into a variety of quantitative and image pro- ducts, which were then distributed to users. A system of five geostationary satellites operated during the operational year of the Global Weather Experiment (see Figure 1). The satellite over the Indian Ocean was also a U.S. satellite and its operation is discussed later in this report by Kahwajy (see Chapter 3). The GOES system consisted of operating satellites at 75°W and 135°W, the ground command and data acquisition station at Wallops Island, Virginia, the ground data acquisition station in Washington, D. C. , and a central data distri- bution system. A total of five different satellites were called upon to parti- cipate in the operational U.S. satellite system during the FGGE Build-up and Operational Years from December 1, 1977 - November 30, 1979. These were the Synchronous Meteorological Satellite (SMS-1), a NASA prototype for the GOES, launched in May 1974; SMS-2, launched in February 1975; GOES-1 , launched in October 1975; GOES-2, launched in June 1977; and GOES-3, launched in June 1978. At the start of the FGGE Build-up Year, the operational configuration of satellites had GOES-2 at 75°W (GOES-EAST) and SMS-2 at 135°W (GOES-WEST), fixed over the Equator at an altitude of about 36,000 km. The two satellites were positioned to provide overlapping and continuous coverage of the Western Hemisphere. On April 4, 1978, GOES-1 replaced SMS-2 at 135°W; GOES-1 was re- placed on July 13, 1978, by the newly-launched GOES-3 which remained at the GOES-WEST station through the end of the Operational Year. At the GOES-EAST 15 o CO • •r- ' — » »♦- -!-> o C o O) •!- CO ^ o ■M a. 1 E -^ O 3 o s- to o ^- £ >> o s- s- o fO o o E O CO •^ in w, ,^ M- o w> E o Ol (U o t/l +-> (D 3 CO >> •1- ■*-> I— •r- n— 4-> (U C +J (O (d o 3 «/) O O" CM >> ^~ s- s- o «o M- C O lO fO to T— S- +-> (U 10 > O O 0) o tJ Ol O 1 CO 1 T— • r— 0) s- o 3 in o> o CM O 16 position, SMS-1 replaced GOES-2 on January 26, 1979, and a switch to SMS-2, which held for the duration of the experiment, was completed on April 19, 1979. The primary instrument of the SMS and GOES was the Visible and Infra- red Spin-Scan Radiometer (VISSR). The VISSR provided concurrent observations in the infrared (IR) (10.5 to 12.5 ym) and in the visible (VIS) (0.55 to 0.75 ym) regions of the spectrum. The VISSR provided a full-disk view ewery 30 min- utes. (More frequent scanning could be obtained at the expense of spatial coverage.) The VIS channel provided 1-km daytime coverage whereas the IR chan- nel provided 8-km daytime and nighttime coverage. During the 2-year period of the FGGE observational phase, two differ- ent polar-orbiting satellite systems were used. At the start of the Build-up Year, the NOAA-5 satellite was operational. The primary instruments carried were the Vertical Temperature Profile Radiometer (VTPR) with eight channels varying in central wavenumber from 668 cm"^ to 833 cm"l and the Scanning Radi- ometer (SR), a 2-channel instrument sensitive to radiation in the 0.5-0.7 ym visible region and in the 10.5 to 12.5 ym infrared (IR) "atmospheric window" region. NOAA-5 had a nominal orbital altitude of 1,464 km, crossing the Equa- tor southbound at 9 a.m. local mean time with a nodal period of 115 minutes. In October 1978, the next generation of polar-orbiting satellites was introduced: the TIROS-N series (see Schwalb, 1978, for a detailed review of the TIROS-N series). During that month, TIROS-N was launched, followed in June 1979 by the launch of NOAA-6. These satellites had nominal altitudes of 866 km and 830 km with a southbound Equator crossing time of 3:14 a.m. and 7:32 a.m. local mean time, respectively. The nodal period was 101 minutes. 2 . 1 Primary FGGE environmental satellite sensors The TIROS Operational Vertical Sounder (TOVS) was a 3-instrument system consisting of: The High Resolution Infrared Radiation Sounder (HIRS/2) — a 20-channel instrument, which makes measurements primarily in the infrared por- tion of the spectrum (see Table 1). The instrument was designed to provide data that will permit calculation of (1) a temperature profile from the sur- face to 10 mb; (2) water vapor content in three layers in the atmosphere; and (3) total ozone content. The Stratospheric Sounding Unit (SSU) — employing a selective absorption technique to make measurements in three channels. The spectral characteristics of each channel were determined by the pressure in a carbon dioxide gas cell in the optical path. The amount of dioxide in the cells determined the height of the weighting function peaks in the atmosphere. The Microwave Sounding Unit (MSU) — a 4-channel Dicke radiom- eter, which makes passive measurements in the 5.5-mm oxygen band. This in- strument, unlike those making measurements in the infrared region, is little affected by clouds. 17 Table 1.— HIRS-2 channels TIROS-N/NOAA-A HIRS-2 Channel 1 Nomi nal Wavenumber i [cm-i; 1 669 TIROS-N/NOAA-A HIRS-2 Channel 2 Nomi nal Wavenumber [cm-i; 679 TIROS-N/NOAA-A HIRS-2 Channel 3 Nomi nal Wavenumber cm-V 690 TIROS-N/NOAA-A HIRS-2 Channel 4 Nomi nal Wavenumber cm"l 702 TIROS-N/NOAA-A HIRS-2 Channel 5 Nomi nal Wavenumber [cm-1 716 TIROS-N/NOAA-A HIRS-2 Channel 6 Nomi nal Wavenumber [cm-i; 732 TIROS-N/NOAA-A HIRS-2 Channel 7 Nomi nal Wavenumber cm"l 749 TIROS-N/NOAA-A HIRS-2 Channel 8 Nomi nal Wavenumber cm-1 900 TIROS-N/NOAA-A HIRS-2 Channel 9 Nominal Wavenumber [cm"l' 1030 TIROS-N/NOAA-A HIRS-2 Channel 10 Nom inal Wavenumber (cm-' ) 1229 TIROS-N/NOAA-A HIRS-2 Channel 11 Nom inal Wavenumber (cm-' ) 1345 TIROS-N/NOAA-A HIRS-2 Channel 12 Nom inal Wavenumber (cm- ) 1490 TIROS-N/NOAA-A HIRS-2 Channel 13 Nom inal Wavenumber (cm- ) 2190 TIROS-N/NOAA-A HIRS-2 Channel 14 Nom inal Wavenumber (cm- ) 2210 TIROS-N/NOAA-A HIRS-2 Channel 15 Nom inal Wavenumber (cm-' ) 2250 TIROS-N/NOAA-A HIRS-2 Channel 16 Nom inal Wavenumber (cm-' ) 2275 TIROS-N/NOAA-A HIRS-2 Channel 17 Nom inal Wavenumber (cm- 1) 2360 TIROS-N/NOAA-A HIRS-2 Channel 18 Nom inal Wavenumber (cm- ) 2514 TIROS-N/NOAA-A HIRS-2 Channel 19 Nom inal Wavenumber (cm- 1) 2680 TIROS-N/NOAA-A HIRS-2 Channel 20 Nominal Wavenumber (cm-1 ) 14440 ■■ , -?• , ■ (Vis Channel) i':55J^V 18 The Advanced Very High Resolution Radiometer (AVHRR) was a 4-channel scanning radiometer sensitive in the visible, near- infrared and infrared window regions. This instrument provided data for central processing at full resolu- tion (1.1 km). Also onboard the TIROS-N series was the ARGOS Data Collection and Platform Location System which was provided by the Centre National d 'Etudes Spatial es (CNES) of France. This was a random-access system to acquire data from fixed and free-floating terrestrial and atmospheric platforms. Platform location was made possible by ground processing of the Doppler measurements of carrier frequencies. Data collected from each platform included identification as well as environmental measurements. The primary uses of the ARGOS system during FGGE were the location determination and interrogation of the drifting buoy and constant level balloon systems. The ARGOS system worked very well since the operational start of TIROS-N on December 15, 1978. The data were moved from the satellite data stream and stored in computer disk storage until the data were sent via land line to New York City, NY, for transmission to Paris, France, through a commun- ications satellite and, again, by land line to Toulouse, France, for final processing by CNES. The remotely accessed data were planned to be delivered to CNES within three hours, and this requirement was met in most cases through- out the Operational Year. 3. U.S. OPERATIONAL SATELLITE PRODUCTS FOR FGGE RESEARCH 'i Three major NESS products were incorporated into the FGGE research data set: Satellite cloud motion vectors, sea surface temperatures, and ver- tical temperature sounding data. These products were operationally produced as part of the NESS commitment to various national and international programs. A decision was made in NESS to keep FGGE support operations separate from nor- mal operations, since FGGE required the products in a delayed delivery mode; thus, there was no operational impact on the NESS data processing facility. The U.S. national archive of magnetic tapes of all operational satellite pro- ducts was used as input data to produce FGGE tapes according to the Formats for International Exchange of Level II Data Sets. These tapes were prepared and delivered on a regular schedule to the FGGE Level Il-b Space-Based and Special Observing Systems Data Center in Norrkoping, Sweden. Each product line has had a unique history during the FGGE period and will be discussed individually. ^ ^^ 3.1 Cloud Motion Vectors ■ NESS produced these vectors from GOES images three times each day, at OOOOZ, 1200Z, and 1800Z, using computer automated techniques for most low-level (900-mb) vectors and manual film loop methods for middle- and upper- level vectors. A cross-correlation, computer model provided the automated detection and displacement measurements of selected digital GOES imagery to determine low-level wind vector estimates. Infrared data were aligned and displacement measurements were computed from two images that are 30 or 60 minutes apart. 19 Approximately 150 low-level vectors were produced from each satellite for each production time. Middle- and upper- level vectors were produced by meteorologists view- ing a time-lapse movie loop of six GOES images with a movie projector which dis- played the movie on a digitized plotting board. The resulting cloud displace- ments were translated into cloud motion vectors by a computer program. The meteorologists also appended cloud height estimates to these vectors and made a final assessment of both manual and automated vectors before finalizing the operational wind vector data set (Bristor, 1975). Approximately 1,400 satellite cloud motion vectors were produced each day from both techniques. ^ery few problems occurred during the FGGE period with cloud motion vectors. A few minor changes were made to the International Level II Data Ex- change formats effective October 1978, but this did not cause any delays in NESS processing. Table 2 summarizes the NESS production of cloud motion vectors for the FGGE research data set. 3.2 Sea Surface Temperatures Sea surface temperature (SST) observations produced by NESS during the two years of FGGE had a varied history of three different computational methods and two different satellites. The initial method was a statistical histogram analysis of data from the NOAA-5 SR sensor's infrared spectral window with atmospheric attenuation corrections derived from the VTPR sensor data (see Brower et al for a detailed discussion). This resulted in nearly 10,000 (4,000 of highest quality) observations daily with a horizontal reso- lution of about 100 km. On March 16, 1978, an equipment malfunction on NOAA-5 stopped the flow of SR data. A backup technique was quickly placed into the SST operation which used only VTPR data. The SST observation used a single spot value from the VTPR thermal window band which was moisture-corrected and determined cloud free by using various combinations of VTPR channels. Because this method worked only in cloud-free areas, the quantity of SST's dropped to 4,000-5,000 observations per day, but the accuracy remained comparable to the SR method. The TIROS-N SST processing system was declared operational on March 1, 1979, with an overlap operation with the NOAA-5 SST system beginning Janu- ary 1, 1979, and provided for a series of changes to the SST products. Some of the more significant changes were: The horizontal resolution of a nominal SST observation was in- creased to about 50 km. The number of highest quality observations was increased to approximately 30,000-40,000 per day or nearly one million per month. Cloud detection and atmospheric attenuation correction was im- proved through the use of TOVS data. - ,. 20 The retrieval method used was a maximum likelihood technique (a stand- ard statistical procedure) which was designed to estimate the mean clear radi- ance by checking only the warmest observations from a target of AVHRR infrared data. This method was simple and computationally fast and had been thoroughly tested on NOAA-5 SR data. The mean clear radiance was moisture-corrected using a combination of the various TOVS channels. Evaluation of the early operational observations showed systematic- ally colder satellite measurements as compared to ship observations in the tropics. Stricter cloud tests in the tropics eliminated much of this cold bias after February 1, 1979. Satellite SST observations in the middle-and high- latitude regions tended to have a slight warm bias during the summer. Table 2 summarizes the NESS production of sea surface temperatures for the FGGE research data set. 3.3 Vertical Temperature Profiles and Clear Radiances One of the more important types of data included in the Level Il-b data set is also one of the most complex operational satellite products: Vertical temperature profiles. This product is of great interest to the global atmospheric modeller, because it fills the void of information about the vertical temperature structure over the oceans. Two different satellite and production systems were used during FGGE: The VTPR instrument on NOAA-5 and the TOVS instrument on TIROS-N and NOAA-6. In the NOAA-5 data flow, radiance values derived from calibrated and earth-located VTPR channel data were passed to a clear radiance program, which operated to eliminate the effects of clouds through comparisons of adjacent VTPR data, sea surface temperature, and approximate first guess vertical temper- ature structure. The temperature profiles were obtained through a modified, minimum- root-mean-square solution of the radiative transfer equation (see McMillin et al , 1973, for further details). The average daily number of clear radiance retrievals was 1300-1400, and approximately 900-1000 temperature profiles were processed from the clear radiance retrievals. On March 1, 1979, the sounding processing system using TOVS data from TIROS-N was declared operational after about three months of test and evalua- tion following the turnover of the spacecraft to NESS. The operational system was comprised of the following components: Preprocessor, TOVS atmospheric radiance module, TOVS mapper, TOVS retrieval module, and output products module. The preprocessor reformatted the raw data from HIRS/2, MSU, and SSL) into a form convenient for other modules, and made limb corrections and water vapor attenu- ation corrections. The atmospheric radiance module computed clear-column radi- ance measurements by statistical regression techniques of various instrument channel combinations based upon the determination of a clear, partly cloudy, or cloudy target (see Smith and Woolf, 1976; and Smith et al., 1979; for a detailed discussion of the science of retrieval methods). After all the needed data were mapped and analyzed, the retrieval module accessed the clear radiances and produced the retrievals of atmospheric temperatures, water vapor profiles, geo- potential height, total ozone, average cloud amount^ and albedos. The parameters were quality controlled and processed into numerous products for the user. Some of these parameters have not met operational standards and, thus, have limited 21 Table 2. --Summary of monthly totals for satellite-derived observations archived by NESS for FGGE. Sea Surface Cloud Motion Atmospheric Clear Date Temperature Vectors Soundings Radiances Jan 1978 373648* 48488 22875* 38125* Feb 1978 337489* 43395 22192* 34142* Mar 1978 265169* 45590 22972* 37660* Apr 1978 163718* 46069 21329* 34966* May 1978 169176* 46507 • 22347* 35472* Jun 1978 .• 173079 49736 28192 46851 Jul 1978 162474 52427 .. 27841 46635 Aug 1978 . 125024 53811 23492 39756 Sep 1978 138764 47109 24056 37529 Oct 1978 135540 46577 24830 38174 Nov 1978 125504 50147 27189 41565 Buildup Year Total 2,169,585 529,856 267,315 . 430,875 Dec 1978 144488 45008 29620 42828 Jan 1979** 1123900 41285 • 217693 ' 218155 Feb 1979 657330 41483 191192 . 191195 Mar 1979 933952 45076 235237 238840 Apr 1979 813396 39443 228410 230524 May 1979 984650 44431 230412 230420 Jun 1979 961848 44491 247558 247836 Jul 1979 904778 44305 241765 : .. 244493 Aug 1979 1046012 49566 , 240030 245320 Sep 1979 1033572 48058 239023 . 244094 Oct 1979*** 963998 50323 324730 325086 Nov 1979 708996 52518 390349 390349 Operational Year Total 10,276,920 5«,987 2,816,019 2,849,140 Grand Total 12,446,505 1,075,843 3,083,334 3,280,015 * Estimated ** TIROS-N Data Operational January 1, 1979 *** NOAA-6 Data Operational October 16, 1979 22 distribution (e.g., tropopause pressure and temperature). The number of clear radiance and atmospheric profile observations derived has averaged about 7,500 per day with a horizontal spacing of 250 km. With the opera- tional use of NOAA-6 TOVS data on October 16, 1979, the daily total doubled to 15,000. However, by early November a hardware problem in the ground data handling system restricted the amount of TOVS data that could be processed. An effort was made to process as many orbits of TIROS-N data as possible to provide a nearly complete, 1-satellite, global coverage at the expense of losing many NOAA-6 orbits. The period November 23-28 was hit hardest with a total of 50 hours of missing TIROS-N data. Table 2 summarizes the NESS production of vertical temperature pro- files and clear radiances for the FGGE research data set. In order to provide FGGE with operational TOVS soundings for as much of the Operational Year as possible, a special effort was instituted to proc- ess TOVS data from January 1 through February 28, 1979, in a delayed "catch- up" mode. Raw TOVS data which had been archived during this period were proc- essed through a parallel operation from March 16 to May 4, 1979, which pro- vided operational sounding data for FGGE as though it had been acquired in real-time processing. This special effort also allowed NESS to test its data handling and computer facility for the dual TIROS-N/NOAA-6 satellite operation. During the start-up of any new complex system of satellite instru- ments and ground data handling hardware and software, various problems and errors will arise which will affect the final output products; such was the case with TIROS-N. The scheduled launch of TIROS-N was delayed several months; however, the FGGE time schedule did not slip. NESS scientists had to compro- mise with a sounding system placed into an operational state without benefit of several months of problem-solving time, in order to meet the Level Il-b data delivery schedule. Two proposed operational sounding parameters (tropo- pause and ozone parameters) were deleted from the FGGE data set for the dura- tion of the Operational Year because of the lack of time for complete test and evaluation. Other data and systems problems resulted in the production at specific times of unusable or unreliable products which were also deleted from the final data set. These items are listed in Table 3. 4. DATA MANAGEMENT OF U.S. OPERATIONAL SATELLITE PRODUCTS FOR FGGE RE- SEARCH All U.S. operational satellite products to be archived for FGGE re- search were to be delivered to the FGGE Level Il-b Space-Based and Special Observing Systems Data Center in Norrkoping, Sweden. The medium of exchange was standard computer-compatible magnetic tapes structured according to the formats for the International Exchange of Level II Data Sets as established by the WMO (see Volume 3, FGGE Implementation/Operations Plan - FGGE Data Manage- ment Plan; Appendix 10). The original, agreed-upon shipment schedule to the Swedish Data Center was to have in the mail by the fifteenth of a month all data for the previous month (example: by May 15, 1978, ship all data for April 1-30, 1978). In most'cases, this schedule was met during the Build-up Year with GOES and NOAA-5 data. By end of the Build-up Year, NESS computer resources 23 Table 3. — Deleted data and precautionary notes for TIROS-N and NOAA-6 sounding data in the level Il-b data set. Time period of FGGE Level lib archive tapes from NESS for 1979 Items Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. t . Deleted all tropopause parameters 2. Deleted all ozone parameters 3. Deleted all layer mean temperatures above 100 mb, if the observation is poleward of 74°N and 74°S 4. Deleted 1 1 precipitable water parameters 5. Deleted complete observation if between 60°S-75°S latitude and MSU channel 4 temperature is less than 219 K 6. Deleted complete observation if between SS'S-eCS latitude and MSU channel 4 temperature is less than 215 K 7. Precautionary note: Possible low tempera- ture bias r.c r surface and in regions of persistent lOw-level cloudiness 8. Deleted complete observations because of bad data between these times: 18 July 1500 GMT-1 700 GMT 24 July 1400 GMT-1600 GMT 29 July 1300 GMT-1500 GMT 1 August 0900 GMT-1 130 GMT 9. Precautionary note: A few bad retrievals may be present due to an ingest problem 10. Precautionary note: Some observations may be mislocated up to 70 km due to TIROS-N spacecraft attitude problem. 4 Nov. 0930 GMT-8 Nov. 1523 GMT 11. Deleted all NOAA-6 temperature data above 3 mb level 12. Precautionary note: Numerous data cover- age gaps occur due to ground system hardware problem 23-28 Nov. 13. Precautionary note: Possible low temper- ature bias in soundings derived from cloudy microwave retrievals made in pre- cipitation / Jan.-end of FGGE I Jan.-end of FGGE I Jan.-ll Feb. I Mar.-ZI Apr. 1-21 Jan. I Mar.-N Apr. 15 Apr.- 5 May 29 Apr. -26 May I Jan.-15 Jul. . .^ . . . A . . . A . . . A / Jul.-3 Aug. 4-8 Not. 16 0,1.-30 Nov. 23-2X Nov. I Jan.-end of FGGE 24 were more difficult to obtain because of check-out of the TIROS-N operating sys- tem. The shipment schedule was modified so that the shipment of data tapes would be made by the end of the month instead of the fifteenth. This modifica- tion allowed greater flexibility within NESS for tape production but still was well within the Swedish Data Center processing schedule. The new schedule v/as met throughout the Operational Year for the GOES cloud motion vector and TIROS-N sea surface temperature products. The TIROS-N clear radiance and sounding pro- ducts encountered the following numerous problems which caused extended delays in the scheduled shipments: The delayed start of the TOVS operational system, the catch-up processing, NESS archiving software problems, a tape format change requested by the Swedish Data Center which was not fully resolved until early June, and the need to reprocess all TOVS data from January 1, 1979. The proc- essing and shipping of TOVS data tapes gradually got back on schedule by the end of the Operational Year. A normal, monthly shipment of tapes during the Build-up Year con- sisted of six tapes: One tape each of cloud motion vectors and sea surface temperatures and two tapes each of clear radiances and soundings. With the start of TIROS-N operational data, one tape was produced weekly for each of sea surface temperatures, clear radiances, and soundings; the dual TIROS-N/ NOAA-6 operation raised the total to two tapes per week for both clear radi- ances and soundings. A number of tapes had to be re-sent during the Build-up and Opera- tional Years for a variety of reasons: Parity errors, unreadable tapes, for- mat overflows (asterisk-filled fields), and undelivered tapes. In many cases, the only action that had to be taken by NESS was to make a copy of the back-up tape in question and reship. Other problems usually involved minor computer processing and reshipping. One deviation from the FGGE Data Management Plan should be noted: The first test tapes sent to the Swedish Data Center were produced at 1600 bits per inch (bpi) instead of the required 800 bpi. The Swedish Data Center had no problem reading the tapes, so a bilateral agreement between NESS and the Swedish Data Center was reached that all U.S. operational satellite tapes would have a density of 1600 bpi for the duration of FGGE. This reduced substantially the total number of tapes which had to be shipped, especially of TIROS-N products. In addition to the regular shipment schedule, NESS also participated in the End-to-End Test for the first Special Observing Period. A NESS scien- tist hand-carried to the Swedish Data Center magnetic tapes of all NESS opera- tional products, including the first available TIROS-N products, for the period January 15-19, 1979. A few minor problems occurred in processing the data tapes at the Swedish Data Center; however, all of these were solved with minimal delay. The consensus of the participants in the data-gathering and merging phase of the End-to-End Test was that the FGGE Level Il-b data producers had reliable data processing operations, which would enhance the chances for suc- cess of the Global Weather Experiment. 5. CONCLUSIONS AND RECOMMENDATIONS The NESS participation in the Global Weather Experiment as the U.S. Operational Satellite Data Producer progressed quite smoothly through the Build-up and Operational Years despite the quantity of data that had to be 25 exchanged and the mid-stream switch-over to an entirely new polar-orbiting satellite system. Several items should be mentioned in retrospect, which may help in the planning of future experiments. 5.1 Level II data exchange formats It is highly desirable that the final data tape formats for an inter- national experiment be compatible for use on as many computer systems as possible. However, when only two data centers are involved in the exchange of data, especially in quantities associated with satellite data products, the Level II data formats used for the FGGE are rather inefficient from a computer system processing and a tape logistics point of view. If the two centers can agree on a higher density recording level and the use of binary recording code instead of EBCDIC code, fewer tapes would have to be produced, shipped, and stored, and less computer time would be required. Data producers may have a workable data format already in use which would be compatible with the receiving data center, thus saving time and money for both centers. 5 . 2 Acceptable error rate for exchanged tapes The data management plan for an experiment of the scope of FGGE should list the acceptable error rates for any given tape that will be tolerated at each data center before that tape is rejected and the data producer is asked to reprocess the data. The goal of each data producer should be 100% correct- ness; however, in processing millions of data parameters, a few inadvertent errors are bound to occur. The error rate could be as small as 0.01% of all ' observations on a tape before reprocessing is requested. During FGGE, several tapes were reprocessed by NESS in which only one or two parameters in one observation were incorrect out of a total of 100,000 observations. 5.3 The use of TELEX messages for tracking i In a large international experiment such as FGGE, communication among participants is always a problem. The use of TELEX messages for the expedient flow of vital information between NESS and the Swedish Data Center was greatly appreciated despite the high cost of the service. 5.4 Tapes lost in shipment It had been expected that some problems would exist in shipping magnetic tapes via international air mail, but this did not materialize. The few tapes that had parity errors or were unreadable may have been due to the passing of the tapes near magnetic fields of motors, but this would be very difficult to prove. Only two shipments of ten tapes each appeared to have been lost in the mail, but they eventually arrived after replacement tapes had already been shipped. These shipments had been delayed by Swedish Customs despite the regular customs form that had been attached to the box. Experi- ment planners should check early with all Customs Offices for proper clearance procedures. 26 CHAPTER 3 GOES, INDIAN OCEAN By F. Kahwajy (NESS) F. Mosher (U. Wisconsin) 1. - INTRODUCTION j'- - • Five geostationary satellites were planned for the FGGE operations, originally including a Soviet satellite at 60°E. The Soviets, however, in- formed the World Meteorological Organization (WMO) that their geostationary satellite would not be available in time to support FGGE. Contingency plans were then implemented to minimize the impact of the absence of the USSR satel- lite. This action resulted in a decision whereby the United States provided a replacement satellite which was then operated by the European Space Agency (ESA) for the operational year of FGGE. In implementing this back-up plan, the United States moved the GOES-1 satellite to a position over the Indian Ocean and provided an antenna system and electronic equipment for satellite acquisition and control. This equip- ment was installed at Villafranca, Spain, for ESA operation. The satellite data were recorded on digital video cassettes and computer-compatible tapes and shipped to the University of Wisconsin. The Space Science and Engineer- ing Center (SSEC) at Wisconsin processed the tapes and measured cloud drift winds which were sent to Sweden as part of the FGGE Level Il-b data base. The full resolution digital archive of the GOES-Indian Ocean images is stored at SSEC and is available to scientific investigators. The GOES-Indian Ocean performed well except for the problem of the infrared sensor having sporadic outages after March 1979. 2. GROUND STATION AND SATELLITE OPERATIONS The GOES ground station at the ESA Villafranca site consisted of a 13-meter S-band antenna system and two vans which contained the satellite control, data conditioning equipment, and the recording equipment. An antenna under construction to support another experiment was diverted to Villafranca. Much of the command and control system was borrowed from NASA and NOAA stations. Existing contracts were modified to include new equipment. Thus, a complete ground station including command and control telemetry receive and transmit, antenna receive and control, synchronizer data buffer, and recording equipment was assembled in less than six months. At Darmstadt, Germany, ESA made changes in their control center to accommodate the GOES satellite. New software was written using NOAA-suppl ied specifications to handle GOES telemetry and to provide a GOES orbital track- ing system. NOAA provided the U.S. management of this crash effort, while fund- ing was shared by the National Science Foundation, NASA, and NOAA. RF Systems Inc., installed the antenna system at Villafranca, while Westinghouse Electronic Corporation provided the system engineering, integration, and testing. SSEC provided the data recording equipment. The Air Force air- lifted the equipment to Spain. ESA prepared the site at Villafranca, includ- ing the provision of concrete pads for antenna and equipment vans, the pro- vision of electrical power, and the provision of communication links between Villafranca and the control center located at Darmstadt, Germany. A logistic system was established to provide spare parts from the United States, while other consumables such as magnetic tapes and archive cassettes were furnished by ESA. Manpower to operate and maintain the station was provided by ESA. 29 NOAA trained the station operational and maintenance personnel and helped ESA develop detailed procedures for the operation of the satellite and ground station equipment. The operational procedures were essentially the same as those for the U.S. operated GOES as described by Herkert, et al . (1975) and Jessie, et al. (1975). The GOES-1 was initially moved from its position at 135°W to about 10°W where both NOAA and ESA could operate the satellite. After ESA was fully trained in the satellite operations, the satellite was moved to 58°E over the Indian Ocean. It arrived in November 1978 and was fully operational in time for the start of FGGE. The normal operation schedule of the satellite had scans every half hour. Table 1 has the normal operations schedule, and Table 2 has the opera- tions schedule for June, July, and August 1979. In addition to the half-hour scans, there were periods of 15-minute rapid scans from 8:00 to 9:15 GMT and 20:30 to 21:15 GMT each day. These 15-minute scans went from scan count 300 to scan 1500 (approximately 43°N to 43°S) and were used for making mesoscale wind sets in the tropical regions. During the months of June, July, and August; rapid scans of 10 minutes between images were made from 6:00 to 9:20 GMT between scans 450 and 1250 (approximately 3rN to 22°S). These rapid scans were used for mesoscale wind sets in the tropical MONEX region and to measure sea surface stress using sun glint in the Arabian Sea region. The infrared sensors on the GOES-1 were calibrated weekly and en- coded using the NOAA standard table. Since the GOES-1 had been used for several years previously as an operational satellite, its infrared calibra- tion characteristics were reasonably well known. Prior to FGGE, a number of calibration encoding tapes were prepared. On a weekly basis, a computer tape which contained the calibration shutter information was shipped from Spain to NESS at Suitland, Maryland. The same calibration procedure as described by Bauer and Lienesch (1975) was used for GOES-1. The infrared calibration information was telexed to Spain where the correct calibration tape was loaded. Hence, the quality of the infrared sensor calibration was maintained throughout FGGE to the same degree as the other U.S. geostation- ary satell ites. The image data from GOES-1 \uere recorded on digital cassettes and computer tapes. The digital cassettes contained the full resolution output of the satellite. This prototype recorder was designed and built at SSEC by E. Suomi. It uses a modified Sony video recorder which can store 21 giga- bits on a single cassette. Five cassettes per day were used. In addition to the cassettes, computer-compatible tapes were made of reduced resolution image data using the Offline Data Ingest System (ODIS) supplied by SSEC. The ODIS had a TV monitor in addition to the tape capability so one could see what data were being recorded, played back, or just being sent from the satellite. The archives were quality controlled by playing them back into the ODIS and checking the image quality. The ODIS was also used to record reduced resolution data for use by cloud tracking groups at SSEC in the U.S., Deutsche Forschungs- und Versuchsanstalt fur Luft- und Raumfahrt in Germany, >■;■ 30 Table l.--Mormal GOES-1 FGGE VISSR operation schedule (Note: See Table 2, this chapter, for June, July, and August 1979 schedule.) TIME VISSR action . TIME VISSR action GMT GMT 0000 Full scan 1200 Full scan 0030 Full scan 1230 Full scan 0100 Full scan 1300 Full scan 0130 Full scan 1330 Full scan 0200 Full scan 1400 Full scan 0230 Full scan 1430 Full scan 0300 Full scan 1500 Full scan 0330 Full scan 1530 Full scan 0400 Full scan 1600 Full scan 0430 Full scan 1630 Full scan 0500 Full scan 1700 Full scan 0530 Full scan 1730 Full scan 0600 Full scan 1800 Full scan 0630 Full scan 1830 Full scan 0700 Full scan 1900 Full scan 0730 Full scan, retrace to 300 1930 Full scan 0800 Scan to 1500, retrace to 300 2000 Full scan, retrace to 300 0815 Scan to 1500, retrace to 300 . 2030 Scan to 1500, retrace to 300 0830 Scan to 1500. retrace to 300 2045 Scan to 1500, retrace to 300 0845 Scan to 1500, retrace to 300 2100 Scan to 1500, retrace to North Limit 0900 Scan to 1500, retrace to North Limit 2130 Full scan 0930 Full scan 2200 Full scan 1000 Full scan 2230 Full scan 1030 Full scan 2300 Full scan 1100 Full scan 2330 Full scan 1130 Full scan Table 2.--G0ES-1 FGGE VISSR operation schedule for June, July, and August 1979 TIME VISSR action TIME VISSR action GMT GMT 0000 Full scan 0840 Scan to 1250, retrace to 450 0030 Full scan 0850 Scan to 1250, retrace to 450 0100 Full scan 0900 Scan to 1250, retrace to 450 0130 Full scan ' 0910 Scan to 1250, retrace to North Limit 0200 Full scan 0930 Full scan 0230 Full scan 1000 Full scan 0300 Full scan 1030 Full scan 0330 Full scan 1100 Full scan 0400 Full scan 1130 Full scan 0430 Full scan 1200 Full scan 0500 Full scan 1230 Full scan 0530 Full scan, retrace to 450 1300 Full scan 0600 Scan to 1250, retrace to 450 1330 Full scan 0610 Scan to 1250, retrace to 450 1400 Full scan 0620 Scan to 1250, retrace to 450 1430 Full scan 0630 _ I'^l"- to 1250, j2e_tr£C^ to 450 1500 Full scan ►0640 Scan "to 1250, retrace to 450 1530 Full scan •'oeso Scan to 1250, retrace to 450 1600 Full scan "0700 Scan to 1250, retrace to 450 1630 Full scan "0710 Scan to 1250, retrace to 450 1700 Full scan "0720 Scan to 1250, retrace to 450 1730 1800 Full Full scan scan or alternatively for calibration 1830 Full scan \ 1900 Full scan "0640 Scan to 1250, retrace to North Limit 1930 Full scan "0700 Scan to 1250, retrace to 450 2000 Full scan, retrace to 300 "0720 Scan to 1250, j^et^ra^c^ to 450 2030 Scan to 1500, retrace to 300 0730 Scan' 'to 1250, retrace to 450 2045 Scan to 1500, retrace to 300 0740 Scan to 1250, retrace to 450 2100 Scan to 1500, retrace to North Limit 0750 Scan to 1250, retrace to 450 2130 Full scan 0800 Scan to 1250, retrace to 450 2200 Full scan 0810 Scan to 1250, retrace to 450 2230 Full scan 0820 Scan to 1250, retrace to 450 2300 Full scan 0830 Scan to 1250, retrace to 450 2330 Full scan 31 and Laboratoire de Meteorologie Dynamique in France, calibration data for NOAA, and navigation data for the satellite attitude determination for SSEC. The digital cassettes are considered a Level I type archive, and digital computer tapes from these cassettes are available from SSEC. Log books of the GOES-1 operations and calibration during FGGE will be kept at SSEC in Madison, Wiscon- sin, after the completion of FGGE along with the digital image archive. The ODIS tapes were recycled during FGGE. During FGGE there were several minor problems with the GOES-1 data, and one major problem -- the intermittent failure of the infrared sensor. The first failure of the infrared sensor occurred on 24 March 1979. The infrared sensor came back on and then failed again several times during FGGE. Table 3 lists the dates when the sensor was on or off. The failure of the sensor was characterized by a slow drift of the data toward either all black or all white. This drift would last for several hours. During this drift toward failure, the calibration was worthless. When the sensor came back on, the calibration appeared to be correct. This failure of the infrared sensor caused problems in tracking clouds, because the cloud height determination and the cirrus cloud tracking were heavily dependent on infrared data. The TIROS-N data were used at SSEC to measure cloud heights during times of the GOES-1 infrared failures. This processing will be described in a later section. The cause or a fix for the infrared failures was never determined, even though a considerable amount of effort by NESS, SSEC, and NASA was applied toward this problem. There were several minor problems with the GOES-1 data. These were generally noted on the log sheets which accompany the archive. The period of December-February had a minor problem with dark segments being in the visible data. The period of mid-December to mid-January had repeated segments 8 pixels long on the visible data. During the spring eclipse period, the satellite was put in the analog sun pulse mode which caused a periodic image distortion of up to six visible pixels. The only minor problems which affected the wind determination efforts were the repeated segments and the analog sun pulse image wiggle. These problems added 3-4 m/sec to the random error of the cloud tracked vectors. All of the minor problems were corrected after they were discovered. 3. CLOUD TRACKING OPERATIONS AT SSEC 3.1 Operations Plan The archive and ODIS tapes were shipped from Villafranca to the Space Science and Engineering Center at the University of Wisconsin. There the data were processed for cloud tracked winds which were then shipped to the Level Il-b center in Sweden. The operations plan at SSEC is detailed in the U.S. FGGE Data Management Plan, Vol. 2, Annex 3. SSEC produced three types of cloud track wind vectors during FGGE. The first type was a tropical high density wind set in the regions from 15°N to 15°S, utilizing two U.S. geostationary satellites (GOES-East and GOES- West). This wind set was referred to as the "Tropical wind set". The second was a macroscale wind set utilizing the full disk images from the geostationary satellite over the Indian Ocean. This data set was comparable 32 Table 3.--IR sensor functioning of Indian Ocean GOES during the FGGE period Date On Off Intermittent Date On Off Intermittent 1978 ., ■ v-',' -•■' 160-161 • X 335-365 X ' " . ■' M !'■ ",■-■' ■' ■■' 162 X 1979 ,; ■; , :' ,,;■.'• 163-173 X 001-057 X 174 X 058 X 175-176 X 059-075 X (Only 7 visible channels) 177 ' X 076-082 X 178-180 X 083 X 181 X 084-085 -- X,' -■ 182-290 X 086-089 X 291 X 090 X 292-296 X 091-117 X 297 X 118 X 298-299 X 119-125 X 300 X 126-128 X 301-304 X 129-130 X 305 X 131-145 X 306-307 X 146-151 X ■ ■ ■ 308 X 152 X 309-311 X 153 X 312 X 154 X 313-331 X 155 X "; . ' ,. ■■ 332 ! ^, X 156-158 X ■ V 333 X 159 X 334 X 33 in coverage to the cloud wind sets produced operationally by NESS. This data set was referred to as the "Indian Ocean wind set". The Tropical wind set and Indian Ocean wind set were produced for the entire FGGE Operational Year. The third type of cloud track vectors, referred to as the "MONEX wind set" was a high-density wind set using images from the geostationary satellite over the Indian Ocean. The coverage was approximately 30°N to 20°S, depending upon synoptic conditions for a 100-day period beginning with 1 May 1979. The tropical wind set was measured at the 18 GMT synoptic time slot, the Indian Ocean set at 00 and 12 GMT (except when the infrared sensor failed when only the 12 GMT was available), and the MONEX winds at the i06 and 18 GMT time period (except when the infrared sensor failed when only the 06 GMT was avail- able). The average daily number of winds produced for the different wind sets is as follows. For the tropical wind set, the average daily number of winds v/as approximately 1450. For the Indian Ocean wind set, the average daily number of winds during 1 December - 30 April 1979 was approximately 1350. Because of the numerous failures of the infrared sensor between 1 May - 30 No- vember 1979, SSEC maximized its efforts to derive winds from visible images centered near 0900 GMT and so it is not possible to clearly separate the num- bers of winds produced for the Indian Ocean and MONEX wind sets. Thus, the combined Indian Ocean wind sets averaged approximately 1700 winds during the period 1 May - 8 August. During the period 9 August - 30 November, the average daily number of winds for the Indian Ocean wind set was approximately 1250. A general description of cloud tracked winds, the processing, and the problems is contained in Mosher (1978). 3.2 Cloud Tracking The cloud motions were measured on the McIDAS (Man-computer Inter- active Data Access System) (Chatters and Suomi (1975), Smith (1975)). This is an image storage, display, and processing system consisting of data ar- chive, data access, video display, operator console, and computer control sections. Central to the system is a computer which controls the display section, operator console, and computer peripherals. Data enter the system either from an antenna on the roof which receives the stretched SMS data, the archive tapes, or computer tapes such as the ODIS. The real time or archive ingestion of data is done by using a data interface box which con- verts the incoming visible and IR data into 8~bit bytes, averages the elements in a line to produce equivalent 1/2, 1, 1-1/2, 2, 3, and 4 mile resolution data, packs the data into 24-bit words, and then puts the data directly onto the digital disk. The Indian Ocean processing was done using 3-mile resolution data with 30 minutes between images. The Tropical data set was done using 2-mile data with 15 minutes between images except in areas of active convec- tion where 1-mile data with 7-minute intervals were used. The MONEX data set used 2-mile, 15- (or 10-) minute data. Using the McIDAS, it is possible by simple key-ins to enhance an image, magnify it, combine adjacent images into loops of any length, locate and track clouds, and display the results as a vector plot superimposed on the original image. Two independent signal systems allow double looping of infrared and visible images, with instant single key transfer from one to the other. 34 Tracking may be done by either of two primary methods: cursor tracking of the cloud to the nearest TV line and element (pixel tracking), and image match tracking of the cloud to better than TV line-element resolu- tion (correlation tracking). Pixel tracking has been facilitated by the addition of a function called the velocity cursor. The operator positions a cursor over the cloud to be tracked using a joy stick. The velocity cursor function then automatically displaces the cursor from one picture to the next according to the position of a second joy stick. The velocity cursor can be used by itself for single pixel tracking, or it can be used in conjunction with the correlation tracking. Correlation tracking requires the operator to roughly track the cloud by placing the cloud within a box for each pic- ture in a set. The computer then performs a correlation analysis to align the brightness field and "fine tune" the operator's tracking. Correlation tracking is the more accurate, but it requires well-defined clouds moving in a single layer flow pattern. Single pixel tracking using the velocity cursor can be invoked by the operator for tracking clouds in multi-layer flow patterns, or for matching the motion of a pattern if individual clouds cannot be tracked. i 3.3 Image Alignment Image alignment was done using a satellite navigation system, de- veloped at Wisconsin by Phillips (1974), which models the satellite orbit and attitude and uses this information in the transformation from satellite to Earth coordinates and vice versa. The orbits used for the Indian Ocean data were those measured by ESA. The attitude of the satellite was deter- mined from land mark measurements at Wisconsin. The accuracy of the naviga- tion is generally better than a full resolution pixel for relative alignment between images, and several pixels of absolute location. Cloud tracking was done using 3 images so that a single cloud motion can be measured twice as a quality control measure. The systematic difference between the velocity measured on the T1T2 sequence and that measured on the T2T3 sequence is a measure of the error due to the relative alignment of the images. For a sampling of data sets between January and July 1979, 53% had residuals differences of .5 m/sec or less; 90% had residuals differences of 1 m/sec or less; 5% had residuals differences greater than 2 m/sec. 3.4 Cloud Height Determination Cloud heights were determined from the infrared temperatures and temperature profiles. When the infrared sensor failed, the TIROS-N sounding channels were used for height determination. This will be described in a later section. The visible-infrared cloud height algorithm (Mosher, 1975) used visible data to determine optical thickness from which infrared emissivity and physical thickness can be inferred. At the beginning of FGGE, the de- rived infrared emissivity was used to recalculate the infrared temperature. The derived temperatures, however^ showed sporadic gross errors. The recalcu- lated infrared temperature portion of the cloud height algorithm was removed 35 in January 1979. The remaining algorithm had a two part decision tree. If the emissivity calculation showed the cloud to be a blackbody, an average tempera- ture of the cloud was used. If the cloud was not a blackbody, the coldest infrared pixel was used. During the end-to-end test of the SOP-1, it was noted that the cloud tracked winds from SSEC were on the average colder than those of NESS for the tropical wind sets where there was overlapping coverage. An in- vestigation of the differences was made using both height determination systems on the same clouds. This confirmed the differences. The SSEC's method aver- aged 1.5°C colder than NESS' method. The cloud height algorithm was changed from picking the coldest IR pixel to picking the coldest value in an infrared histogram which has more than two pixels in it (the NESS algorithm). This eliminated the differences. The revised algorithm was installed at SSEC during April 1979. The work of Hasler, et al . (1977) has shown for trade cumulus that the wind at cloud base is most closely related to the drift of the cloud. The cloud height algorithm used the estimated cloud thickness derived from the optical thickness and put the wind level at the cloud base for clouds with bases below 850 mb. However, the temperature reported with the wind was that of the cloud top. The temperature profile used for all of FGGE was the climatological profile described in Mosher (1975). The original FGGE plans called for using the NMC analysis for the profile, but a series of problems prevented the in- stallation of the computer link to NMC until the end of FGGE. While the cli- matological profile works well in the tropical oceans, midlatitude regions with synoptic conditions of strong troughs or ridges can have height errors of up to 100 mb. Cloud heights were either automatically determined with each wind measurement, or were manually assigned by the operator. The manual height assignment mode was generally used when there were multiple cloud levels, thin cirrus, or TIROS-N derived cloud heights. The manual height assignment method of Hubert (1975) was used in that a "representative" cloud height would be measured and this height would be assigned to a "fleet" of vectors from clouds in the general vicinity with similar motion characteristics. In the Indian Ocean data set, there were an average of approximately 18 vectors per fleet. 3.5 Cloud Selection . . . Not every cloud that appears to move is a valid tracer of winds. Gravity waves, orographic clouds, thunderstorm cores, etc. are not repre- sentative of the wind field. Thunderstorm outflow, sea breezes, etc., while being representative of mesoscale wind flow, are not representative of synop- tic scale flow. The FGGE cloud trackers were meteorologists (with at least a B.S. degree) who had enough synoptic experience to recognize when flow patterns were associated with synoptic scale systems. In the training for FGGE, the cloud trackers were instructed to ignore flow patterns, such as thunderstorm upper level divergence, if it influenced an area smaller than 200 km square. If features such as cloud clusters in the tropics were pre- sent which had an influence radius larger than 200 km and were locally 36 divergent, the operators were instructed to measure at least four vectors around the system. If the operator could not measure vectors all the way around the system, he was instructed to measure only one vector which showed the large scale flow. The operators were trained to ignore features such as gravity waves and orographic clouds. They were told to concentrate on the meteorologically significant systems such as jets, storms, etc. They tried to distribute the vectors within the constraints of cloud availability to obtain reasonably uni- form coverage. The operators were instructed to attempt to measure midlevel clouds, especially in the tropics. However, plots of the cloud winds show that there were generally not enough midlevel clouds available to adequately depict the midlevel circulation patterns. 3.6 Quality Control The quality control procedures outlined in Appendix A of Volume 2, Annex 3 of the U.S. FGGE Data Management Plan were used except that overlays of conventional data measurements were not available. The operators did have fax maps of the NMC global analysis available for reference in the Indian Ocean processing. Three images in a sequence were always used in order to allow the cloud to be tracked twice and compared for consistency. If for some reason one of the three images at the scheduled times was unusable, the sequence time was shifted until a sequence of three could be obtained. The satellite images were sectorized into TV-sized areas for wind processing. For the Indian Ocean, the Earth's disc was divided into 16 sectors for processing. As the winds were measured, the vectors were displayed on the graphic overlay and edited when necessary. Figure la shows the visible image of a sector over India on 5 May 1979, 0930 GMT, and the measured low level wind vectors. Figure lb shows the infrared image and the high-level flow of the same region. 3.7 Summer MONEX Wind Set ,^ . ., In addition to the normal FGGE Indian Ocean wind set, an enhanced wind set was produced during the summer MONEX period. The area of enhancement was from approximately 30°N to 20°S, 20°E to 95°E. The time period of enhance- ment started on 1 May 1979 and lasted for 100 days. The operational procedure of the MONEX enhancement was to measure the large scale flow using 3-mile data with half-hour time resolution. Then an enhanced wind set was produced using 2-mile data and lO-or 15-minute time resolution on the rapid scan images just prior to the normal wind images. The large scale vectors were displayed on the rapid scan images, and the enhancement was done by filling in the voids around the previously measured vectors. Since the large scale winds were measured generally on the 0930, 1000^ and 1030 GMT images, they were included in the 12Z reporting period for FGGE. The rapid scan enhanced wind sets were generally at 0830, 0845, 0900 GMT or 0850, 0900, 0910 and were included in the 06 GMT reporting period. When the infrared channel was available, a similar situation occurred with the normal winds in the 00 GMT period and the MONEX winds in the 18 GMT period. Even though these MONEX and Indian Ocean wind sets were put in separate reporting periods, they should be considered as part of the same data set because there is less than an hour between them. 37 f^ f ■ : ;^^Vit-/*- ^ *f ' . ■V ''^^^^^^^^HE ^ .. N -c^K^ 0.' ^ ^^ -- 1^ yLJ^^^^^^^^P' ^-'■' ^ ., xT^^ ^^ ?^.v^^ "^ ^B- ^. :„^-^^ '^\.b^ 1 ;';^j|A<^-^ A', i^ ^9 '^^BRjii,''^^-^^ ^"^ * ■asw-'^'^l'li^. '^^\»^^^- m -^■- jjj ^- -i^WL^. ' '^t^'- .^iwi,- \«, .' ^BBBp' ii^ - ^3psiN*>^^^ *~ ^^ ^— - ' \<1U- ^^Bl':^' ,,^sif^'' ^*.-^- li^ ^ **WI Figure la. --Visible sector over India on 5 May 1979 showing the low-level cloud wind field. Figure lb. --Infrared sector over India on 5 May 1979 showing the high-level cloud wind field. 38 3.8 TIROS-N Cloud Heights The GOES-1 infrared sensors failed several times during FGGE (see Table 3 for dates). The infrared channel provides several functions. It allows tracking of clouds at night. It is very helpful in the detection and tracking of cirrus. Finally, it is used to obtain the heights of the cloud tracers. Without the infrared sensor, it was impossible to obtain winds at night for the GMT data set. During the daytime hours, there were several options available: guessing at the cloud heights (high or low as was done in the early days of cloud tracking with ATS), obtaining cloud height information from a different data source, or not processing the data. The second option was chosen by obtaining cloud height information from the TIROS-N HIRS and microwave sounding channels. The cloud heights were obtained using the multi spectral CO2 absorption method of Menzel , et al . (1978). The cloud pressure is determined from the ratio of the deviations in cloud produced radiances and the corresponding clear air values for two or more spectral channels. The method works well for all types of clouds, includ- ing thin cirrus. The vertical resolution archived is roughly 50 millibars. The operational procedure had the cloud tracker first process the TIROS-N orbits by identifying cloud features and measuring several heights for each of the major cloud features. These heights were stored in a file along with the location of the measurement. Next the GOES visible cloud tracking sequence was displayed. The TIROS-N cloud heights were plotted on the GOES images at the location of the measurement. Since there was a time difference of up to three hours between the overflight of the TIROS-N and the GOES se- quence, the plotted cloud heights did not always fall on top of the cloud. The cloud tracking meteorologist would make a subjective judgment on what cloud heights went with what clouds. A short study was done to see if the TIROS-N cloud heights differed significantly from the normal GOES heights. For one day (6 May 1979) when the TIROS-N data were available and the infrared sensor was working correctly, the normal procedure of selecting clouds on the TIROS-N data and displaying the results on the GOES images was followed. The operator selected the cloud to which he would have assigned the TIROS height and then measured the cloud height using the GOES infrared data. The mean difference between the two data sets was 13 millibars (TIROS being lov/er) with a standard deviation of 66 millibars. The sample size was 87 cases distributed over the Indian Ocean. The TIROS cloud top temperature was an average of 0.8°C warmer with a standard deviation of 7.8°C. 4. SUMMARY AND CONCLUSIONS The geostationary satellite coverage of the Indian Ocean during FGGE was successful. In less than a year a joint effort by NOAA and ESA was able to configure a ground station in Spain for operation by ESA. The GOES-1 was moved over the Indian Ocean. SSEC at the University of Wisconsin provided the archive and data processing support for the cloud tracking operations for the 39 FGGE Level Il-b data set. The full resolution digital image archive of the entire FGGE year of the GOES-Indian Ocean, GOES-East, and GOES-West is avail- able at SSEC. Computer tapes from this archive are available from SSEC. The only major problem with the GOES-Indian Ocean program was the " intermittent failure of the infrared sensor after March 24, 1979. A total of 159 days have no infrared data and an additional 18 days have only partial coverage during the day. The infrared sensor failed on eleven different occa- sions. During the times of no infrared coverage, the cloud heights were ob- tained from TIROS-N data and the cloud tracking was done on the visible images. Analysis of TIROS-N cloud height process showed no significant bias in height assignment as compared to the infrared derived cloud heights. The only notice- able deficiency of the cloud wind data set without the infrared data was a tendency for the jet cores to be poorly defined and the wind maxima to be missed occasionally. v The FGGE Level Il-b data set for the Indian Ocean contains over 510,000 cloud tracked vectors. An additional 529,000 vectors were generated at SSEC as part of the GOES-East and West tropical data set. The GOES Indian Ocean program depended heavily on international and interagency cooperation. The successful completion of the program in spite of short lead times and equipment failures speaks highly of the many people all over the world involved with the program. '.' ) ' 40 CHAPTER 4 SOUTHERN HEMISPHERE DRIFTING BUOYS By E. Kerut (NOAA/NDBO) 1. INTRODUCTION This chapter summarizes the performance of the 64 U.S. drifting buoys deployed in the Southern Oceans during the Global Weather Experiment. Forty- six buoys were deployed by ship and 18 buoys were deployed by aircraft as part of the drifting buoy array established during the experiment. An examination of the buoy performance indicated that approximately 56 percent of the buoys were operational after six months, and 40 percent after one year. ' As part of the U.S. drifting buoy development program, extensive test- ing was performed to verify system performance prior to the start of the exper- iment. End-to-end systems tests were performed to establish system interface compatibility and to determine if corrections to production buoys were needed prior to buoy deployment during the experiment. Data quality analyses were performed on systems prior to and during buoy deployment periods. ' o- '• An extensive test and evaluation program preceded the deployment of operational buoys. Pre-FGGE system end-to-end testing was conducted to uncover system level problems before the shipment of production buoys from the manu- facturer for deployment in the experiment. Predeployment and deployment test- ing was performed to evaluate the quality of the data from each of the buoys being deployed. A description of each test and the evaluation results follows. Finally, buoy performance during the First and Second Special Observing Periods (SOPs) is summarized. '•' - Central to the drifting buoy system is the utilization of the polar orbiting ARGOS satellite system for data telemetry and position location. Oper- ated by the French Centre Nationale d'Etudes Spatiales (ONES), the ARGOS Data Collection System (DCS) provides a means to locate and/or collect data from fixed platforms and moving buoy and balloon platforms. Buoy data were acquired through ARGOS when the TIROS-N and NOAA-6 satellites were in view. All data were collected at control data acquisition stations in Gilmore, Alaska^ and Wallops Island, Virginia; and then via landlines transmitted to the spacecraft opera- tional control center in Suitland, Maryland, for preliminary processing. The data were then transmitted to the ONES in Toulouse for processing by Service ARGOS. Processed data were then disseminated as appropriate. Meteorological data was disseminated worldwide on the Global Telecommunications Service (GTS). 2. PRE-FGGE TEST AND EVALUATION The primary objective of pre-FGGE test and evaluation programs was to conduct end-to-end system tests prior to the shipment of production buoys from the manufacturer. The tests were performed on development/prototype buoys in an effort to determine difficulties or problems attributable to buoy components/ subsystems that could still be corrected on the production systems. The tests included system checkout of the drifting buoys, satellite, communications links, data processing, procedural matters, and the training of personnel. Buoy sensor and position data were processed by Service ARGOS in Toulouse, France, and the results compared with data from a ground-truth system developed by the NOAA Data Buoy Office (NDBO). 43 Table 1 .—Statistics on buoy sensor performance for barometric pressure and water temperature. Barometric Pressure (mb) BUOY 1641 BUOY 1642 BUOY 1643 OVERALL - Mean Difference Between Buoy and Ground Truth 0.34 0.87 0.83 0.73 Standard Deviation 0.48 0.43 0.46 0.46 No. Observations 66 107 115 288 Water Temperature CC) 0.15 0.15 0.40 0.25 Mean Difference Between Buoy and Ground Truth Standard Deviation 0.32 0.17 0.17 0.21 No. Observations 49 106 107 262 •"( 1 44 Three prototype buoys were evaluated during a six-week intensive test period. The following buoy performance characteristics were measured and evaluated: Performance and accuracy of the two primary meteorological sensors (barometric pressure and sea surface temperature) Performance and accuracy of the position location system Number of actual and good data transmissions per day ' " Total elapsed time for various types of data received from Service ARGOS and processed or analyzed by NDBO. 2.1 Buoy Sensor Performance ' Buoy data received through the TIROS-N satellite were compared with ground-truth data. The buoys under test were equipped with Paroscientific baro- metric pressure sensors with a specified accuracy of +^ mb with a range of 900 to 1050 mb. The digital resolution (one count) was 0.15 mb. The ground-truth pressure system included three Rosemount transducers with a specified accuracy of +^0.6 mb over the range of 900 to 1050 mb. The ground-truth pressure sensors were averaged to provide a more accurate reference. The statistics on buoy pressure sensor performance are summarized in Table 1. All three buoys showed a positive mean difference when compared with the ground-truth reference stand- ard. The range of pressures measured during the tests was from 1004.2 to to 1032.2 mb. The post-calibration check of the ground-truth barometers indicated that the ground-truth pressures averaged 0.8 mb low over the range of pressures measured. The test buoy pressures fell within the specified accuracy with the ground-truth correction taken into account. The buoy water temperature sensors were Yellow Springs Instruments thermistors with a specified accuracy of j^l°C with a range of -5° to +35°C. The digital resolution (one count) was 0.16°C. The ground-truth temperature sensor system included several Action Instruments platinum-resistance trans- ducers with a specified accuracy of +0.4°C over the range of 0° to 30°C. The ground-truth sensor readings were averaged to provide an accurate temperature reference. The buoy water temperature measurements statistics are summarized in Table 1. The data fell within the specified accuracy range. 2.2 Buoy Location System Performance The buoy location system performance was evaluated in a controlled test. Five buoys were moored at fixed locations. Data were collected from the buoys starting two weeks after the launch of TIROS-N on October 13, 1978, and continuing through December 12, 1978. During this period, 224 position fixes were selected from the DISPOSE file obtained from Service ARGOS. The location data can be summarized as follows: 45 The mean radial error for all fixes from the actual location was 0.26 km. The standard deviation of mean radial error was 0.16 km. The largest radial error was 2.15 km. More than 96 percent of all position fixes were within 0.72 km. The results of the controlled tests indicated that the position- fixing system is accurate, reliable, and virtually error-free. The accuracy and reliability of the ARGOS location system well exceeded all initial esti- mates. It proved to be reliable, accurate, and suitable for use whenever accuracy to within 1 km was required. , In the effort to effectively evaluate the overall system performance early in the experiment, a large data set was collected and analyzed. Statis- tics on the number of position fixes were analyzed. Time delays in obtaining, transmitting, and processing buoy data were analyzed statistically to provide estimates of minimum and maximum delays for data to be updated on the computer files of Service ARGOS. Estimates of delays in accessing data from the files were also made using remote terminals. .. , 2.3 Data and Buoy Location Passes A simplified approximation of the average number of satellite passes per day in view of a buoy is shown in Figure 1. The number of passes depends on the latitude of the buoy and the minimum elevation angle at which the satellite can receive the buoy data. Using operational U.S. FGGE buoy data, the mean number and standard deviation of both good data passes and buoy locations per day were calculated as a function of latitude over a one-month period for buoys located from 20°S to eS'^S latitude. These data are shown in Figure 1. Curve A shows the increase in the average number of good data passes per day as the latitude increases. Comparison with the curves in Figure 2 indicates that data receptions appear to be obtained for buoy elevation angles as low as zero degrees. Curve B in Figure 1 shows the average number of good buoy locations per day as a function of buoy latitude. This curve follows curve A very closely. The buoys transmitted eyery 51.36 seconds, and at least three good transmissions are required to calculate buoy location. Figure 3 presents the same buoy location data in a different form. Curve C shows the percentage of data passes that resulted in a good buoy loca- tion. This percentage (70 percent) is independent of buoy latitude. Curve D shows the percentage of satellite orbits per day that resulted in a good buoy location. This curve is similar to curve B of Figure 2. 46 11" 10" 9 " 7 ■• < 12 Q K U Pi CO o < o >^ o ;d m O CO < ft < < Q O d z a > <; 4 •- 3 •- I"- ■ NO. OF DATA PASSES • NO. OF LOCATIONS 20 30 40 50 60 70 LATITUDE rS) MEAN NO. OF STANDARD MEAN NO. OF STANDARD LATITUDE DATA PASSES/DAY DEVIATION BUOY LOCATIONS/DAY DEVIATION 20° 4.74 24" S 4.89 29" S 5.11 30»S 5.14 33'S 5.48 40»S 6.19 47»S 7.00 50»S 8.22 51.5»S 8.62 55* S 10.11 55. 5" S 10.07 60° S 10.63 64° S 11.37 0.94 0.97 0.89 0.79 0.70 0.56 1.00 0.58 0.79 0.32 0.47 0.56 0.63 3.37 3.63 3.67 3.52 3.85 4.37 4.74 5.07 6.00 6.41 7.00 8.30 9.00 0.97 0.49 0.68 0.70 0.82 0.84 0.71 1.04 1.07 1.34 1.07 1.49 1.07 Figure 1. --Average number of data passes and buoy locations per day as a function of latitude. 47 o o ffl u- o < Q U PL. en < o d > 11t lO" 6-- 4-- 3-- 2" 1-- 0° 10" 20* -»- ■f- -h -f- ■+ -H ■♦■ -K ■+■ ■+■ -♦- 4- H 5 10 15 20 25 30 35 40 45 50 55 60 65 70 LATITUDE ("S) Figure 2. --Average number of satellite passes per day in view of a buoy at a given latitude for various angles above the horizon. T ', 48 100 - 90 •• w t 80 W K o U 70 fc hJ hJ 60 u H < 50 w b O 40 g u 30 u b: u Ah 20 10 ■• ■ % OF ORBITS IN SIGHT OF BUOY RESULTING IN A GOOD LOCATION • % OF TOTAL ORBITS PER DAY RESULTING IN A GOOD LOCATION D I I 20 ■»- + -»- + 30 40 50 LATITUDE (»S) 60 70 PERCENT OF TOTAL LATITUDE . DAILY OF 20«S 24.0 24«S 25. 8 29«S 26.1 30» S 25.0 33»S 27.4 40* S 31.1 47«S 33.7 50»S 36.0 51. 5* S 42.6 55" S 45.6 SS.S^S 49. 8 60»S 59.0 64»S 64. PERCENT OF ORBITS IN SIGHT OF TRANSMITTER 71.1 74. 2 71.8 68.4 70.3 70.6 67.7 61.7 69. 6 63.4 69.5 78.1 79.2 Figure 3. --Percentage of satellite orbits resulting in a good buoy location as a function of latitude. 49 2.4 Data Transmission Per Pass Service ARGOS data from the DISPOSE file were analyzed to determine transmission statistics on all passes for which buoy data were obtained. The average number of transmissions per orbit for passes over five different buoys at various latitudes were calculated, and the results are summarized in Table 2. The overall average was calculated to be 13.43 transmissions per orbit. The maximum number of transmissions received on any sample orbit was 19. 2.5 Time Delays in Data Transmission and Processing The U.S. FGGE buoy data transmissions stored in the Service ARGOS AJOUR file were analyzed to determine typical maximum and minimum delays in obtaining the most recent buoy data. The AJOUR file indicated when it was last updated, and that time was compared with the time that each buoy trans- mitted data. Also, the dead time in orbit from a particular location was calculated to estimate the average time required to transfer data from the satellite to Service ARGOS and to update the AJOUR file. Early results of this analysis for a one- satellite system were as follows: Maximum time delay - 10.35 hours Average of the maximum time delays - 8.46 hours Minimum time delay - 1.27 hours Average time for data to be transferred from the satellite to Ser- vice ARGOS and to update AJOUR file (no dead time in orbit) - 80 minutes. 3. PREDEPLOYMENT AND DEPLOYMENT DATA QUALITY ANALYSIS The objective of this effort was to evaluate the performance of each of the U.S. FGGE drifting buoys deployed in the southern Pacific Ocean during the November 1978 to February 1979 time period, extending from buoy activation to just after launch. Forty-six buoys were deployed by five ships. Buoy data obtained from Service ARGOS were evaluated daily and compared with ground- truth data reported by each ship. The evaluation period started at the time each ship left port and continued until after the buoy had been launched and ground-truth data had been compared with buoy data received via the satellite. Shipboard ground- truth data received via radio communication links were compared and evaluated daily with buoy data obtained from Service ARGOS. If buoy sensor data were questionable or out of tolerance, NDBO management and the U.S. FGGE Project Office were notified immediately so that a decision could be made whether to deploy the buoy, substitute another buoy, or forego deploying a buoy at that location. Also, in certain instances, instructions concerning buoy operation or maintenance were sent to the ships. 50 Table 2. --Average number of transmissions per orbit for data passes over buoys at various locations. AVERAGE NUMBER OF TRANSMISSIONS STANDARD NUMBER OF BUOY ID LATITUDE PER ORBIT DEVIATION ORBITS SAMPLED 1602 44 "S 13.07 4.19 100 1608 22"$ 13.76 3.97 50 1611 66°S 14.24 4.08 102 1621 45°S 12.93 4.58 60 1634 32 °S 13.07 4.30 96 Overall 13.43 4.36 403 51 Each ship was requested to activate its buoys and to provide an up- date of the launch schedule by buoy ID and location well in advance of the time that daily buoy checkout reports were required. There was some reluc- tance on the part of the ships to activate the buoys, since they transmitted at a frequency of 401.65 MHz, which is slightly above the frequency of some of the shipboard navigation equipment. However, tests conducted by NDBO did not show an interference problem. Typically, several messages were sent to and received from each ship before the message procedure became routine. Messages were transmitted and received over the AUTODIM circuits using the Naval Oceanographic Office (NAVOCEANO) Communications Center located at the National Space Technology Laboratories (NSTL) near Bay St. Louis, Mississippi. Buoy data were obtained by accessing the computer at Service ARGOS. Calibration tables and other data had been previously sent to Service ARGOS. The system periodically received satellite data and updated the AJOUR file with the latest buoy data. The Telex system was used as the primary means of accessing the AJOUR file. In most instances, this file was accessed at least once daily. • i. J The overall data flows are shown in Figure 4. Note that there are two completely separate flows for the data. The ground-truth data from each of the ships go through radio and hardwire links to the U.S. FGGE Project Office, with an information copy going to NDBO. The buoy data are stored in a tape recorder on the satellite, dumped to one of the satellite ground Con- trol and Data Acquisition (CDA) stations, and transmitted to the Data Pro- cessing and Services System (DPSS) at the National Environmental Satellite Service (NOAA/NESS) and then Service ARGOS, where the data are processed. The latest data are stored in the AJOUR file, which is accessed by means of a dial-up terminal. In order to time-correlate the ground-truth and buoy data, ground-truth data were taken by the ships at the same time that buoy data were being received by the satellite. In addition, the AJOUR file was accessed and the buoy data obtained before the buoy data from the next satellite pass were used to update the file. The mean and standard deviation of the buoy-measured barometric pres- sures and sea surface temperatures were calculated daily for all buoys on board each ship. These statistics were compared with the ground-truth data. Typi- cally, there was better correlation among the buoy sensors than between the buoy sensor mean and the ground truth. As buoys were launched, fewer remained on the ship and more emphasis had to be placed on the ground truth and analyz- ing the data. Daily verbal and weekly written reports were provided to NDBO and the U.S. FGGE Project Office. Ground-truth pressure and temperature data were taken by ship per- sonnel to coincide with the closest morning and afternoon local satellite passes. Equatorial crossing time and longitude were provided by NOAA/NESS for each satellite orbit. Planned and updated buoy launch location and time data were requested and obtained from each ship. Using the NOAA satellite and buoy launch data, equatorial crossing time and longitude for the morning (descending) and afternoon (ascending) orbits that passed closest to the ship were determined and the data sent to each ship. Each ship determined the time 52 ZD uj O -I —3 —> Q. =) 1 _l _j «r el Z 1— 1 •— 1 Q z: q: O UJ CD H- Q Z o > oo oo y a: 00 - ' ' '''■,' ' ' •a: >- D 1 SL «=t o 2: o OO Q ct t— ' ■ ' ■; ' rr ^: o '. c; — _J ' :>- z *— • e> —I o i"-i 1— 1 U- 5 1— U- Q < UJ O oi>n C5 t— 1 ini U_ O zz =3 1-. UJ s: _i • -D y OO O f^ • q: CJ z:; Q- t/1 —1 O UJ o: I— •— < *=-'.' I— on CO o ■M (T3 -o >> S_ CU Q. •a i_ I I cn o => I— a: Q£ eC CD H- Q O h- =} <: CO o 53 correction for its location. Ground-truth data were taken at this time and at 30 minutes prior to and after this time. Ground-truth data were also taken 3 hours prior to launch, at launch, and at 1-hour intervals for 9 hours after launch. The procedures and steps required for the data analysis evolved dur- ing the program. The steps that were followed varied some from ship to ship, but in general followed the flow shown in Figure 5. Procedures were followed for monitoring buoys on a daily basis and during the launch sequence. Each ship was requested to provide the buoy deployment schedules, which also indi- cated which buoy (by Service ARGOS ID) was to be deployed at each position. Each ship was also requested to activate all of its buoys shortly after the ship left port. The AJOUR file at Service ARGOS was accessed daily and all buoy data that had been updated since the last update were obtained. A de- tailed data report from each ship was prepared which evaluated sensor perform- ance by calculating the mean and standard deviation of all sensors that were within tolerance. Since the sensor data from each buoy on a ship were usually collected within minutes of each other, it was relatively easy to evaluate the data. These detailed data reports from each ship were compared with FGGE buoy checkout status reports which were received daily from each ship. If a problem was found, it was brought to the attention of NDBO and the U.S. FGGE Project Office and appropriate action taken. Similarly, FGGE buoy deployment status reports prepared by each ship following the launch of each buoy were compared with the appropriate data from the AJOUR file and the results analyzed. This analysis and the day-to-day tracking of the sensors were used to determine whether the buoy was operational at launch and whether to put the sensor messages in DRIBU code format on the Global Telecommunications System (GTS) for distribution. ■, The predeployment performance of FGGE buoy barometric pressure and sea surface temperature sensors were expected to meet or exceed the following standards during the predeployment analysis: Physical Standard Parameter Range Resolution Mean Deviation Barometric 900 to 0.15 mb ± 1 mb 0.6 mb Pressure 1050 mb Water -5° to , 0.16°C + rc 0.5°C Temperature +35°C r Redundancy of the received data permitted the detection of most random and burst errors in the overall end-to-end system, which included the satellite processing, communications links, and the ground-truth processing. Differences between the standard and the measured pressures were less than 1 mb, and aver- aged 0.6 mb over the range of pressures tested. The sea surface temperature data were checked while the buoys were on deck prior to deployment. Due to varying amounts of solar and other radiation and different physical locations on deck, the sea surface temperature sensor values varied as much as 2°C. Dif- ferences between sensor values and ground truth were as much as 3°C. 54 o 1— CO _j CO UJ ■>- >- t— Z h- C3 O O O O 1 C- S^ UJ UJ UJ y- \- 1— 1— * on 13 Z z «1 s. UJ O O UJ X oo Q. • • u. o o ►- ».-« O I ^_ Q 1— _J UJ CJ t-^ Q UJ cC UJ o •-D O UJ CO • O r^ o oo q: 3 r^ • O. UJ UJ oo =D o ^-^ cr -r UJ 1— 1 > (— Q CO u_ UJ ci: 1— » z o u_ on o 3: < U- o I I NALYSIS DATA H SHIP — FGGE BUOY CHECK- OUT STATUS REPORT (BCSR) DATA MES- SAGES (GROUMD TRUTH) SATELLITE EQUA- TORIAL CROSSING TIMES OF BUOY MEAR I ME < >- : —1 c 1— < < I O ( 3UJ on J-O ANALYSIS DATA AT/ LAUNCH T t— >- _i UJ n S ^ IJ o >- en a on o u- 1 n eC O Q. Q »- U. i-H ^r Q. UJ 1— I UJ • o 3: >- UJ T »— O Q. o UJ Z3 UJ eC O CO Q£ UJ o 3: o => o <: CO 00 UJ I >- O 1 _J UJ ^-^ a. en zc UJ 1— Q oo^—rD r) ct ci: >- (— oo (— 1 O et Q n> 1— QQ Q QQ OO ---Z ID UJ 1— 1— O CD z q: ai CD UJ o cs u. s: d.--^ ] vt •r" (/I >» «o c ro «d •M to T3 I lO Q. O) 4-> «/) T3 C to 0} L. 3 •o a; s. 3 C7» U- o ct: T^ UJ Cl o CO r> q: 1 u_ ^--^ _i UJ OO - on _i >- ZD UJ 1-^ O O CJ 1 < ZD — ) — . Q CO «t > 55 After the buoys on a ship had been activated for several days, the differences in pressure between the individual buoys and the average of three or more buoys sampled at the same time were determined. If this difference was greater than 1 mb for a particular buoy, the sensor data were declared out of limits. A similar procedure was followed for the temperature, except that a wider tolerance was used and the temperatures were monitored over a longer period of time. The calculated averages of the deviation of the good sensors from the mean of the sensor values for each satellite pass are listed below for each ship: Average Pressure Standard Average Temperature Standard Ship Deviation (mb) Deviation {°C) ORCADAS 0.5 1.3 POLAR STAR 0.3 0.7 ACUSHNET 0.5 ^ 1.1 MAUMEE 0.4 ""-'": 1.5 . ^ .f^f 1^ BLAND 0.4 0.4 There was good pressure correlation. The temperature correlation was poor for the reasons described previously. 4. PERFORMANCE EVALUATION FOR FIRST AND SECOND SPECIAL OBSERVING PERIODS (SOP I AND SOP II) As part of the FGGE SOP I and II, 64 NDBO drifting buoys were deployed in the Southern Oceans. Forty-six were launched by ship and 18 were launched by aircraft. The purpose of the additional air-launched buoys was to reestablish the buoy network in preparation for the second Special Observing Period. After launch, the buoys were monitored daily, using the World Meteorological Organi- zation DRIBU messages and other available data. The DRIBU messages were proc- essed at NDBO on a weekly basis in order to edit the data, calculate perform- ance statistics, and provide a summary report on overall network performance. Buoy positions as shown in Figure 6 were plotted on a weekly basis on a polar chart of the Southern Hemisphere. DISPOSE file listings for the U.S. FGGE buoys were obtained from Service ARGOS to provide additional data for buoy failure analysis. Additional information on buoy sensor performance was ob- tained by comparing buoy data with historical statistical data and synoptic maps of pressure and sea surface temperature obtained from the Bureau of Meteo- rology, Melbourne, Australia, and from the National Weather Service, National Meteorological Center, Suit! and, Maryland. ■ . Weekly and monthly summary reports were prepared for engineering and management evaluation of buoy performance. A typical report is shown in Fig- ure 7 for the period January 5, 1979, to March 5, 1979. The mean, standard deviation, maximum and minimum values for the buoy daily drift velocity, baro- metric pressure, and sea surface temperature are plotted. The last reported buoy position and battery voltage are also reported. Buoy sensor, position fix, and network performance are computed for all the operational buoys during the report period. 56 Figure 6. --U.S. FGGE drifting buoy tracks as of November 5, 1979, 57 Itp- r. mm >- 5£ u. en — r- gsi g m (OU> tou)^ni^niiomofniou>«oiooi ^Aj<8Aia>so> — 9«o — — — ^ jr^~ z..oaoea>tftmtft'— o~«oo»ar«oimo>r>a»aDO|'<>a>9>oioi'OMie ^ • fiw»j>r*r»ftja»vowTftjMoo»o'»»r'^OB — — *'r»o»j' x3>tf»ino»ou)oroin«n'^in'\i(Oft iP(9.r— w«^ ij''fnj^ou)in''i7onjmor>>-Mr> iczu.o>r*'or'ininor-(D — inooxnoxutor-ottooo'O'^'^oaD'o — roa>u>>oioic«oatf~«or*TOvDj'^tpo«v»p»f^'\i»*>'\jU)o^'\i — — or*«\jinif»»^ OT't^r^r^— ioiPiP<7>(7ir-ifloo»or-»flaDOioa3CTicno> — — ODooiODoo — ooio — — (TCD'or'jD oioto>oi(i*aotoioioi(7>otO)o(Tto(na)0)0)oa>o)0)a)ooafoc7>(nooooooto30totoi(T>0) K •u.— (O— OlO — lO'OOro^lOiO'^IO — U> — fOaXD — mrnCDCDr^dO — '^'^(S — '^ — CDtD'^OOtOinflD «/*< ajftjK* — iprutDiD— •/>(Ti'>^Q)r-o» — j'Oi — oi'vjnjfu — — — «vijtD'~-r~— U)*r»inr-j'0» — u»jC rv» — — — — footnocnai'Mo ajoowturo'Ofowwnj — f^ftj — fuoifofu — oi — — \jm«\»oo — Ui 0000000(TlaO)0100000000000C3000000000000000CJOOOOO g o •ifcfflonjwoinonotPotfij-ajoB^ — ^if» — — — 00 — a)«\iQ)o«\iaiu) — t^tucD^M — a o — r-wo»r» — o tt>- 10 (Bcnu) o>nj at oi o(vj ^ OD — Ki ^ X oi — o r* ^ ot O) (s ^f^o^ y «\j*\jin — — wr'iu'^flDW O U,0^00)T — CD— ODOirVilPUXO'^ — n — (DlO'OOtCOOlOr-'OOr^'VjO'^tOOOtni— iPrOCDlOOf) Tf\j(nnjr^(D(\j'^cDioinfVjnj(Oio<\jCDO>'^oir>u>(7>^o(0^'~a>oooooB (0< — oi(7)O>O'-ODaDODO>0)OCDO — — CDCO — — — — — — — — — — — — — — — — — fVlCTl — — — OXDWO) y oai(7io>oioa>(7)(nonoioo)oooo)(pooooooooooooooooooo)c oooioxno) >-H-ru^rooa)o>cr>njoDo — \d r~^rnr\iU>rtinf'ioDa>rootfDK>ir>ininoooir>0)0)CDineDOD'7--o> — cnotj* (OiO(OiO(OU3U)(OiO(OU>u>r^(OU}U)iou>iO(O(OiotoiO'^iXi''iAjnj<\i aj> »- — oooooo — ooo — j-o — ooooiooooooooo o — ooooo — o — o ao — oo — §"= I x«\i(no»io — «DAj'*i— yr-- *vnO(DMin'or-o»r-«flfur-«0'ointwo"^r- — o»r^a»r>oDif o— ODOtio — H-o — r>oa-a)flBcnA»r««\j ur — — — xrui*> — — »o — nj — <\i — — CO — oi rtj Aj K» ftj — — oj — — — — — *u— _ — — Sor^iox T — K>i~ — oiinoofvj a- omcnoo — rom — ooo(DO>x«fl'\joiaio)iDOnO'\j«Oir'Oo — «vi— x »-aifVJiowinM — m— ftjooj-foj-MfuajoixfUforonKtru— rurofo- — f\j tut\j--rr^nii\if^<\j >U) _» • — z «<Mr-in«\ja)cnnjftjoDr»oDaio>u>XJ"ODj-cDK»Mr-f^r"0>o — a-oj'— ly owx^fflirjor-j-^Mx — j-iDirxfliDininioinj-xxj-iowx xiniowT^wMwioft. xiniO:ru)»n 3 — — o>TfOK>j'Oj'tPU)moDa3 — — xr^x- (jioioDOMOOM — o — Tr-ciK>o»oowvDiLr»*xi»oo»flD«u o 5zoiDr»inr-— tnr"iPO>u)r-MOBooinooo»oiT)njp-i>ou>iooa)cPODO>oiiOiniftMooa)iu»oynj«nr» oif)U3ior-(Or-or-M(D — 0>xr~r^r*r-air~fuc .-j-^ioin — lOfu- ifloo o— oiosoiatoiinonj ►- ' yjuir-— CDMOJODfuftjj-otiPt^ oo>a)U)o>onjoji\)r-j-Mmoiru — oxxjr-jotooftODiOAi — ftiwifnoftf flL>-oj-MiDtf>j-Ajr~«oj-iD*oioiDU)CD — Mifioo— »fiU)fUoo>(J»Mx«D — *V) M g» in o £ o O o o o _ UiQTOQDGOOOODOO^CDOtCDu^OOOOTOOl >-vr~r-r-r-r-r-r~r-r-(^r-r-r-r~r»' 0>- — ODODJ— fUf^Ol(\Ji0'Oj-jlDj'__-_ . _._._ _-,«_.. ^_.w _iO — — <\JOO(\j — ooooomo 0000-— OM — o*\j — ooiofonj- o — ooooo — — — CQ<\»AJ — fUAi- — furv.njfv»nj — fXjfvjojMru'Vi ojoj- fufVjfV) — fUfU — fo — (Vi _ — — — u£ — — — — — — o — o — oo — o — — ooo — — o — — — o — — ooooo — oocoooooo — f\iMxinr~a)oo — fUfoj-in»Dr~ooo>o — rorojintor~o — «vj OOOOOOOO — — r^J(^Jr\J(^JfVJnJ(^J«^JfVJ'VJ'OM^orOfO»»1^<^K(^^J■ f •j-j-iniPin <(OU)(DU)U>U)lO(DlOlO(£)l£>(OU}(DlOU}(0(Oa>U)a3tDU)U)(O(aiOlOU)i0(OUDl0U)l0lD(DlOUilOtOtOU>tD 0(/> . — — • — — — — — — __ — — — — S — AiMj-jof^cDCDo- <\jMj-iniOf^flDCT>o — (\jfij-(r)»Dr»a)<7>o — njMjiniDf^cooifi'-ODcno- ftj >-0DOooooooo — roaiajfUCjfuroOjrvjai'^Mr^roMMfiMfOTT ■• r 3- inifiin O — (Dli3(D(OU)U}lOU)lD(OlOlI)lOkOU}U>(OU}lOlOUC30UUOlOlO(OU)lACO(OU>lOtDlOa>inij&iOiatriti3 ffiOinu^intninininuiininiPuitriuiiPioinu^ininininu^uiinintfia^iOioininuiu^uitnintniou^m ^ « f^ • Oi Oi II u u to Z H < O K^ -« z ft o h u >* ft H . ^5 f" in •— < H >* f w O S m ^ 1 N r»v CO • t- "— o> A 1 ^ ir> K CiJ u >» 5 < 3 C «d PERFO CO AST A 1 >» s. <0 K ?-. 3 ^r _ ?-. ^ «a U (£> c o W »-i >* >* ■t-> O O D U o> CQ CQ Q. O • ^ • m 'T r«* o eo irt • OJ CM &. II o> 3 IONS LID 1 o H < z < i> < > !z S K — *»• OBSK DATA OS u °9 ft u; X. CC ^ ^^ i^ 'X. ^ r- , H t=» ' u z fc Z 58 The comparisons of means and standard deviations of pressure and temperature from the drifting buoys with means and standard deviations of pressure and temperature for comparable locations from synoptic charts of pressure and temperature for the same dates confirmed the overall excellent performance of the drifting buoys. As of November 6, 1980, the original 64-buoy network had logged in over a total of 18,000 buoy days of operation since the start of the Global Weather Experiment. Considering all failure modes (except four deployment failures), the average time to failure was 319 days. The average time to failure is defined as the total buoy network operating time divided by the number of failed buoys. The average age of failed buoys was 290 days. The average age of failed buoys is defined as the total of failed buoy operating times divided by the number of failed buoys. The cumulative failure distri- bution is shown in Figure 8. 5. CONCLUSION The feasibility of economically and reliably deploying meteorologi- cal drifting buoys in the remote data-sparse areas of the world has been shown during FGGE. The performance of the U.S. drifting buoys during the experiment indicates that about 80 percent of the buoys can be expected to remain operational for at least one year. These drifting buoy-derived data have also proved extremely valu- able in overcoming weather forecasting problems arising from the deficiency of surface observations in the oceans of the Southern Hemisphere. The exper- iment has provided the operational experience and data network performance characteristics needed for the planning, design, and implementation of future drifting buoy monitoring systems. An operational polar-orbiting satellite network, reliable data proc- essing, and a data dissemination network have been established to provide a global environmental monitoring capability during the next several years. Planning is underway to continue this capability into the next decade. It is expected that drifting buoy technology will play an increas- ingly important role in climate-related research programs during the next decade. Climatically important ocean processes are presently poorly under- stood, due mainly to the relative lack of long-term, synoptic time series data defining large-scale oceanic variability. Comprehensive data sets are needed for ocean climate diagnosis, for model development and validation, for process-oriented studies, and for investigations of ocean-atmospheric coupling. Meteorological drifting buoys with increased measurement capabili- ties, in conjunction with other remote measurement systems, will play a key role in providing the needed data sets. The use of drifting buoy data for improved surface analysis in the Southern Hemisphere has been amply demon- strated during FGGE. Within the next several years, operational drifting buoys may become a major source of meteorological data for forecasting pur- poses along with other remote measurement systems. 59. o "* in o T~ in • o cu 00 S- ■«i- 3 n3 4- o o CM 3 ■^ UJ O CO CD CD Ll. O M- (O O CO C o o •r— CO 4-> CO , . 3 CO J3 >. o CO *^ o T3 S- CO +-> s CO •r~ o '3 -o C\J o cu 3 > o CQ 'r— ■^ -t-> C\J (0 3 o E T~ 3 CVJ O o 1^ CD 00 o in o lo o lo CO CO CM CM t- 1- o o CM lO K- ^.^ E ro CD c O fO U) to T- 1— 1 c o •1— c lO CM u> ^ a XJ < c (TJ \u 1— o > to 1 — 1 o u. c *" O o +-> >- ro < t_5 o to in +-• to x: o c 3 ro o in 5 O) in &- 3 CM cn G3H0Nnvn SNOomva dO uaai/^nN aAiivnni^no 66 340 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 Figure 3. --Balloon locations at the time of final transmission. llf 340 I I I I I I I I I I I M I I I I I I I I I I I I I I I I I I I ^° 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 Figure 4. --Last location--Canton balloons SOP-I. 67 of last locations that occurred in regions of convective storms gives an esti- mate that more than 87 balloons were destroyed by convective storms. Counting the number of last locations that occurred near the magnetic cutdown lines gives an estimate that 58 balloons were terminated by the built-in magnetic cutdown system. Other balloons were lost to electronic failure, balloon failure, and pressure cutdown. The pressure cutdown system terminated the flight if the balloon descended into the commercial aircraft flight region. Since balloons fly at a constant density surface, they could descend below the set pressure cut level if the ambient air temperature was too warm. This would occur at high latitudes. Many balloons were undoubtedly cut down when they flew too far south. The ARGOS system performed extremely well. On the average, data and locations were obtained for each balloon four times per day. This resulted in approximately 50,000 wind and temperature measurements for the complete balloon program. The balloon platforms transmitted 1 watt to the satellite. However, our tests have shown that 0.1 watt would have given location and data for 70 percent of the overpasses. Balloons were located that had gone to the surface and were hanging in trees. Figure 5 shows the plotted data for 18 hours of balloon trajectory for all balloons flying on 16 June. The trajectories result in a presentation similar to a streamline analysis. Even with the limited data it is possible to construct a wind map that gives a picture of the tropical circulation that cannot be derived in any other way. Interesting features seen on the map are a blocking circulation in the Indian Ocean that is defined by a few balloons caught in this circulation, and slowly rotating counterclockwise. Balloons crossing Africa jam up against this circulation and eventually spill out across South Africa and then move rapidly across the South Indian Ocean, turning north- ward along the west coast of Australia. There are three waves in the equatorial Pacific. The most pronounced wave is over the north part of South America. The Tropical Constant Level Balloon System produced a useful data set for the study of the wind circulation in the tropics. In addition, the experi- ment demonstrated that a balloon system is a practical and economically feas- ible means of collecting meteorological data in the tropics. The cost of each balloon flight was approximately $2,500. This resulted in 50,000 wind data points for a cost of approximately $16 per measurement. 58 O O O o o o o ^COMi-O^CMCO I I I I' I I 1 I I I I I I o o o o ^ lO (O K ' I I I I I o -* CO o CM CO o o CO o 00 CM (Tl o 1 — 1 CM C O 3 •O CM WD I — 1 O C CM o CM en O c o •r- CM 4^ o CO to c ^ o o o CO 1 — X) o ^ -* 1 — T- fa o 4- o CM ^■ >> S- o o o 4-> o 1— O) ''~i o ro CO 1— O '. CD LO OJ t- O 3 CD ^ lZ o CM o o •* « 69 .A CHAPTER 6 AIRCRAFT DROPWINDSONDE PROGRAM By 0. Scribner (NOAA) J. Smalley (NCAR) 1. INTRODUCTION The purpose of dropwindsonde operations was to probe the tropical air mass. In the tropics the coupling between the pressure and temperature fields and the wind field is weak. Thus a separate measure of wind is needed there. It was the primary purpose of the dropwindsonde program to make soundings to obtain wind profiles. Pressure, temperature, and humidity data were collected at the same time but were of secondary importance. The flight tracks were pre- planned. They were chosen to cover as much of the tropical ocean areas between 10°N and 10°S as possible, but to avoid large atmospheric disturbances or locales sounded by surface ships. Many operational considerations also influenced the final flight track configuration. These and many other details are covered in Chapter 7 "Dropwindsonde Operations". 2. HISTORICAL BACKGROUND 2.1 The Origins of Omega Windfinding As best we can determine, John Beukers of Beukers Laboratories (1964) was the first person to propose that upper air windfinding might be done by a system which used retransmitted navigation aid signals to sense the winds aloft. Certainly, Beukers was the first to build a system (1967) that actually did the job, under a contract resulting from the strong support and amplification of Beuker's ideas by Chris Harmantas of the Weather Bureau. From 1968 to 1974, Beukers Laboratories designed and built increasingly more sophisticated NAVAID windfinding equipment. All of this early work was done with surface-based rawinsonde systems, first with LORAN-C, then the Omega, and later with Navy VLF Communications stations and the Soviet NAVAID system as signal sources. By the middle 70's, Beukers had even built equipment which could mix signals from different NAVAID systems as a source of positioning data for rawinsondes. Once proven for the case of land-based rawinsonde observations, it was a natural progression to think about using similar techniques for ship-based rawinsondes and dropwindsondes launched from aircraft. The NAVAID concept is particularly suited to these latter applications because, in contrast to the widely used radar or radio-direction finding (RDF) rawinsonde systems, no stable reference platform or mechanically complex precision tracking antenna are needed. Ship- board NAVAID equipment built by Beukers was used in the 1974 GARP Atlantic Tropical Experiment (GATE) and a shipboard system built by VAISALA and TRACOR was used in FGGE. Aircraft dropwindsonde equipment was used in both experi- ments. The principles of windfinding with NAVAID signals have been discussed in the literature in great depth by Acheson (1970, 1978); Beukers (1967, 1972, 1975); Govind (1973); Olson (1977); Passi (1973); Poppe (1971, 1973, 1974, 1979); and others. Suffice to say here that systems designed to work with the worldwide 73 Omega NAVAID system use a sonde (balloon-borne or airdropped parachute supported) which receives the Omega signals (usually the 13.6 kHz transmissions) by means of a miniature Omega receiver in the sonde. The Omega signals then modulate a telemetry transmitter (usually in the 400-406 MHZ band) in the sonde, which then transmits to the receiving/processing equipment. Customarily, in most NAVAID windfinding systems, the sonde also contains pressure, temperature, and humidity sensors which generate additional signals which also modulate the telemetry carrier. The retransmitted Omega signals are subsequently processed by a com- puter which determines the changes in position of the sonde. The changes in position are a measure of wind velocity. 2.2 Evolution of the FGGE ODWS The Omega Dropsonde Windfinding System (ODWS) is an improved version of similar equipment (described in several articles by Rossby, Govind, Cole, Pike, Norris, Saum and Lee, 1973) used successfully during the GARP Atlantic Tropical Experiment (GATE). The GATE system, and the dropwindsonde to work with it, were both developed, tested, and operated (in GATE) by a group of able scientists and engineers at the National Center for Atmospheric Research. It represented an important innovation in meteorological instrumentation - especially for the tropics where wind measurements are vital. For the first time one was able to obtain profiles of wind speed and direction^ as well as temper- ature and humidity profiles, as a function of pressure, from the flight altitude of an aircraft to the sea surface. And the flexibility and range of aircraft operations made it possible to obtain upper air data over ocean areas not readily accessible by other means. But the GATE system required people with a high degree of skill and experience for successful operation. The system needed to be made simpler to operate and to be more compact and reliable, in order to make it suitable for use by trained weather observers and technicians operat- ing from diverse types of aircraft at locations around the world. Prior to FGGE entering the picture, NOAA's Environmental Research Laboratories (ERL) decided to begin work on a next generation of ODW systems which would take advantage of technological progress and the GATE experience. Dr. Stig Rossby (by then relocated to the Research Facilities Center of ERL at Miami), with considerable assistance from the NCAR team, began the develop- ment of a set of specifications for the new ODWS. When the Carrier Balloon System was eliminated from the FGGE plans, the U.S. GARP Committee agreed with the U.S. FGGE Office's recommendation to replace the carrier balloons with an aircraft dropwindsonde program. The FGGE effort was joined with the ERL initiative in the new ODWS program. TRACOR, in Austin, Texas, won the competition for the procurement of ten systems and was awarded the con- tract in June 1976. Dr. Rossby accepted technical responsibility for the joint ERL/FGGE effort and managed the test program using NOAA aircraft. The TRACOR "First Article" was conditionally accepted in early 1978, although the software continued to evolve for another year as further TRACOR equipment 74 and sonde tests and operations during the first FGGE Special Observing Period (SOP-I) uncovered areas of needed software improvement. The equipment proved to be significantly easier to operate and to be more reliable than the GATE version. When the FGGE Project Office opted for the ERL dropwindsonde system as a major U.S. effort for FGGE, the Project Office entered into an agreement with NCAR for a companion effort to improve the sonde over that used in GATE. NCAR's Research Systems Facility undertook the work, with J. Smalley as team leader (described in an article by J. Smalley, 78/79). That development was completed in the early summer of 1977 with the delivery to NOAA from NCAR of a set of drawings and specifications for the improved dropwindsonde. VIZ Manufacturing Company, Philadelphia, was awarded the contract for 7,500 sondes in August 1977. Vernon Zurick of ERL was technical manager of the procurement^ with Smalley and the NCAR team continuing as advisors. As intended, the new sondes were a major improvement, in reliability and performance, over the GATE sondes. 3. DESCRIPTION OF THE ODWS The ODWS consists of permanently or seirii-permanently installed equipment in the aircraft (the On-Board System) and dropwindsondes (sondes), one of which is expended for each sounding. The sondes, after ejection from the aircraft, telemeter signals back to the aircraft. The On-Board system receives the telemetered signals from the sondes; separates the Omega signals from the thermodynamic data; processes the Omega signals into wind speed and direction and the thermodynamic data into meteorological parameters of Pressure (P), Temperature (T) and Humidity (H); displays the processed data; and records semi-processed data on magnetic tape cassettes for more sophisticated post-processing. Following is a brief description of the Omega Dropsonde Windfinding System as used in FGGE. 3.1 Description of Aircraft On-Board System Very briefly, the On-Board System consists of two antennae (one for local Omega and one for telemetry reception). Omega receivers, sonde receivers, cassette recorders, a computer and other data processing modules, a sonde preheat and baseline module, two solid-state teletypewriters, a chart recorder, a paper tape reader and a sonde launch tube. Figure 1 is a schematic block diagram of the equipment. Besides power and crew intercom, the system requires input con- cerning aircraft navigation. There are several selectable modes of this input, three of which were used by various FGGE aircraft: The best mode, when the aircraft is equipped with an Inertial Navigation System (INS), is to obtain aircraft heading, ground speed, and flight level winds from the INS. 75 ^^ — m 0) E a n c c O o > "^ -a -o QJ e > (D o E i~ to •^ • E T3 s- P^ O) 1. to o •,— 1 s- 1 cn >. CO -c: c; i- CD 82 was no exception. After pre-heat, baselining was accomplished as previously outlined in Section 3,1 of this chapter. Pass limits were within: j^ 2 degrees for temperature i ' + 10 percent for relative humidity +_ 10 mb for pressure ■' »'•/ From the baseline data, the computer automatically provided an offset correc- tion from the factory calibration. But the above limits for that correction were established as being those within which the designed accuracies could be retained with only an unsophisticated offset to correct for small changes from the factory calibration. In FGGE, the operator had the option to over- ride the "fail message" from the computer, as these thermodynamic parameters were of secondary importance. 4. SUPPORT ACTIVITY . v- 4.1 Aircraft Installations .: * The FGGE dropwindsonde missions were flown by both NOAA and USAF aircraft. Two NOAA P-3s and one NOAA C-130 v/ere permanently equipped with ODW systems prior to FGGE and will continue to employ them in research pro- jects for the indefinite future. It was a different story for the Air Force planes. Those installations were temporary for the project and would be removed between the two Special Observing Periods, reinstalled, and perma- nently removed after SOP-II. This dictated that the ODWS console, sonde storage racks, pre-heat chamber, work table, operator chairs, and slave printer be placed on a pallet which could be readily installed and removed. The antenna, cabling to aircraft systems, and, in some caseS; the launch tube had to be engineered for ready removal. There were two USAF installation programs: one for WC-135 aircraft and one for C-141s. Both programs were managed by Kenneth Foulke of the U.S. FGGE Project Office (NOAA). Of the two, the WC-135 program was the less difficult. The WC-135s already had launch tubes, and since their usual missions required considerable electronic equipment in the cabin, the aircraft interfaces were straightforward. Three WC-135S were equipped for a total cost of about $70,000 of FGGE funds, since Foulke could arrange for much of the work to be done at the home base of the aircraft by USAF technicians at little direct cost to the program. The C-141 program was far more expensive, being accomplished en- tirely by a USAF contract to Lockheed, under the cognizance of the USAF depot at Warner Robins AFB. The C-141 is a cargo carrier, which meant that connection to aircraft systems was more complex, since there were no access points in the cabin. Special launch tubes were constructed integral with replaceable side escape hatches well aft in the aircraft. These reconfig- ured hatches replaced the standard hatch doors during times the ODWS was in use. Specially designed, readily removed tripod supports for the parts of the launcher extending into the cabin were fabricated. Expensive tests had to be run to ensure that the launcher location would eject sondes properly and that ejected sondes would not hit the aircraft. Altogether, the direct cost of designing, installing, testing, and removal of the ODWS on six C-141 aircraft approached a half million dollars. 83 An interesting feature of the C-141 installations, applicable only to the four aircraft which would operate in the Pacific, was the removal of all national and organization markings from the aircraft, excepting only the identifying "tail numbers". This was a stipulation of the Mexican Government in order that the aircraft could operate to and from Acapulco. In place of the USAF markings and American flag, a WMO insignia was placed on each side of the tail of each of the FGGE Pacific aircraft. 4.2 Training The ODWS operators for the Research Systems Facility of ERL, which operates the NOAA aircraft, were trained on-the-job by Dr. Rossby who, as technical manager and test director for the on-board ODW systems, had become thoroughly familiar with running the equipment. That task was made easier^ because the RFC engineers and technicians were already accustomed to operat- ing complex computer-dependent scientific equipment on the NOAA airplanes. A much more lengthy and formal training program was needed for the weather observers who would operate the ODWS on the Air Force aircraft. Lt. Col. Rich Chappie and Master Sergeant Lee Weiher of the USAF Air Weather Service (AWS) conducted a recruiting and screening program of volunteer AWS weather observers to select the ODWS operators for the par- ticipating USAF aircraft. They selected four senior non-commissioned offi- cers (one of whom was Weiher), and twenty-two airmen to undergo special training for the FGGE ODWS project. Each of the senior NCO's would later be an ODWS operator supervisor at a FGGE operating location, and the other twenty-two became the equipment operators for the WC-135 and C-141 aircraft. Following flight qualification by the Air Force, the trainees were divided into four classes. Each class attended two weeks of "ground school" at the TRACOR factory conducted by TRACOR engineers and software specialists. Half of that time was spent in "hands-on" practice with two ODW systems which had been set up as simulated flight trainers; the other half in classroom work to provide an understanding of sonde preparation, of Omega windfinding theory, of the reasons for the many actions required of an operator to ob- tain a valid sounding, and of the special operator actions needed in FGGE. Each class then went to Miami for a week of basic flight training in NOAA aircraft, under Dr. Rossby' s direction. During that week, each operator flew four practice missions and made a number of soundings using production sondes. Purposely, both phases of the training was arduous, both mentally and physically, in order to ensure competent operators for the project. All but one trainee finished as qualified ODWS operators - a tribute to the AWS selection process. Shortly before deployment of crews and air- craft for the FGGE SOP-I, each operator was afforded one more flight for pro- ficiency and refresher training. Upon deployment, en route to the operating locations, still more sondes were dropped as a crew coordination exercise. 84 The costly and demanding training program, and the large number of expensive sondes expended before a single FGGE observation was made, paid off handsomely. A new system, with operators who until selection were not at all familiar with running systems of this complexity, combined to produce good results from the first day on station. 4 . 3 Logistics Support ' Good equipment properly installed, proficient operators, an adequate supply of sondes, and carefully planned missions served to get things started on the right foot, but more was needed to minimize subsequent shut downs. The best of aircraft and electronics equipment will malfunction from time to time. Provisions for restoration to service had to be made if the FGGE aircraft dropwindsonde program was to succeed. The NOAA and Air Force aircraft units were both accustomed to oper- ating away from home base, and had already-developed maintenance support arrangements. While some missions were lost due to aircraft outages, the only devastating impact on FGGE from that source occurred at Ascension Island during the first SOP. Smalley, in his report on the Atlantic operations, discusses that issue in Part 2, Chapter 7, of this report. Suffice to say here that maintenance problems with the WC-135 aircraft at Ascension resulted in less than half of the scheduled missions being flown over the Atlantic in SOP-I. As a result of that disappointing performance, the WC-135s were not used in SOP-II. Instead, two additional C-141s were outfitted with the ODWS between the two SOPs and were deployed to Ascension for SOP-II, with excellent results. The logistic support for the ODW systems themselves had to be de- veloped specifically for FGGE. Long before deployment, three types of spare parts kits were procured. With the help of the TRACOR engineers, maintenance parts were identified, based on statistical probability of failure, in three categories: Each ODWS had its own set of running spares of small items, kept with the equipment. A second set of more expensive spares, containing parts with a lower probability of failure, was provided for each operating location, to support a number of ODW systems. A third set of vital but unlikely-to-fail parts was procured and kept at one central location along with a complete back-up system. The spare parts are of little value without expert repairmen to use them. The NOAA RFC has a few individuals with the necessary background and experience to enable them to become ODWS repairmen in a relatively short time. Also, since RFC plans to retain the systems indefinitely, it was worthwhile to expend the required time and effort to develop the new skills. Thus, the ODW systems on NOAA aircraft were maintainable within NOAA re- sources. 85 For the Air Force aircraft, NOAA contracted with TRACOR to place a qualified maintenance technician at each operating location during each of the two SOPs. Excellent service was obtained from these individuals. ODWS outages were a serious problem only during the second SOP in the Pacific, where a rash of failures of the NOVA computer core memory boards exhausted the pre- positioned spares and threatened for a while to collapse operations there. It is suspected, though not proven, that lack of air-conditioned storage between SOPs, during the time that the equipment was not in the C-141 aircraft, con- tribiuted to premature failures of the NOVA core memory boards. NOAA was fortunate to locate a source from which reconditioned NOVA core memories could be bought on short notice (lead time quoted by the manufacturer and its dis- tributors for new boards or repair of old ones ruled out those sources). Emergency purchase of the reconditioned boards and employment of extraordinary means for their rapid shipment rescued the Pacific systems, but not before several missions were lost. 5. SOME COMMENTS AND CONCLUSIONS ■ '' ■ Despite the computer problems in the Pacific during the second SOP, and the C-135 maintenance problems during the first SOP, the FGGE Dropwindsonde Program, overall, was a success. It made a critically important contribution to the FGGE data set; for without it, large areas of the tropical oceans would have been void of the vertical wind profiles considered essential by the scien- tists who planned the experiment. The success was partly due to careful plan- ning of flight operations, installation, logistic support, and operator train- ing, and of course to the timely development and procurement of the ODWS on- board system and sondes. But the critical factor was competent and dedicated people in NOAA, NCAR, the Air Force, and perhaps most importantly, in the three business firms which produced the system, built the sondes, and converted the C-141 cargo carrier into a scientific research platform. >., '-Vv »r(.;' ■•. 86 CHAPTER 7 DROPWINDSONDE OPERATIONS PART I: THE PACIFIC By Edward Tiernan (U.S. FGGE Project Office) 1 . INTRODUCTION Prior to the Global Weather Experiment, San Cristobal, in the Galapagos Islands, was the only operational upper air station in the 8 million miles of open ocean area between the west coast of South America and the International Date Line and between the tenth degrees of latitude north and south. Its data were occasionally supplemented by intermittent upper air observations made by the University of Hawaii on Fanning Island at 160 degrees west. A total data void existed between these two island stations. Starting in December 1978, two upper air observations a day from tiny Penryhn Island became available as a result of a special U.S. effort for the Global Weather Experiment. A month later, two observations per day from Canton Island were added, and the Fanning Island schedule was also increased to tv70 observations daily. Early in January, ships equipped with special upper air systems arrived in the area and commenced upper air observations at local noon and midnight. By mid-January huge Air Force C-141 aircraft were flying daily sorties over the area, dropping windfinding sondes every 350 kilometers along specially designed tracks. A major effort was underway. The winds aloft over the tropical Eastern Pacific were being measured as they never had been before. Similar efforts were underway in the Atlantic and the Indian Oceans. It was the beginning of the first Special Observing Period (SOP) for the Global Weather Experiment. It is not likely that such a massive effort will be mounted again soon. However, the experience gained in the planning and execution of such a large- scale experiment could be useful to others in the future. This chapter presents a brief discussion of the factors taken into consideration in the planning and implementation of the operational aspects of the Pacific Dropwindsonde Program as well as the major problems encountered in its execution. 2. OPERATING LOCATIONS, FLIGHT TRACKS AND AIRCRAFT UTILIZATION Four Air Force C-141 aircraft from the Military Airlift Command were available for the Pacific area. To obtain as many spatially independent drop- windsonde observations as possible, the strategy for employment of the aircraft was quite simply to fly them as high and as far as possible in the zone between 10 degrees north and 10 degrees south latitude. To accomplish this, two suit- able operating locations in, or as close as possible to, the zone had to be found. When all the requirements for the operating locations such as location, runway length, maintenance facil ities, and logistic support capabilities were considered, it was determined that Hickam AFB, Hawaii, and Acapulco International Airport, Mexico, would be the best choices. A third possibility was Howard AFB, Panama. However, C-141 s operating from Howard AFB could not reach Hawaii nor could nonstop round-robin flights from either Hickam or Howard (round-robins) extend far enough eastward or westward to provide the required coverage of the central portion of the Eastern Pacific. Plans were therefore developed for the use of Hickam AFB and Acapulco International, with Howard AFB to be used in the event of problems in obtaining operating rights at Acapulco. 89 Figure 1 in Part 4 of this Chapter illustrates the primary tracks planned from Hickam and Acapulco (PIO, P20, and P30). If possible, each of the three tracks would be daily, with one of the four aircraft held in reserve. The actual tracks flown on any given day were selected based upon the location of the Intertropical Convergence Zone (or any organized deep convection) and the location of upper air observing ships on that day. A fundamental principle of the drop strategy was to avoid deep convection where the descending sonde would be subject to small-scale air currents in the convective cells rather than the large-scale flow desired. Drops in the vicinity of upper air observing ships were to be avoided to maximize the unique spatial contribution of each observa- tion. The initial deployment plan was to place two aircraft at each operat- ing location. The four aircraft would then be rotated through the system of tracks. However, because of limited parking space for aircraft at Acapulco, it was not feasible to have three on the ground there on a routine basis. Hence, the reserve aircraft would have to be kept at Hickam AFB. The first shuttle (P-20) was to be flown westward from Acapulco and the second eastward from Hickam; this rotation would continue throughout the operation. On alternate days there would be one and two aircraft at Acapulco with three and two at Hickam. Con- sidering the relative capabilities for maintenance and logistic support at Hickam AFB and Acapulco, the reserve aircraft should have been stationed at Acapulco, which had the lesser support capability. However, with an in-bound shuttle to Acapulco ewery other day, which could carry spare parts for either the aircraft or observing system, it was felt that Acapulco could be adequately supported. 2.1 The Move to Panama Plans for the eastern Pacific coverage were essentially complete by mid-June 1978. Clearance from the Government of Mexico (GOM) for the use of Acapulco was all that was still required. It was not until November 20, 1978^ that the U.S. was informed that the GOM wanted the operations from Acapulco to be conducted under a formal agreement between the WMO and the GOM. Even at this late date, we remained optimistic that the first SOP would start on time; we seriously underestimated the task. Intensive efforts were made to conclude the agreement on time. However, the U.S. Embassy in Mexico City was notified by the GOM on January 19, 1979 (four days after the scheduled beginning of SOP-1) that "Mexico regrets that it will be unable to participate in the first SOP". The contingency plan for operation from Howard AFB in the Panama Canal Zone had to be implemented. 2.2 SOP-1 While efforts were going on in Mexico City to conclude an agreement for the use of Acapulco, personnel and aircraft for the eastern Pacific program assembled at the predepToyment staging base at Norton AFB, California^, on January 11th and 12th. On January 13th, when the four aircraft were scheduled to deploy, and with the situation concerning Acapulco still unresolved, only the two aircraft bound for Hickam AFB departed on schedule. The other two C-141s, flight crews, ground support personnel, and observing system operators remained at Norton AFB awaiting clearance to proceed to Acapulco. 90 The first scheduled sortie for SOP-1 was to be the Acapulco to Hickam shuttle on January 15. However, with no clearance from Mexico, the planned shuttle could not be flown. In its place two round-robin sorties, a P-30 to the west and its mirror image to the east (called a P-30 FLOP), were flown from Hawaii. On January 17, 18, and 20, round-robins were flown from Norton AFB south- ward to 8°N latitude. On January 21^ with the situation in Mexico settled at least for SOP-1, one aircraft was deployed to Panama to fly one round-robin sortie per day from there and a third was sent to Hawaii to provide a reserve aircraft there. • ' ' . The first round-robin sortie from Panama westward to 113°W longitude was flown January 23. The two round-robins were flown from Hawaii^ providing coverage from the date-line eastward to 130°W longitude. This was the mode of operation for the remainder of SOP-1. The Hickam-Acapulco shuttle had to be cancelled for the duration of SOP-1 and the area between 113°W and 130°W could not be covered. Figures 2 and 3 in Part 4 of this chapter illustrate the tracks actually flown in the Pacific during SOP-1. The final day of SOP-1 operations for the entire dropwindsonde program was originally scheduled to be February 13, 1979. However, because of early delays encountered at all locations, the program was extended for an additional week. The last day of operations in the Pacific was February 20; 71 round-robin sorties were flown from Hickam AFB and 26 from Howard AFB. A total of 1739 drop- windsondes were launched. Thanks to the flexibility demonstrated by the USAF in accomplishing the sudden change of operating locations at the beginning of the SOP and, above all else, the skill, professionalism, and dedication of the air- crews, system operators, and ground support personnel, the first SOP was opera- tionally a success. 2.3 SOP-2 The agreement between the WMO and the GOM for use of Acapulco Inter- national Airport was finally signed on April 24, 1979, and diplomatic notes between the GOM and the U.S. were exchanged the next day. For the first time all barriers to the full implementation of the original plan for the Dropwindsonde Program in the Pacific were removed. The shuttle flights between Hawaii and Mexico could be flown and unbroken longitudinal coverage of the Eastern Pacific could be provided. (The text of the agreement and the Protocol of Execution are reproduced at the end of this section of Chapter 6.) The first of two C-141s landed at Acapulco in the early afternoon of May 8. To all those who had suffered through the frustrations and uncertainties of the first SOP, it was an emotional moment. In accordance with the agreement with the GOM, all national emblems had been removed from the aircraft; only a small WMO emblem and an aircraft number were displayed on the tail. The second aircraft arrived one hour later. Two similarly marked aircraft arrived at Hickam AFB on the same day. The second SOP started well when, according to schedule on May 10, the first shuttle flight to Hawaii from Acapulco was launched along with round-robins from both locations. The second day. May 11, went equally well when the shuttle 91 was launched from Hickam AFB to return to Acapulco and two more round-robins were flown according to plan. The first two days were 100 percent; only 28 more to go for a perfect record. It was a week later before we had another 100-percent day.^ and overall we would only experience 15 days when the Pacific dropwindsonde pro- gram was 100 percent according to plan. The vulnerable point in the plan was Acapulco, where ewery other day there was only one aircraft available for launch, and where repair facilities and logistic support were minimal for C-141s. It was, however, a known vulnerability and a calculated risk necessitated by the resources available to the program. Of the three tracks to be flown daily in the Pacific, the P-10 round-robins from Acapulco were the most vulnerable. When aircraft or observing system problems prevented a take-off on those days when only one aircraft was available, the sortie v/ould be missed for the day. Further, even on those days when two air- craft were available, if one failed, priority was given to launching the shuttle to Hawaii, and the P-10 track would be missed. The shuttle was not only impor- tant for its scientific value; it was essential for logistic support to Acapulco. During the 30 days of operations for SOP-2, the P-10 or P-11 tracks were missed 14 times. The Hickam-Acapulco shuttle tracks were m'issed on only one day (May 22) and the westward round- robin from Hav/aii was flown eyery day. On the one day the shuttle was missed, a special westward round-robin {P-15) was flown, along with an eastward round-robin from Hawaii (P-30 FLOP). This combination of round- robins longitudinally covered all but 5 degrees of the shuttle track. Hence, except for those 5 degrees on the 22nd of May, the Pacific program provided longitudinal coverage in the active region from 100°W to the date-line e^ery day. A total of 81 of a planned 90 mission sorties were flown in the Eastern Pacific during SOP-2. While 14 planned sorties along the P-10 series of tracks were lost, five unplanned sorties were gained by utilizing the fourth or spare aircraft. 90% of the planned number of sorties were flown and 1518 dropwindsondes were launched (83% of those planned). . ^ ■ < At Acapulco, any concerns the U.S. participants may have had about the degree of Mexican support and cooperation, in view of the long and difficult negotiations, were yery quickly put to rest. The Mexican representatives showed a keen interest in the program and lent encouragement and assistance well beyond that required by the Protocol of Execution. 9S APPENDIX 1 TRANSLATION AGREEMENT ON THE GLOBAL WEATHER EXPERIMENT BETWEEN THE WORLD METEOROLOGICAL ORGANIZATION AND ;; THE GOVERNMENT OF MEXICO CONSIDERING that the World Meteorological Organization is planning a Scientific Experiment relating to the meteorological and oceanographic processes on the global scale within the framework of the Global Atmospheric Research Pro- gramme (GARP) of the Organization and the International Council of Scientific Unions; , ■.,-.;■ ■ > , CONSIDERING that Mexico is situated in proximity to areas in which special aircraft operations will be conducted as part of the Experiment and has at its disposal the appropriate installations, and that it has therefore been recommended that an aircraft operational centre of the Experiment should be established at Acapulco; CONSIDERING that the Government of Mexico has considered with interest this recommendation; NOW THEREFORE the following agreement is concluded as the basis for the co-operation between the World Meteorological Organization and the Govern- ment of Mexico. Section 1 - Name of Experiment The Experiment shall be known as the Global Weather Experiment, herein- after referred to as "the Experiment". Section 2 - Purpose of the Experiment , ; - To obtain a better understanding of atmospheric motion for the develop- ment of more realistic models for weather prediction. - To assess the ultimate limit of predictability of weather systems. - To design a composite meteorological observing system for routine weather prediction of the larger-scale features of the general circu- lation of the atmosphere. - To investigate, within the limits of the period of observation as provided for in this Agreement, the physical mechanisms underlying 93 the fluctuations of climate in the time range of a few weeks to a years and to develop and test appropriate climatic models. few - In order to meet the above objectives, an instrument package will be dropped via parachute from an aircraft, and the instruments will measure winds and temperature and humidity profiles over tropical sea areas of the Pacific Ocean. Such data will be made available to all Member States of WMO as indicated in the WMO FGGE Publication Series: FGGE Report No. 3 - the FGGE Data Management Plan. Section 3 - Conduct of the Experiment * The Experiment shall be conducted by National Co-operating Agencies desig- nated by the Member States of the World Meteorological Organization indicated in the WMO FGGE Publication Series, Implementation and Operational Plan for the FGGE Special Observing Systems: Part B: Aircraft Dropwindsonde System, in co-operation with the World Meteorological Organization, hereinafter referred to as "the Organi- zation". Member States of the Organization, other than Mexico, participating in the Experiment are hereinafter referred to as "other participating Member States". Within the framework of the present Agreement the Organization shall be responsible for the implementation of the Experiment. The Organization will neither accept nor endorse any financial liability that might result directly or indirectly from the implementation of the Experiment. This liability will be the responsibility of the participating Member States. Section 4 - Duration of the Experiment in respect of the aircraft operations Aircraft operations will be conducted during the Period of intensive Observations which is scheduled to take place from 10 May 1979 to 8 June 1979. Section 5 - Co-operating Agencies The designated Co-operating Agencies under the present Agreement shall be: (a) (b) For the World Meteorological Organization: For Mexico: (c) For the other participating Member States: The Secretariat of the Organization Foreign Affairs Secretariat; National Defense Secretariat; Finance and Credit Secretariat; Agl^i culture and Water Resources Secretariat; The Secretariat for Communications and Transport and for Airports and Auxiliary Services. Such national agencies as the Member States shall designate in accordance with Section 16 below. 94 Section 6 - Privileges and Immunities The Government of Mexico grants to the personnel participating in the Experiment the privileges and immunities as set forth in Article VI of the Con- vention on Privileges and Immunities of the United Nations with the reservations made by the Government of Mexico as ratified by it in the Official Journal of the Federation of 10 May 1963. Section 7 - Authorization for Access to and Use of Facilities in Mexico The Government of Mexico authorizes for the Period of Intensive Obser- vations of the Experiment, with additional time prior to and after this period for the appropriate preparation and termination procedures, the use, in so far as practicable, by other participating Member States of the facilities of Aca- pulco Airport as may be required during the duration of the Experiment, in accord- ance with Section 4, and as they appear in the Protocol of Execution attached to the present Agreement. Section 8 - Entry and Departure of Aircraft and Personnel (a) The Government of Mexico shall, upon request which should be made 72 , hours in advance, take the necessary steps to grant in due time the authorization for entry into and departure from Mexico during the Experiment in respect of the aircraft with the emblem of the Organi- zation and personnel of the other participating Member States assigned to the Experiment. (b) The Government of Mexico reserves the right to verify the identity of personnel assigned to the Experiment and to inspect equipment, materials and instruments which will be used during the Experiment. Section 9 - Importation and Exportation of Materials, Equipment, Supplies. Goods and other Property The Government of Mexico shall, upon request, and in accordance with the Convention on Privileges and Immunities of the United Nations, take the necessary steps to authorize the admission without restriction into Mexico for use during the Experiment and in due course, where appropriate, the removal from Mexico, of materials, equipment, supplies, goods and other property of any other participat- ing Member State. Section 10 - Fiscal Exemptions Materials, equipment, supplies, goods and other property, including motor vehicles, belonging to the other participating Member States, assigned to Mexico for the purpose of the Experiment, and imported into Mexico for use during the Experiment, shall, on request and in accordance with the Convention on Privi- leges and Immunities of the United Nations, be admitted free of tax, customs and import duties and other charges, subject to exportation after the conclusion of the Experiment. Detailed lists of such property shall be sent to the Co-operating Agencies of Mexico designated in Section 5. 95 Section 11 - Landing Fees and Other Similar Charges No fees shall be payable by participating Member States for aeronautical activities in Mexico for the purpose of the Experiment. However, the cost of ser- vices rendered in respect of the use of equipment and special facilities shall be reimbursed in accordance with customary rates. Section 12 - Expenditures and Payments All expenditures and payments resulting from the execution of the present Agreement and relating to the provision of services to the participating Member States or their designated Co-operating Agencies shall be entirely borne by those Member States. Section 13 - Liability r; '■ •' 'lUiivv ^ r (a) Each Co-operating Agency of a participating Member State shall be responsible for claims for damage to property or injury to persons with respect only to activities directly related to the Experiment or per- formed by the Co-operating Agency or its employees. (b) Whenever an employee of a Co-operating Agency is involved in a per- sonal capacity in any litigation, the Co-operating Agency shall collaborate with Mexican authorities to facilitate settlement of the litigation. (c) This Agreement will not come into force for any Co-operating Agency of any other participating Member State until it has signed an Agree- ment on liability between the Government of the said other partici- pating Member State and the Government of Mexico. Section 14 - Settlement of Disputes (a) Any dispute between the Government of Mexico and the Organization relating to the application or interpretation of the present Agree- ment shall be settled by negotiation or by any other mode of peaceful settlement of disputes agreed on by the parties. (b) For any dispute of a similar nature arising between anot+ier partici- pating Member State and the Organization or between Mexico and any other participating Member State or between participating Member States the procedure detailed in (a) above shall be adopted mutatis mutandis unless otherwise provided for in a specific arrangement agreed upon between the parties concerned or in a note by which a Member State agrees to be a participating Member State as provided for in Section 16 (a) below. Section 15 - Protocol of Execution The Organization shall negotiate with the Mexican Government for signa- ture a Protocol of Execution which, in accordance with this Agreement, shall re- late to the details of implementation of the present Agreement applicable to each participating Member State, and shall constitute an annex thereto. 96 Each Member State of the Organization shall receive a copy of this Protocol of Execution. Section 16 - Application of this Agreement to Participating Member States (a) In order that this Agreement and the Protocol of Execution may become applicable to any of the other participating Member States of the Organization, that Member State shall deliver to the Government of Mexico a note wherein the Member State agrees to be a participating Member under the terms and conditions prescribed in the Agreement and in the Protocol of Execution and specifying the name and address of its national agency which will act as its Co-operating Agency for the purposes of the Agreement, as soon as it has fulfilled the stipulations of Section 13 (c). The Organization shall receive a copy of the note. (b) Any of the other participating Member States may, if necessary, estab- lish with the Government of Mexico, Supplementary Arrangements, which, in accordance with the present Agreement, shall specify any further administrative and technical details of the required co-operation be- tween the two Governments. The Organization shall receive a copy of such Supplementary Arrange- ments. (c) Such Supplementary Arrangements shall constitute annexes to this Agree- ment, applicable only to the parties concerned. (d) Any Supplementary Arrangements may be amended at any time, by mutual agreement between the two parties concerned. Any amendments shall be notified to the Organization. (e) Any specific arrangement made in accordance with the provisions of the present Agreement shall constitute an annex to this Agreement, appli- cable only to the parties to the arrangement. Section 17 - Notification of Annexes and Amendments The Organization shall notify all participating Member States of all annexes and amendments established in accordance with the provisions of Sections 15 and 16. Section 18 - Duration of Agreement (a) This Agreement shall enter into force upon signature by both parties and shall remain in force until the Government of Mexico and the Organ- ization mutually determine that the Experiment has been completed, but in all events the Agreement shall terminate not later than 30 June 1979. (b) This Agreement shall enter into force for other participating Member States on the date of notification of their acceptance thereof in 97 accordance with Section 16 (a) above, after they have fulfilled the terms and conditions indicated therein, and will terminate in accord- ance with Section 18 (a). Done and signed at Geneva on this twenty-fifth day of April nineteen hundred and seventy-nine For the Government of Mexico For the World Meteorological Organization (Signature) Permanent Representative of Mexico with the international organizations in Geneva Roberto Martinez Le Clainche Ambassador (Signature) Secretary-General D. A. Davies Certified that the above text in the English language is an authentic translation of the original text in the Spanish language. Geneva, 25 April 1979 (Signature) (L. Colson) Chief, Language Branch, WMO Secretariat (Signature) (D.A. Davies) Secretary-General , WMO 98 APPENDIX 2 TRANSLATION PROTOCOL OF EXECUTION Pursuant to the provisions of Section 15 of the Agreement on the Global Weather Experiment between the World Meteorological Organization and the Govern- ment of Mexico; The Government of Mexico, and the World Meteorological Organization, hereinafter referred to as the "Organization"; Have agreed as follows: Article 1 - Conduct of the Experiment (a) Each participating Member State shall detach in Acapulco Airport during the Experiment a representative to co-ordinate activities on the spot and to establish liaison with the Mexican Co-operating Agency, which, for the purpose of the execution of the present Protocol relevant to the Government of Mexico, shall be duly designated. (b) The Organization shall co-ordinate its activities through the FGGE Operations Centre at the WMO Secretariat, Geneva. The main duties of the FGGE Operations Centre shall be to ensure that the planning and conduct of the Experiment are directed at all times toward the achieve- ' ment of the scientific goals of the Experiment. (c) The Government of Mexico shall likewise designate a qualified person as Liaison Officer who will be the contact for the representatives and the Organization. (d) The main duties of the representative shall be: i) to ensure the provision of the operational, administrative and logistic support needed to achieve the scientific objectives of the Experiment; ii) to provide the scientific guidance required for the flight oper- ations from the operations site; iii) to be the focal point for liaison with the Mexican Co-operating Agency concerning all personnel, operational, administrative and logistic aspects of the programme; iv) to co-ordinate the scientific aspects of the programme with the Mexican Co-operating Agency and the participation of authorized Mexican personnel in the programme of in-flight operations. (e) The schedule of operations at Acapulco Airport shall be as follows: 99 Personnel Arrival not earlier than 1 May 1979 Departure not later than 15 June 1979 Aircraft Arrival not earlier than 5 May 1979 Planned departure around 10 June 1979, with possible extension up to 15 June 1979 should aircraft maintenance and/or FGGE observational programme so require. v^.^; i.;n ;- (f) All participating aircraft shall undergo customs and health inspection upon landing in Acapulco. (g) All the operations of participating aircraft shall be subject to the legal provisions for aviation. (h) The representatives of other participating Member States shall provide the Mexican representative with the operations schedule before each flight from Acapulco as well as a report of the programme accomplished immediately after each flight terminating in Acapulco. (i) In the aircraft being used for the Experiment, no arms, photographic or remote sensing equipment are permitted, whether installed in, or carried aboard these aircraft. (j) Aircraft used in the Experiment are required to display the emblem of the Organization. No other emblem shall be used. The use of the emblem of the Organization, of its name and of abbreviations of that name through the use of its initial letters by any Member State participating in the Experiment is formally authorized for the purposes of this Exper- iment by the Secretary-General of the Organization. The said aircraft shall not be subject to the registration under laws or regulations of Mexico. Article 2 - Personnel matters (a) Personnel from other participating Member States shall obtain the appro- priate visas prior to the commencement of operations, and shall be sub- ject to the Mexican immigration regulations. (b) The Government of Mexico reserves the right to determine the number of the personnel of the other Member States participating in the Experiment. (c) Lists of participating personnel shall be exchanged between the repre- sentatives as well as information on any amendments thereto. (d) Appropriate provisions shall be made for authorized Mexican personnel to participate in the flight programme. Up to three authorized Mexican persons shall participate in each flight. Matters relating to the Mexi- can personnel participating in the flight operations shall be co-ordinated between the representatives. 100 (e) Personnel participating in the operations of the Experiment shall not wear military uniform. Article 3 - Hospital and Medical Services The Mexican Co-operating Agency shall: (a) Provide information on the hospital services and premises, as needed. (b) Provide a list of the names, addresses and telephone numbers of recom- mended doctors and dentists in private practice in Acapulco. Article 4 - Specific Undertakings on the Part of the Mexican Co-operating Agency (a) The Mexican Co-operating Agency shall assist the other participating Member States with regard to the following: i) the necessary arrangements for using offices and hangar space at Acapulco Airport; ii) authorization for access to and use of facilities at Acapulco Airport; iii) entry and departure of aircraft and personnel to and from Mexico; iv) importation and exportation of materials, equipment, supplies, goods and other property needed for the Experiment; v) fiscal exemptions in accordance with Sections 9, 10 and 11 of the Agreement on the Global Weather Experiment between the World Meteorological Organization and the Government of Mexico. (b) The Mexican Co-operating Agency shall, in so far as practicable, arrange for the provision of: 1) free parking space at Acapulco Airport for a fleet of up to three C-141 aircraft to be used in the Experiment; ii) such amounts of open storage space at Acapulco Airport as may be required for the storage of equipment and supplies intended for use in the Experiment. This space shall not exceed 200 square meters. Article 5 - Specific Undertakings on the Part of the Organization (a) The Organization shall arrange that the other participating Member States, either jointly or individually and for the duration of the operations ex- cept as otherwise provided above, or as may be agreed at some future date, shall provide, or arrange for the provision of the aircraft, all the tech- nical equipment and supplies and all the personnel required for the con- duct of the Experiment. 101 (b) The Organization shall maintain at its Secretariat in Geneva, Switzerland, an FGGE Operations Centre for the international co-ordination of the Ex- periment. Article 6 - Term The present Protocol of Execution shall enter into force upon signature by both parties and shall be coterminous with the Agreement on the Global Weather Experiment between the World Meteorological Organization and the Government of Mexico. The present Protocol of Execution shall enter into force for other par- ticipating Member States on the date of notification of their acceptance thereof, in accordance with Section 16 (a) of the Agreement on the Global Weather Experi- ment and, upon fulfilment of the terms and conditions laid down therein, shall terminate on 30 June 1979. Done and signed at Geneva on this twenty-fifth day of April nineteen hundred and seventy-nine For the Government of Mexico For the World Meteorological Organization Permanent Representative of Mexico with the international Organizations in Geneva Secretary-General D. A. Davies Roberto Martinez Le Clainche Ambassador Certified that the above text in the English language is an authentic translation of the original text in the Spanish language. < r (Signature) (Signature) (L. Colson) Chief, Language Branch, WMO Secretariat (D. A. Davies) Secretary-General, WMO 102 CHAPTER 7 DROPWINDSONDE OPERATIONS PART 2: THE ATLANTIC By J. Small ey (NCAR) 1. INTRODUCTION The FGGE aircraft dropwindsonde program included flights from Ascension Island in the South Atlantic Ocean. A track was flown, generally east along 7.5 degrees south latitude and west along 2.5 degrees south latitude (Figure 4 in Part 4 of this chapter). Ascension Island is a British possession used exten- sively for radio and undersea cable communications. The United States also maintains a facility there, mainly for satellite tracking. The writer was desig- nated the FGGE Director of Operating Location 4 (OL-4), i.e.. Ascension Island. This chapter is a report of FGGE Atlantic Ocean dropwindsonde operations from his viewpoint. 2. SUMMARY OF OPERATIONS During SOP-I, a single WC-135 was on station at OL-4 for dropwindsonde operations. A variety of maintenance problems prevented the planned second air- craft from ever joining the operation and higher priority national requirements precluded the reassignment of another WC-135 to Ascension Island for FGGE SOP-I operations. The one aircraft which operated at OL-4 during SOP-I arrived on station one week late because of maintenance difficulties which, unfortunately, continued to plague the operation throughout the SOP. The aircraft only flew 11 times during the 29 days it was on station. It turned out that the supply pipeline was too long for timely repairs. Certain one-of-a-kind failures occurred simply due to the age of the C-135 fleet. The failure causing the greatest number of lost days was a cracked cockpit side window. For these and other reasons, the U.S. FGGE Project Office decided to try to replace the single SOP-I C-135 with two C-141s for SOP-II operations. Fortunately, the Air Force was able to agree to this request. During SOP-II, while there were some maintenance problems with the C-141s, only two scheduled flights were missed and one of these was made up later. Overall, the SOP-II operations were excellent, and Atlantic area drop- windsonde operations from OL-4 achieved the highest SOP-II mission accomplishment percentage of the three areas covered by FGGE dropwindsondes. The following table summarizes OL-4 operations for the two SOPs and clearly shows the dramatic improvement of SOP-II compared to SOP-I. (NOTE: Part 4 of this chapter gives a more detailed summary including depiction of actual tracks flown. ) Drop Summary of SOP-I and SOP-II Number of Number of SOP Missions Flown Sondes Dropped I 11 205 II 29 546 - 105 3. DROP STRATEGY Drop strategy differed somewhat from the other OL's. This was brought about by the fact that Ascension Island was directly on the mission track. As soon as the aircraft took off, it was time to drop the first sonde. For best coverage all the way around the track, it was desirable to have 350 km between the last drop and first or, lacking that, not to exceed 500 km. All of this could have been accomplished by a spiral ascent at the start followed by the first sonde drop. Then as Ascension was approached, the last drop could be followed by a loiter (without turns) until impact and then a spiral descent to the Island. However, this plan was precluded by permissable fuel load limita- tions and crew fatigue considerations. (The runway length and slope, and emer- gency braking requirements limited take-off weight and fuel load.) Accordingly, take-off was immediately followed by climb to altitude in a straight line. The first sonde was dropped on passing through 25,000 feet. Succeeding sondes followed at a nominal spacing of 350 km. In practice the last drop was made when the aircraft was quite near the island. To minimize the flight time, the drop was followed by a gradual deceleration and slow descent. To preclude loss of a signal, descent rate was limited so that the aircraft was still above 25,000 feet when the sonde reached the surface. This usually meant that the island was overflown. ^ In the drop sequence, if a turn was imminent the drop was delayed until the turn was complete. If a turn was required while a sonde was in the air -- the usual case — the turn was held to a 10-degree bank. Past experience with GATE data reduction showed that maneuvers at either the beginning or end of a drop are more difficult to smooth accurately so an attempt was made to accom- plish all maneuvers (turns, altitude, changes, speed changes) in middrop. 4. ON-THE-GROUND PROCEDURE The desire to space the first and last drops no more than 500 km apart caused a change in preflight procedures at Ascension. It was most important that the first drop be made as soon as possible. This meant that the first sonde had to be baselined while the aircraft was still on the runway. There was a difficulty imposed by this requirement. Late in aircraft preflight, there came a point where power was switched from external power to internal power. This switchover generally disrupted the onboard systems and meant re- starting the computer and Omega synchronization. Loss of the computer meant loss of any preflight baselining. Thus, it was necessary to establish a pro- cedure that, even though the aircraft was ready, it should not leave the chocks until the ODWS was ready and the first sonde baselined. This sometimes pro- duced anxiety in the minds of the flight crews, because in routine Air Force operations, aircraft readiness is all important. When the aircraft is ready-- it goes. Any upsets or failures compounded the problem. Nevertheless, in the research environment, the scientific requirements of the experiment were made to take precedence. 106 5. PERFORMANCE OF THE DROPWINDSONDE AT OL-4 During SOP-I, almost every OL-4 sonde was checked for proper removal of the timer and chute cover because of start-up problems with the mechanical assembly. The manufacturer was promptly notified of the items being discovered and remedial action was taken. One of the most prevalent difficulties resulted from a rubber packing band binding the drogue. Although the drogue could not open, it would still stream behind the sonde and the timer would \jery likely be started. If the chute cover was at all tight, the undeployed drogue would not be able to pull it off and the sonde would be a "fast fall". If the cover was not tight, it would most likely be shaken out and the chute deployed^ si nee it was packed in such a way that yery little force was needed to pull it from the chute cavity. Even though the undrogued sonde would be falling rapidly, experience during development test drops showed that the main chute would probably survive opening impact. The other most prevalent occurrence was an excessively tightly fitting chute cover which would not release, resulting in a fast fall even though the drogue did work properly. Even after the above problems were corrected, there continued to be a disappointing number of fast falls during the early stage of SOP-I. When it was clear that the problem was not sonde-assembly or launch-operation related, launching sondes nose first was tried with a considerable improvement in launch successes. The C-135 has an extraordinarily long launch tube extending down from the cabin floor. It appears that during the long fall in the tube, the cap was coming off before the sonde left the tube so that the drogue was already partially deployed upon leaving the tube and subsequently not opening properly. Most sondes launched by Ascension aircraft passed the temperature and humidity calibration checks prior to launch, but many failed the pressure check. Surprisingly, most sondes reported a pressure lower than the reference pressure (cabin pressure). While cabin pressure varied with altitude, it was very steady during level flight. Further work needs to be done in this area to pinpoint the cause and the remedy for the apparent anomolous pressure readings. However, in accordance with the FGGE instructions, failure to pass the pressure check was not usually considered cause for rejecting the sonde, since the thermodynamic param- eters were far less important than the wind data. ■'■,'•»'.■ During a flight with a laboratory pressure standard on board, it was discovered that the reference pressure output does not change as rapidly as cabin pressure. It is probably a case of a large software filtering time constant. Consequently, the word was passed to all OLs that baselining should not take place during change in altitude. A printout of reference pressure was available so the operator could watch it until the value stabilized. Approximately half of the sonde drops were coded in TEMP DROP code and transmitted by voice to the WWW Global Telecommunications System. A drop that had few "reasonable" winds was not transmitted and the next drop used. If a drop yielded only a few unreasonable winds, it was coded, omitting the levels where the winds were unreasonable. The judgment of reasonableness was, of course, sub- jective. After a couple of days or more of flying, one could get a feeling for the expected values. 107 Although one of the levels to be coded was 1000 mb, the real-time system cannot give a wind measurement at this level. The winds reported are at the mid- point of a four-minute window. The data used for the 1000-mb level are distorted because much of the window occurs after impact. The post-processed winds do not suffer from this limitation; however, it would be a significant programming task to change the algorithm to report a real-time wind close to the time of impact. The TEMP DROP code called for surface values of pressure, temperature, and dew point depression, which were difficult to measure accurately. The real- time system reports every ten seconds; thus surface data reported may contain some values collected after impact. Also the data are noisy as impact generally occurs at maximum range. A considerable improvement in estimation of impact would be possible. There is a meter showing received signal strength. As the drops nears the surface, the meter reading is usually very low, but it sometimes clearly shows impact. The Omega signals can be heard with a pair of ear phones and impact can often be detected by the change in the signal character. In the system used during FGGE, there was no way to relate these to the printed output. If, when impact was imminent, the operator could turn on a strip chart and record P, T, H, Omega, signal strength and time, then splash down could almost always be detected. With time known, an extrapolation of the printed data could result in greatly improved surface values. In a more sophisticated implementation, one would store data in memory and display it in some fashion. The human capability of the operator would be used to decide the instant of splash down and the com- puter would calculate smoothed values at that time. 6. PERFORMANCE OF THE AIRCRAFT ON-BOARD SYSTEM At the outset, it should be emphasized that the aircraft on-board systems had many inrportant improvements over the first-generation system during GATE in 1974. A great number of aids were supplied to prompt the operator at each step in the drop sequence. Many steps were interlocked to ensure necessary supporting data were entered on the recording tape. Several self- test features were used. Particular to the installations for FGGE, the on-board system had the fallback option of manually entering Heading and True Air Speed. During the latter part of SOP-I, the TAS signal was lost and the manual option was very valu- able. In a sense, the operator of any on-board system is operating blind. One does not really know if the system is working properly. That is, the question "Are the winds being reported correctly?" cannot be answered. There were some aids, however. First of all, one could listen to the Omega signals. In the Atlantic it was always comforting to hear Liberia come blasting through loud and clear. It was always easy to check synchronization. On the runway on Ascension, one could often hear five, and sometimes seven stations. The onboard system had the convenient feature of listing all the Omega stations, on command, and a measure of their "quality". While it was not a positive indication, one at least knew when real-time winds were bad or doubtful and knew they should not be reported. There was a direct and observable correlation between poor quality signals and unreasonable winds. 108 Two other signals were available to see "how goes it". On command one could print the flight-level wind. It is, as the name implies, the wind vector computed by the aircraft channel of the dropwindsonde system as the difference between the reported TAS and Heading and the movement of the aircraft through the Omega field of signals. It was not uncommon for this computed flight-level wind to diverge from that reported from the Inertial Navigation System. It was not clear what was going wrong, and there was no way to set in a best estimate — say from the INS. Instead, it was necessary to halt operation and go through the tedious steps of entry of NAVigation MODE followed by TAS, HEADING, MAGnetic VARiation, LATitude, LONGitude, DATE, and TIME. When in a hurry, it was never possible to enter latitude and longitude precisely, which probably contributed to new errors in flight-level winds. On command, one could also print the oscillator frequency error. It was never expected to be zero but ideally should be a small constant value. In the laboratory, such a condition was routinely achieved. In flight this was almost never the case. The operators were instructed to observe the oscillator fairly often and zero it out when needed. Fortunately, the single command, OSC, sufficed. The error estimate could be entered also, but experience showed that zero was as good a value as any. It wasn't going to stay where you put it any- way. This particular entry was rather abstruse to the operators and would probably be better monitored and corrected by the computer itself. The culprit causing diverging flight-level wind and oscillator frequency error was most likely disturbances of the Omega signals. It is hard to see how a high-flying, constant-speed, constant-altitude platform such as an aircraft could be the problem. Of course, the aircraft does change its geometric position within the field and flight-level winds do change, but these changes are not as drastic as indicated in the on-board system output. 7. CONCLUSIONS AND FUTURE RECOMMENDATIONS The low percentage of missions flown at OL-4 during SOP-I was at least largely due to the absence of the planned second aircraft. Higher priority missions precluded the assignment of a substitute aircraft when maintenance problems prevented the second aircraft from joining the operation. The fact that FGGE could not compete successfully for this necessary resource highlights an inherent difficulty in conducting a research program with operationally justified platforms. The OL-4 SOP-I I operations went fine, so such problems are certainly not inevitable. However, planners of future programs should remember that research activities are less vulnerable to short- notice develop- ments if they are not dependent on the use of military platforms. 109 CHAPTER 7 DROPWINDSONDE OPERATIONS PART 3: THE INDIAN OCEAN By J. McFadden (NOAA/RFC) 1. INTRODUCTION In the shadow of the Equator, at 7°20'S and lying halfway between Indonesia and the African Coast at ZZ^ZB'E, the island of Diego Garcia became the site of the Research Facilities Center's Indian Ocean Dropwindsonde Oper- ations in support of the Global Weather Experiment (FGGE). Geographically, Diego Garcia was the ideal location from which to launch double sorties daily during the two Special Observing Periods (SOPs). It is most unlikely that either the extensive area covered or the number of missions flown could have been accomplished from any other location in the Indian Ocean area. From the beginning of the FGGE planning phase, it was evident that a centrally located Indian Ocean base was essential to the operation. Several other locations could have been utilized at significantly higher costs and with considerably greater difficulty. These included Singapore to the east, Sri Lanka to the north, and the Seychelles to the west. Any one of these alone would not have been suitable to obtain full coverage of the Indian Ocean and splitting the available resources between two of the locations would have resulted in an operational, logistical, and communica- tions nightmare. Diego Garcia was clearly the best choice. 2. DIEGO GARCIA - THE ISLAND, ITS HISTORY, AND THE PRESENT FACILITIES In 1966, the United Kingdom and the United States signed a bilateral agreement making the island available to the two countries as a midocean com- munications base. Construction of the communications station, the airport, and the permanent base facilities by a Naval Marine Construction Battalion began in 1971 and continues unabated at present with fuel storage facilities, a deep water port, and permanent housing for all personnel. Today, Diego Garcia is a thriving community of over 1500 U.S. Navy personnel and about 25 officers and enlisted personnel from the British Royal Navy. While the base is almost exclusively used for U.S. military endeavors, the island is still part of the British Indian Ocean Territory and under United Kingdom jurisdiction. A Royal Navy officer oversees the island. The other British officers and men for the most part work along with the U.S. Navy personnel on the base. Customs, police activities, etc., all come under the purview of the British Representative and his staff. The permanent party of the U.S. Navy group on Diego Garcia is made up of the Navy Communications Station, the Navy Support Facility, which pro- vides personnel to support air and ship operations, billeting, food service, etc., and the Naval Marine Construction Battalion, which handles all heavy construction underway on the island. Located four miles from the town and a short distance south of the fuel storage area is Diego International Airport. To say that this was a com- plete and modern airfield would not be fair to those who operated there during FGGE. Certainly, the runway was long enough, even though a parallel taxiway did not exist, and the aircraft parking apron was large enough in most situations. The personnel in air operations, weather and communications v/ere talented and quite eager, and the working atmosphere was congenial. Beyond these things, there were many problems. 113 Fueling was a major headache. Although there were three fueling pits, only one fuel line existed which had as its maximum pump rate a flow of 500 gallons per minute with one aircraft being fueled. Add a second aircraft and the flow dropped to 250 gallons per minute. There were no hangars at the airfield in which to perform heavy main- tenance during extremely hot or inclement weather. Ground support equipment was in such heavy use and poor condition that long periods of maintenance down- time for this equipment were required. Special tools for aircraft maintenance were often not available on the island and had to be flown in from the Philip- pines. In a few instances, unnecessary delays resulted. For example, in order to change an engine, it was necessary to order a crane (if it was avail- able) from the motor pool, have it driven to the field, and use it to hoist the engine away from its mount. 3. THE PLANNING PHASE It was clear that in order for our operation to be a success, a great deal of effort and care would have to go into the planning phase. Negotiating with the Department of Defense and the United Kingdom for use of the island facility, making arrangements for support of the aircraft at Diego Garcia as well as other bases between there and Miami, arranging housing, messing^jand transportation for the 35 NOAA employees on the island, logistics, communica- tions, enroute over-flight and landing clearances, etc., all had to be con- sidered and carefully worked out. With the cooperation of the U.S. FGGE Pro- ject Office, in particular Mr. Onial Thomas, the U.S. Navy, especially CDR. Del Ritchhart, and the U.S. Air Force, the planning and preparation for our operation went very smoothly. Negotiations for the use of Diego Garcia were initiated in March 1976 with tentative approval from the Navy Department being issued in July of that year. Final confirmation of its availability was given in April 1978. Once NOAA had obtained these necessary clearances, it was then possible to deal with the U.S. Navy in planning the required support on the island for the RFC operation. Preliminary contacts were handled by the FGGE office and arrangements were made for an on-site visit in September 1978 by NOAA personnel to discuss support requirements. During this trip, discussions were held with Pacific Fleet personnel in Pearl Harbor, Hawaii, and with officers from various support groups at Cubi Point Naval Air Station in the Philippines. During this period of time, parallel efforts were underway in Miami by the RFC supply section to arrange shipment of spares and supplies by military airlift to Diego Garcia. As the shipping point for our supplies was Travis AFB, it was necessary to devise a method of getting them to California and onto a MAC flight along with minimizing the possibility that individual packages might be- come separated during transit or even disappear. The solution was to construct plywood containers which could be loaded with supplies for shipment and then used on the island for their storage. One pallet size and two quarter-pallet size boxes solved our problem. Transfer to Travis AFB was accomplished by truck. 114 For OL-3, Diego Garcia, the plan called for two flights per day from a total of three aircraft, two WP-3Ds and one WC-130B. Flight duration ms approximately ten hours for the P-3s and nine hours for the C-130. Our daily operating plan called for rotating the aircraft, thereby maintaining a schedule of two days up and one day down. While it would have been more desirable for FGGE program objectives to fly the P-3s as often as possible because of their longer range, follow-on projects (MONEX, EPOCS, and the Australia Tropical Cy- clone Program) dictated equal distribution of flight time among all three air- craft. Accordingly, the aircraft operations section of RFC prepared an opera- tions plan that took into account local conditions. Deployment and redeployment plans, communications, safety, etc., were also included in this document. 4. THE EXECUTION - SOP-I Between the completion of the planning phase and the beginning of the operational phase at Diego Garcia, several events of major significance occurred which severely impacted the early part of SOP-I. An understanding of the avail- ability of RFC resources at this time is necessary to explain the relatively poor performance during the first days of the experiment. Around the middle of December, the C-130 was flying its last mission prior to the holidays and our planned 5 January departure for the FGGE. During this short system test flight, the aircraft suffered a catastrophic failure of its #2 engine. The first stage of the turbine exploded, blowing away the clam- shell doors, causing severe damage to the engine mount, and cutting several gashes in the fuselage section of the aircraft. Preliminary estimates on the time required to obtain parts, to build up the engine with a new turbine sec- tion, to replace the engine mount, and to repair the fuselage section was one month, particularly considering labor problems over the holiday period. The aircraft was repaired and departed Miami on 18 January, or two weeks later than originally planned. One P-3, N42RF, had participated in the Winter MONEX program in Malaysia during November and December and because of a needed phase inspection and the total flight hour commitment for the FGGE and the follow-on Australia pro- gram, RFC elected to store the aircraft at Cubi Point, NAS, in the Philippines over the holiday break and to return there on 5 January to begin a normal fi ve- to seven-day phase inspection. Unexpected maintenance problems, such as fuel leaks, a prop seal leak, and a cracked center fuel tank cover plate produced a delay in the departure date. This was further complicated by clearance problems for Singapore resulting in a delay in the arrival of N42RF at Diego Garcia until 18 January. As a consequence of these two events, RFC began SOP-I at OL-3 with one aircraft, N43RF, and about 20 people. We quickly revised our operating plan to fly single missions with one P-3 starting on the 15th and double missions begin- ning on the 18th after the second P-3 arrived. This was not to be. On 17 January, N43RF suffered a failure of the turbine stage of the #2 engine and we were quickly introduced to the rather painful experience of 115 maintaining an aircraft at a facility where neither tools nor spare engines are kept on location. Nine days after our request for a replacement engine went out, we finally received shipment. Two days later, we were back in the air. The remaining portion of SOP-I was relatively uneventful from an oper- ational point of view. True, we did have a gear box failure on N42RF which also necessitated an engine change, but the logistics system was primed by then and we were back in the air in less than half the time for N43RF. During SOP-I, the Naval Weather Service Environmental Detachment at Diego Garcia had the capability of receiving only the Japanese GOES satellite which provided coverage of the Indian Ocean area east of the island. Weather information west of the island was derived from ship and aircraft reports. The principle tracks covered during SOP-I are shown in Figure 5 in Part 4 of this Chapter. In all, 61 missions were flown during the period. On 20 February, NOAA secured its operation on Diego Garcia and de- parted the island for Singapore. From Singapore, one P-3 went to Australia for the Tropical Cyclone Project and the other P-3 went to Guam to begin the Equatorial Pacific Ocean Climate Study. The C-130 returned to Miami. 5. THE BREAK PERIOD During the two-month break period between phases, primary attention was given to analyzing our performance during SOP-I with an eye to improving our operation. One of the major areas of concern was the inadequate housing the NOAA personnel had during the first phase of the project. Aside from the lack of air conditioning, the communal type living in cramped quarters created a serious morale problem near the end of the period. After working and eating with your co-workers, it was also then necessary to sleep along side of them. There was simply no privacy to be had. In attempting to obtain better quarters for SOP-I I, we were made aware that naval activities v/ere on the upswing in the Indian Ocean and, if indeed we were permitted to return to Diego Garcia in May, the quarters that we would be assigned would be even more austere than we had during SOP-I. This was considered by RFC to be unacceptable and after consulting with the FGGE Project Office, the process of trying to obtain clearance to operate the second phase from Sri Lanka was initiated. Although NOAA did obtain permission from the government of Sri Lanka to operate from that location, the decision was made after much thought and discussion to return to Diego Garcia and make the best of the situation. RFC did go through the process of drawing up a contingency operations plan for Sri Lanka in the event it was necessary to use it at any time during SOP-II. 116 6. THE EXECUTION - SOP-II All three aircraft were scheduled to depart for Diego Garcia on Tuesday, 1 May, with intermediate stops in California, Hawaii, Guam, and Singapore. On Friday, 27 April, ten minutes before the end of the business day, RFC received a call from Lockheed Aircraft Company indicating that all P-3s had been grounded pending an inspection for cracks in the support brackets in the outboard fuel tanks. The procedure required defueling of the aircraft, opening and airing the tank with blowers, and inspecting the support brackets. Upon completion of the inspection, the tank was to be resealed and refueled. The entire process required at least two days per aircraft, particu- larly in our situation where outside contractors and subcontractors play a key role in accomplishing the work. As the alarm was sounded after RFC and con- tractor personnel had departed for the weekend, no useful work began until 30 April. This resulted in one aircraft, the C-130, leaving on schedule on 1 May followed by one P-3 on each of the two following days. Arrival in Diego Garcia was also staggered with one aircraft arriving on each of three days beginning on 8 May. Other than an engine failure on N42RF and complications because of the increased Fleet activity, SOP-II was fairly routine. The NWSED was re- ceiving TIROS-N data during this period and the satellite information was yery useful for track selection. Toward the latter part of the observing period, several attempts were made to coordinate our flights with the Summer MONEX flights originating in Bombay. For the most part, owing to communica- tions difficulties, this was unsuccessful. Primary flight tracks are shown in Figures 8 and 9. Operations on Diego Garcia were secured on 9 June with the depar- ture of the C-130. All spare parts and unused supplies were prepared for shipment by MAC prior to our departure. Subsequent airlift to Miami occurred without any problem. 7. OMEGA GROUND MONITORING STATION In support of the FGGE and under the direction of Dr. S. A. Rossby, RFC established an OMEGA receiving station on Diego Garcia to monitor the quality of the OMEGA signals. The purpose of the monitoring was threefold: (1) To detect transmitter outages; (2) To detect and note times of anomalous propagation, and; i '. ' ' i ' , (3) To detect and note times of phase anomalies resulting from Sudden Ionospheric Disturbances (SID). Each of the above have the potential of seriously degrading the quality of wind measurements made by the OMEGA Dropwindsonde System (ODWS). 117 With regard to numbers (1) and (2) above, if the effect is severe, there is nothing to be done but flag the questionable data. In the case of a SID (3), there was some thought that we might be able to model the effect of the disturbance with the ground station and apply some correction to the affected wind measurements. However, very few SIDs of any consequence were observed in SOP-I and II. Therefore, no corrections will be attempted as the effort is not warranted. The receiving system consisted of two Tracor 599-R OMEGA navigation receivers, a Tracor 304-D rubidium frequency standard, and a Wintronics MIIE multipoint strip chart recorder. Using the frequency standard as a reference, the phase of each re- ceived signal was plotted on the chart recorder. Data obtained during the two SOPs were sent to NCAR as were all other data obtained during the exper- iment. 8. CONCLUSIONS AND RECOMMENDATIONS , '.;,.;/ The success of our mission in the Indian Ocean can only be attributed to the hard work of a dedicated group of people. Regardless of how poor oper- ating conditions are, there is still a tremendous sense of pride among the individuals in the organization to demonstrate that the job can be done. Add to this a certain inter-plane rivalry that exists in RFC and you end up with a healthy operational situation. I think the results of these were clearly demon- strated in the FGGE. If the question ever arose as to whether RFC would return to Diego Garcia for a follow-on program, the answer would be yes. (This is not to say that another location would be more preferable to our personnel.) Within the next year or so, the construction of permanent quarters will be completed and air-conditioned spaces will be available for all visitors to the island. Addi- tional facilities, including a hangar and new fueling systems, are being con- structed at the airport, and the complete ground support facility is being upgraded to handle a much heavier load of air traffic. The major recommendation I would make if a follow-on program is planned is to utilize four aircraft in the operation. While RFC was reasonably success- ful in meeting program objectives with three aircraft, the burden was quite heavy considering the lengthy downtime experienced on three separate occasions when an aircraft required an engine change. A 100 percent mission achievement would be within the realm of probability with the additional aircraft. As a final comment, it should be pointed out that the island is a beautiful place. The Navy personnel we worked with, both British and American, were understanding and very cooperative. There was a mutual support understand- ing among all of the aircraft operators whereby we all helped each other. I am happy to say that NOAA's presence saved the day on several occasions for both the ASW P-3s and the S-3A from the "MIDWAY". Then, too, they saved us a number of times. The experience of being on Diego Garcia was a memorable one. To have missed it would have been disappointing, but no one seems to be in a hurry to return. 118 Table 1. --Diego Garcia - Significant events SOP-1 DATE EVENT January 5 N43RF departs Miami January 13 N43RF arrives in Diego Garcia January 13 N42RF is in Cubi Point because of maintenance delays January 15 N43RF begins SOP-I. N42RF delayed because of clear- ance problems January 17 N43RF loses #2 engine because of turbine failure. Mission completed January 17 N42RF departs Cubi Point - cracked nose wheel noted during refueling in Singapore January 18 N42RF arrives in Diego Garcia January 18 C-130 departs Miami January 19 Single missions resume January 26 C-130 arrives in Diego Garcia January 27 Daily double sorties commence January 28 N43RF in commission February 3 N42RF aborts due to gear box failure of #2 engine. Engine change required because necessary stands and engine supports are unavailable on island. February 8 N42RF in commission February 8 C-130 loses pressurization six (6) hours into the flight. Major portion of mission was completed. February 10 C-130 - Bleed air duct rupture mission aborted February 18 N42RF departs Diego Garcia for Singapore February 20 N43RF and C-130 depart Diego Garcia 119 Table 2.--0L-3 Diego Garcia - Significant events SOP-11 May 1 May 2 May 3 May 8 May 9 May 10 May 15 May 16 DATE , EVENT C-130 N6541C departs Miami N42RF departs Miami N43RF departs Miami C-130 arrives in Diego Garcia N42RF arrives in Diego Garcia SOP-II operations begin; N43RF arrives in Diego Garcia C-130 grounded by hydraulic leak N43RF flies modified N40 track with termination in Colom- bo, Sri Lanka. Stopover was made to test feasibility of conducting operations from that location should it be necessary. C-130 hydraulic line part obtained from USS CAMDEN, Seventh Fleet support ship N42RF loses #1 engine. Navy was reluctant to release on-site T56-14 engine to NOAA C-130 down because of broken stud on oil scavenger pump C-130 problem corrected with assistance from Naval Marine Construction Battalion May 25 Daily cable to MONEX Office, Bombay initiated contain- ing OL-3 FGGE flight schedule Engine change on N42RF completed Northeast corner of track N31 truncated because of Seventh Fleet operations in that area N42RF "buzzed" by A-4, based on carrier USS MIDWAY N42RF departs Diego Garcia for Singapore N43RF departs Diego Garcia for Bangkok C-130 departs Diego Garcia for Singapore. FGGE mission at OL-3 terminated May 17 May 17 May 21 May 22 May 'i ?9 May : 31 June 3 June 7 June 8 June 9 120 Table 3.--0L-3 Diego Garcia - NOAA participants POSITION NUMBER OL-FGGE Director 1 Pilots 8 Navigators 4 Flight Engineers 6 Mechanics 4 Loadmaster (C-130) 1 ODW Operators and Electronics Maint. 8 Supply 1 Engineering Technician 1 Avionics Technician 1 121 CHAPTER 7 DROPWINDSONDE OPERATIONS PART 4: FLIGHT TRACK AND MISSION SUMMARY By 0. Thomas (NOAA/OAl) 1. INTRODUCTION This summary of flight tracks and missions flown is a supplement to the information provided in Parts 1, 2, and 3 of this Chapter. It is presented as a reference source for performance statistics and as a quick-look recapitula- tion of the accomplishments of the Aircraft Dropwindsonde Program (ADWP). 2. SOP-l SUMMARY The initial projection for SOP-l was to fly a total of 180 dropwind- sonde missions over the three oceans. 174, or 97%, were flown, although over a 37-day period rather than the originally scheduled 30 days. The 7 day extension was arranged because clearance difficulties and maintenance problems resulted in a slow start for the ADWP program. Only 24 missions were flown in the first 8 days vs 48 planned. Clearance difficulties were most severe for the eastern Pacific base, where it finally became necessary to abandon original plans for SOP-l opera- tions out of Acapulco, Mexico. Instead, these operations were temporarily conducted from Norton Air Force Base in California and then moved to Howard Air Force Base, Panama Canal Zone for the duration of the SOP. Maintenance/ aircraft availability problems were most serious at Ascension Island in the South Atlantic where only 37% of the planned number of missions were accom- plished. 3. SOP - I I SUMMARY SOP-II ADWP operations started out much more smoothly than had been the case in SOP-I. Clearance problems with Mexico were resolved, and the utilization of Acapulco made possible a daily shuttle flight between Hawaii and the North American continent, thereby covering an oceanic region which had been data-void during SOP-I. Another crucial improvement was achieved in the Atlantic where the substitution of C-141 aircraft for the WC-135 used during SOP-I contributed to accomplishment of 97% of the planned missions during SOP-II. Overall for the four operating locations during SOP-II, 92% of the scheduled missions were flown during the 30-day operating period. 125 O Q. X s- +-> ai O &- 01 o s- D. ■o o to T3 C •r- Q. O S- o li- fe s- u S- O E I I I O) J3 rt3 = sJ 2 = ^ •M ■M +J -(-> -t-> -M +-» « i. in c c = s m c r = 3 ; s c c 12 ; — ^I^'T +-" 0) r* £ •■- E OI ■r- 2 •>- E •^ = <_) m in fo ^ a: fT3 OJ <0 OJ oc rO OI 13 OJ nj 0) SO -^ 4-» m -M •a: z .— z — ■s.— z •— S r— «^ (/» .— 1/1 3 x> ^ 3 J3 J3 J3 1 s- 01 cj <-) , O) J3 CO 1/1 Q. U> U. *J f- ■0 i_) 13 1/1 c -^ z s = c ^^ s- z z ; = r z : 1 C s. a. = (U o< M <: Q. u — (/I Z3 •M i. W m OJ DIEGO GARCIA EAST OL 3 C S. 4-> s. < ja a)*J Qj = s < -1^ -u 00 u ■- (0 .— -w en s. -H C »-l u. ./) c ^, 1. 1 X 1 x> (o •<- «I Q. _ < _ CJ < ■ 5 iZ _ _ 4^ = OJ CJ ■tJ 3 u JO ■— I ^^ t— CVJ J3 4-> -r- c 3 <_) ■w L. j: w u CL —15 1 2 Q 01 J3 "-^ .— U- 1— — = a. a- LlJ -0 C OJ ■13 > t/1 s c ■.- (U > c 2: ^J 5 a. C3- f^ r^ un r^ CO CM 01 r^ ~ 10 in on 1 ^ ^ VO "3- in , 10 CJ1 r^ 00 ro r^. «T C^ CO •a* m C\J m CO CM LYING TIME HOURS AND NTMS' • 1 ■ UD oi oi oi d 01 «!• U3 00 m d d CO d •sO d 00 CM d 10 d kO d CO f^ 01 P^ d d ro CSi 3- ^ tn 10 in 10 m in in ■=r LO in r^ 00 10 r^ 00 00 10 ^ r^ CJ1 CO 00 r^ CD 00 «a- CO r^ 00 en I^ 01 r^ U3 10 f^ 10 01 a parioaiioo * CM 3 uaqiunN * ' 0; pai^Dune"! m t— 1 ro in f^ p^ ^r 00 CO V£> m ,_4 in ^^ 10 00 r^ CM OI 01 CO C« r^ CO ro CM ID m lO 10 01 10 1^ ,— 1 r^ r*% r»1 m un C*1 in un m m CTi CO 01 10 P-. 00 £31 CO 00 01 uo CO CO 01 r^ r^ f^ 10 CO CO CM .43qmnN * 00 CM « SUOLSSIW CNJ m ■cr m m •a* ro CM ■a- tn in «s- in 10 lO ^ IT) LT) 10 10 lO in in in 10 LO LO m lO 10 in 10 in m in U1 m J.0 jaqiuriN ]s z ■z. 1 z Iz 2= 1 1 1 Z Z Z IZ z z z c c z 2: 03 CO CO CO CO CO CQ CD CO CO CO CO CD CO CO 00 CO CO CO CO LU 01 < a: et •a: cc r^ eo cy>lo 1.^ CM ro ^■ m U3 I-. CO OI r •"^ ""^ ■^ '^ CNJ CM ^ c ~ CM CM C^J ixi CM ro m " '~* '"' ■^ r Cvj \— 126 Table 2. --Detail dropwindsonde operations Global Weather Experiment California/Canal Zone OL-1 SOP-I DATE AIR- CRAFT TYPE TRACK FLOWN DROPWINDSONDE FLYING TIME (HOURS AND TENTHS) AA = Air Abort AMD = Aircraft Mechanical Difficulty 1979 Q UJ ■z. < Q UJ ^? (/I I— EP = Equipment Problems (ODWS) NF = No flight POS = Poor Omega Signals TM = Track modified to avoid convection 15 Jan Awaitina Diplomatic Clearancp 16 Jan Awaiting Diplomatic Clearance 17 Jan C-141 P-01 11 11 6 11.2 DroDS beqan 155 18 Jan C-141 P-02 11 10 4 11.3 Drops beqan 15S 1 streamer 19 Jan C-141 MS; EP (Program would not load) 20 Jan C-141 P-02 10 9 6 11.2 Minor EP 21 Jan C-141 Stand down for move to Canal Zone 22 Jan C-141 Moving to Canal Zone 23 Jan C-141 P-03 18 16 9 10.3 24 Jan C-141 P-03 20 17 8 10.3 2 streamers 25 Jan C-141 P-03 20 19 3 10.5 1 streamer 26 Jan C-141 ?-.T^ 18 18 9 10.6 Minor EP 2; Jan C-141 14 14 8 10.7 Special track to California 28 Jan C-141 15 14 3 10.2 Special track to Canal Zone 1 streamer 29 Jan C-141 P-05 9 8 4 8.6 AA (Proqram bombed reoeatedly) 30 Jan C-141 NF; EP (will not hold proqram) 31 Jan C-141 P-05 21 20 14 10.9 Winds suspect on drops 14 thru 19 1 Feb C-141 P-05 18 17 14 10.9 EP (cannot load calibration tapes) 2 Feb C-141 P-05 19 14 10 11.0 EP (proqram bombs) 3 streamers 3 Feb C-141 P-03 17 14 3 11.0 EP f proqram bombs) 4 Feb C-141 P-03 14 10 8 11.2 EP (proqram bombs) 5 Feb C-141 P-03 20 18 9 10.7 EP (proqram bombs) 2 streamers 5 Feb . C-141 P-03 20 19 16 10.9 EP (manual OP) 1 streamer 7 Feb ' C-141 P-03 23 20 18 10.6 8 Feb C-141 P-05 10 8 5 11.2 EP (program bombs) 9 Feb C-14i P-05 19 19 11 10.5 EP (tape reader/launcher) 10 Feb C-141 P-03 21 21 10 12.0 I/C with Albrook EP 11 Feo C-141 P-05 22 18 12 10.2 FP (manual) 3 streamers 12 Feb C-141 P-05 1? 10 8 10.4 FP (nn baseline) 1 ^itreamer 13 Feb C-l^l P-03 21 20 10 11.1 1 streamer 14 Feb C-141 P-05 13 12 11 8.2 TM 1 streamer 15 Feb C-141 P-05 22 19 17 11.1 EP 16 Feb C-141 P-03 19 17 8 11.0 1 streamer EP (baseline) 17 Feb C-141 P-05 20 17 11 11.0 TM 18 Feb C-141 P-05 19 17 8 11.0 19 Feb C-141 P-G5 19 19 11 10.5 20 Feb C-141 P-05 15 15 3 9.8 TOTAL 530 480 297 330.3 127 CVJ I to fO S I O S- S- +-> OJ o CO c o -M s- O) Q. O OJ XJ c o CO Q. C S- ■o fa +-> o I CO d 1- s. c a c X c o a >, o i/- ■u w V -— r— > - — ■ s- 1/1 3 C e Ol d s- o o ^~ s- t5 Ol c 10 c E M- c/) G C- D ■,- re <*- 3 "O •r- 1/1 irt s- (U s. CM •1- Q -1- ■1-1 — Q o I. (J O O O u CJ oa o ^J H- 1— ^— - > 01 s. +J t/> ' — ' 1 — re > (J > a XI t. n tA i/> c h- s. a ro •^ ■1- X3 cj E »— o a t rt x: CM x: T' ■^- •.- a; re ■•-> C r- c re Si a^xj J- E 5 o c Q S- X H O > re .CO 'f- 'f— (. o ^ — ' c ^- 4-- - — ■ X o 111 Ci_ <+- a r c i. 4-> Air Abort Aircraft M( Equipment 1 No flight Poor Omega Track modi Q XI (/► a Ol re t/> L (/ CU ■u c z r c s- 4- re q +- XJ O' a Sh o> •ij a O to c c re s- .c O re <- 1. s_ z S- s_ cr s. t 01 4-> XI a OJ 4-J >d i re Ol 0) -o .a VI -C n: Q 0) E Ol Q > re t/l 43 (/) o 4- E re • I-- n. o ■^ i XJ $ o re , — in u n. 01 0) t/1 II M n 1) H II u. c (. c +J i/> cr s- 1- «t Q a. ii. oo E d S uj z: O P -t-J o S- ^_i ^ • c c o J H OJ E 3 4J l/> 4-' 5 0) t 3 4-> "?: 3 rt s_ ct o a. Q- 3 o. o 1 — cC Q. ~ CO t_i <_ !; cC Cl. LD UJ <_) LU c/) Ll. J3 in 1-. c CNJ CO O O o 00 o CT> CM CO O r-\ - I— * cn o CO CM r-H CM o ^ cr ID O o 00 o LO CD CM f— 1 cn c> 00 o t— 1 O in in CO cr> I— 1 »-H .— 1 r- 1 »— 4 »— 4 .— 1 ro Si9 NO ir> O o lO CM ■* 00 o O CO O^ CO r^ o 00 O o o ID a- c (Ti CTl lO CO o lO 'd- Ol -o ID i^ en in f— 1 ^A r-l .—4 f- 1 — 1 r-^ UJ CNI a z o 00 o z VIVQ LT) O f^ CM CO r^ O o c^ CTi lO o- o CO r^. r^ * «* «* '^ ^ "* ^ ■* «* W "* ■* ^ >* »* *~ ' ' *^^ ' ' '~~* t — t '~^ ' ' ' ' 1 — 1 ' ' 1 — 1 *"' ^"^ • * ' ' ■ — ' ' — ' ■ — ' '—^ ' — 1 ' — ' ' — ' ' — ' ' — ' r— 1 < — 1 1 — • .— 1 1— ' r-l r-l f— 1 « — • • — 1 1 1 u t t 1 t 1 1 o tJ 1 c_) tj 1 1 c_> (_) 1 c_) 1 O 1 (_> 1 <_> 1 1 1 L_l o 1 1 O 1 1 1 1 1 1 1 C_l C C C c c C c c c _ c C c c -O -Q J3 X) -Q ■a XI _ Xi XI X3 XI x> XJ XJ XJ XJ XJ XJ LU CT> re re re re re re re re re re re re re re re re re OJ ^ 0) Ol 0) 01 0) CU 0) o» 0) 0> OJ -J 1— r^ ^o -r> n —3 ^3 T) -3 -3 o -3 -o T 'T) -3 "^ o -^ u. Li. U. U- U- ll. Ll. LL. Li- U- U- Ll. U. u. ■a- ll. IJ_ u. Li. ■< < a> y- Q ^ \n ^£) •-- 00 cr. o r— f CM ro ■3- o-i kO l-~ CO CTi o »— 1 . — 1 CM ro «* in ko r^ CO cr> ^^ CM ro •=1- LD lO r^ 00 cn O o *-H ^^ w-i -— t .-H r CSI CM CM CM CM CNI CM CM CNI ro ro ■"* 1—1 '"' r-H ■"* ""^ '~~* *~* •"* CNJ 1— 128 Table 4. --Detail dropwindsone operations Global Weather Experiment SOP-I Hawaii - West OL-2 DATE AIR- CRAFT TYPE TRACK FLOWN DROPWINDSONDE FLYING TIME (HOURS AND TENTHS ) AA = Air Abort AMD = Aircraft Mechanical Difficulty 1979 Q UJ 3: (_> z o LU o 00 1— ta z o EP = Equipment Problems (ODWS) NF = No flight POS = Poor Omega Signals TM = Track modified to avoid convection 15 Jan C-141 Ground sbnrt; FP fOmpga nnf) 16 Jan C-141 P-30 13 13 6 10.8 17 Jan C-141 P-30 5 4 2 7.0 AA; EP (Omeqa weak baseline inoo) 18 Jan C-141 P-30 13 12 6 10.2 Heavy cu west end 15 Jan C-141 P-30 17 13 8 11.0 20 Jan C-1.41 P-30 11 10 5 6.9 AA; AMD (fuel valve) 21 Jan C-141 P-30 5 4 2 7.4 AA; EP (Omeqa & Baseline prob) tl Jan C-141 P-30 2 2 1 4.1 AA; AMD (enq. inst. converter 23 Jan C-141 P-30 20 19 10 10.9 24 Jan C-141 P-30 20 15 D 10.6 EP (manual OP, 4 streamers) 25 Jan C-141 P-30 22 20 11 10.7 26 Jan C-141 P-30 16 12 7 10.2 EP (Omeqa bad after drop 14, 2 streamers) 27 Jan C-141 P-30 19 16 9 10.6 EP (manual only, 1 streamer) 28 Jan C-141 P-30 22 17 10 10.7 .25 Jan C-141 P-30 21 19 9 10.5 30 Jan C-141 P-30 19 18 . 10 10.2 31 Jan C-141 P-30 4 ■0 4.4 AA; EP (enq. hvd. leak) 1 Feb C-141 P-30 14 14 7 , 10.1 EP; launch tube 2 Feb C-141 P-30 20 20 9 10.3 3 Feb C-141 P-30 21 19 9 10.4 EP; launch tube 4 Feb C-141 P-30 21 21 11 11.0 5 Feb C-141 P-30 19 15 9 10.7 4 streamers 6 Feb C-141 P-30 19 18 9 10.5 7 Feb C-141 P-30 19 19 9 10.7 8 Feb C-141 P-30 2 6.0 AA; EP (computer out) 5 Feb J C-141 P-30 19 18 9 10.3 10 Feb CO 141 P-30 21 18 7 10.5 3 streamers 11 Feb C-141 P-30 19 19 9 9.8 12 Feb C-141 P-31 20 19 10 10.3 13 Feb C-141 P-31 20 18 9 10.1 1 streamer 14 Feb C-141 P-30 14 14 7 8.8 AA; AMD (autopilot) 15 Feb C-141 P-30 14 14 7 10.6 16 Feb C-141 P-30 15 14 9.9 Several deviations for Wx 17 Feb C-141 P-30 11 11 4 7.5 Late T.O. for AMD; cut short for crew duty 18 Feb C-141 P-30 18 18 6 10.3 19 Feb C-141 P-30 17 16 3 10.3 20 Feb C-141 P-30 20 19 5 8.3 TOTAL 572 518 241 342.6 129 1 O 1/1 ■•-» i. 03 CO I 1— o o o +-> CO +-> fO ra S- •1 — O) u Q. S- O fO o OJ -a o E O) O O) to •r- -o Q c •r- s Q. O s- T3 CO ■fJ (U O I 0) JQ UJ CO en ;z VD CM T3 LlI C o in en ro LlI >> o VI t/O in in CO r- > 3 C 1 1 o > o JD ro z ro Diffi( ODWS) 'oid c( 1- ro ro o +-> XJ E O o 00 ro ■»-> ro -o > 01 CM ro > — ^ r— (0 O) o; o o> 0) ?' a O) o_ u s_ 4-> > l_ , S- IS, S- u E .— o •1-0) <0 4J ■o 3 x> OJ o a ro ■o o O) Z3 c o c c o S i o a. ro C r— C ro XI o>-c> CM s- a c ro o cr ro ro 4- IC ai oo x: o •>- (u 1 01 4-> O" N r- o •r— 0) (X <4- XI oo in Q X o s_ s. oo D OO s: ro t- cr c c o 4- J3 ^ o CD o +j +J +J cri-O ^ o i_ Ol i/i 1/ 4-> CD -I-' S- -M C -E 01 Q (J "O o i_ t- 1/1 1/ n re £ •.- O 3 X) 01 in 0) c: ro Ol 3 o; s_ 4-> 4-" SZ ■>-> =1 J~ Q-.— -ii CJ 10 > > ro 1 o .^ Q c O a cr O •!- <♦- I. O c 4-> ro c ro *. — •f -. — t- 1- 3 o ro QJ sz c o ex it- 1 — t — I — •■- -r- cr o o s- o 01 1- o C3 O c <+- i+- <+- «« cc uj 2: n. f— O -o ro +-> c: ■a: >- 1— 5 ,— r— ,_ II II II II II II <^ o ro o ro XI to 1— 1 1 ro ro ro CD o .— 1 no CT» 00 r^ CO i^^ r-- m ro , CO Si9 NO r^ CO CO lO r^ * tn r^ m r^ CM r^ ID Ln ID ro r-^ ,_ r^ r^ no CO ID in >* ^ m CM o LU O z o i-i CM ~ 1/1 a z Viva ^ LD IT) CM vo I— < in ^ ^ in »* C3 rM r-H C3 in r— 1 m r\j T-H f—f .— 1 cn ro ^.. 0) s_ 031331103 .-H r-l CO ■^ in tn in i-H .— 1 '^ll o o r^ 00 o ex en ro 3 ro ^ Cl. o cn c o oo 03H3NnV1 r^ 00 r^ ^ 00 CM lO ID CO CO lO CO ■=d- ID kO UD CM 00 tn in ro t-f r—l r-H ro r^ . — 1 •— 1 t~i .-H in r~^ in m "* 1 — 1 ro o ro r-^ CO o in o in i^ z t—i i-H in UT) LD in in in in in o o CD C3 in o in in CD o o o CD in CT O o. ■=1- 1 •* ■ta- ^ >* 1 1 •=t 1 "^r 1 1 1 I <* 1 1 1 1 1 1 1 1 r— 1 t t in <* in in o CD cl CD L 1— U. :^ s: ^ ^ ^ ^ 2: ?^ z: ^ 2: ^ ^ ^ e: ^ 2: ^ ^ s: ^ p^ ^1 2^ ^r; ^ 1 I 1 1 1 z 1 Z 1 z K 1 z 4-> . h-i^ o o o o o o o o o o " C ro ro ro ro CO ro ro ro ro ro r^ s o o CD o Ol AIR- CRAF TYPI oo ro ro ro ro ro ro ro ro ro ro ro ro ro n ro ro ro ro H- 1 1 1 1 1 1 1 1 1 1 1 CO V t 1 1 1 1 1 1 1 1 ro ro ro ro ' — ' ' — ' r~* ro '~* ' — ' ro Q_ o_ CL Q. Q_ CL a. CL Ol CL O- ej o CJ CJ o. CJ <_) a. CJ CJ CJ CJ CJ CL 1 1 1 1 ' 1 ( 1 1 1 1 i- Q, CL CL <-^ CJ CL CJ CJ CL. O 1*- ro c c c c C c c c c c C C ^ c c c c -Q n JD X> JD X3 J3 -Q £1 n J3 X3 -O XI XI XI Q a XI XJ Q 4-> _l uj cr. ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro ro o -2 Lu U- Uu Ll, Ll Lj- U_ U- Li- IL Ll IJL Uu u. Ll Ll '1- Ll_ Ll U- Ll cn 1— cC en o Q r-l fji lO h^ 00 cri o r-l CM ro "* in ID F^-- CO en o 1 — 1 . — f TM ro "* in ID r^ co en o f-H CM ro ^ in ID I-- CO en o C5 * I-- ^ r—l ' ' r— 1 CNI rvi CM CM CNJ CM CM CM CM CM ro ro "" "" *--< ^ T- "^ '" '" CM CM 130 Table 6. --Detail dropwindsonde operations Global Weather Experiment SOP-I Diego Garcia - West OL-3 DATE AIR- CRAFT TYPE TRACK FLOWN DROPWINDSONDE FLYING TIME (HOURS AND TENTHS) AA = Air Abort AMD = Aircraft Mechanical Difficulty 1979 Q LU a: o z o UJ o I/O 1— to z o EP = Equipment Problems (ODWS) NF = No flight POS = Poor Omega Signals TM = Track modified to avoid convection 15 Jan Onlv one a/c on station 16 Jan Only one a/c on station 17 Jan Only one a/c on station 18 Jan Second P-3 arrived on station 19 Jan P-3 AMD (engine rhange) 20 Jan P-3 AMD fenginp change) 21 Jan P-3 AMD (engine chanae) 22 Jan P-3 AMD (enaine chanae) 23 Jan P-3 AMD (enqine chanqe) 24 Jan P-3 AMD (enqine chanqe) 25 Jan P-3 AMD (enqine change) 26 Jan P-3 AMD (enqine chanqe); C-130 arr on sta 27 Jan P-3 N-33 16 13 6 10.0 28 Jan P-3 N-33 15 15 3 10.2 Minor EP (tapedeck) tape may be bad 29 Jan P-3 N-33 16 13 7 9.9^ 30 Jan P-3 N-33 16 IS 7 10.1 31 Jan P-3 N-33 16 14 6 9.9 1 Feb P-3 N-33 16 15 7 10.1 2 Feb C-130 N-23 12 11 5 8.2 3 Feb P-3 N-33 17 16 7 10.3 4 Feb P-3 N-33 19 14 7 10.2 2 streamers - weak Omega 5 Feb P-3 N-34 17 11 5 9.8 INE prob. - 1 streamer 6 Feb °-3 N-34 16 13 5 9.9 7 Feb P-3 N-34 18 15 6 10.1 8 Feb P-3 N-33 15 15 7 10.1 9 Feb P-3 N-34 19 15 8 10.0 10 Feb P-3 N-34 19 13 6 9.9 No Omeqa 5 sondes - Omega weak 11 Feb P-3 N-34 18 15 7 9.8 12 Feb P-3 N-34 16 14 7 9.8 13 Feb P-3 N-34 16 13 6 10.3 14 Feb P-3 N-34 16 14 5 10.0 15 Feb P-3 N-34 13 14 7 9.8 16 Feb P-3 N-33 16 14 7 10.2 17 Feb P-3 N-33 17 15 5 9.3 7 sondes no T&P - sensor cans empty 18 Feb P-3 N-34 10 10 5 9.7 500 km soacinq 19 Feb P-3 N-34 14 14 5 9.8 500 km spacinq 20 Feb NF Both a/c to Singapore TOTAL 388 331 154 237.9 131 Table 7. --Detail dropwindsonde operations Global Weather Experiment SOP-I Ascension OL-4 DATE AIR- CRAFT TYPE TRACK FLOWN DROPWINDSONDE FLYING TIME (HOURS AND TENTHS) AA = Air Abort AMD = Aircraft Mechanical Difficulty EP = Equipment Problems (OOWS) NF = No flight POS = Poor Omega Signals TM = Track modified to avoid convection 1979 Q Q LU O (/I 1— o 15 Jan No aircraft on station 16 Jan No aircraft on station 17 Jan No aircraft on station 18 Jan No aircraft on station 19 Jan No aircraft on station 20 Jan No aircraft on station 21 Jan No aircraft on station 22 Jan WC-135 arrives on station (1 crew) 23 Jan WC-135 A- 10 4 1 1 4.7 AA; EP 24 Jan WC-135 A- 14 18 10 7 8.8 Modified track to 9 hr so crew could flv 25 Jan 25 Jan WC-135 A- 14 22 15 7 10.7 26 Jan WC-135 Crew rest NF 27 Jan WC-135 NS; AMD (bad battery) 28 Jan WC-135 A- 10 21 12 6 8.8 CWI - baseline 29 Jan WC-135 A- 10 16 15 7 8.9 1 streamer 30 Jan WC-135 NF: AMD fhvdraulic line broken) 31 Jan WC-135 NF; AMD (hydraulic line broken) 1 Feb WC-135 NF; AMD (hydraulic line broken) 2 Feb WC-135 A- 14 18 16 7 9.2 3 Feb WC-135 A- 10 19 18 9 9.3 4 Feb WC-135 A-10 23 16 9 10.5 1 streamer 5 Feb WC-135 NF; crew rest 6 Feb WC-135 NF; AMD (cracked windshield) 7 Feb WC-135 NF; AMD cracked windshield) 8 Feb WC-135 NF; AMD (cracked windshield) 5 feb WC-135 NF; AMD (cracked windshield) 10 Feb WC-135 NF; AMD (cracked windshield) 11 Feb WC-135 NF; AMD (cracked windshield) 12 Feb WC-135 A-IO 21 15 7 10.2 EP (TAS synchro failed) 13 Feb WC-135 A-10 24 17 9 10.5 EP (TAS synchro failed) (1 tape head inoo) 14 Feb WC-135 NF; AMD (radar inop) 15 Feb WC-135 A-10 19 17 8 8.6 EP (1 tape head inop) 16 Feb WC-135 NF; AMD (hydraulic Dump out) 17 Feb NF; AMD (hydraulic Dump out) 18 Feb NF; AMD (hydraulic pump out) 19 Feb NF; AMD (hydraulic pump out) 20 Feb NF; AMD (hydraulic pump out) TOTAL 205 152 11 ioo.2 132 Table 8, --Summary of flight tracks for Global Weather Experiment SOP-I PACIFIC OCEAN ATLANTIC OCEAN 1 INDIAN OCEAN DATE 1979 o 1 a. O 1 a. o 1 a. a. in o ) o. p-30 p-30 Flop 1-^ CO 1 D. a. o u. m 1 a. 7 a. Q. o u. t a. 4- r-»Ji — ■ o 1 •a- 1 o I z (NJ 1 Z z 1 z 1 Z LD "a- z 01 s_ +-> o C Q. E en r- tjr, a. 0) o 15 Jan X ' 1 X 16 Jan 1 X X X 17 Jan X X X X 18 Jan X A , A 1 1 19 Jan 1 X X 1 i ^ 20 Jan i X ! X X i ! X 21 Jan 1 1 X X 1 1 X 22 Jan 1 1 X i X 23 Jan X X X 1 ^ 24 Jan 1 X X X X 1 X 25 Jan 1 X 1 X i 1 )( ! X 26 Jan I X X X 1 X 27 Jan 1 X X X X X 28 Jan X X 1 X i 1 X X i 29 Jan X ^ 1 X X 1 X X 30 Jan 1 X X 1 X X 1 31 Jan X X X 1 1 X X 1 Feb 1 X i X X 1 1 X X i 2 !^eb 1 X X X 1 X X ' 1X1 3 Feb X 1 X X ! 1 X X 1 X 4 Feb X 1 X X X 1 X X 1 5 Feb X 1 X X X X 5 Feb X i i X X X X 1 1 7 Feb X X X 1 X X 1 8 Feb X X i X 1 1 X X i 9 Feb 1 X X X X 1 X 10 Feb .X 1 X X X 1 X 1 U l-eb X ! X X 1 1 X 1 1 X 12 Feb 1 X X X X .. 1 X X 13 Feb 1 X 1 X X X X X 14 Feb 1 X X 1 1 X X 15 Feb X X X X X 1 X 16 Feb X X X X X 17 Feb X X X X X 18 Feb X y X 1 X _. .lX 19 Feb X X X X 1 X 20 Feb X X X 1 L _. XX TOTAL 1 2 11 15 34 31 2 2 2 2 8 3 15 1 1 12 11 3 16 3 OCEAN TOTAL 102 11 61 TO' rAL 174 133 Table 9. --Summary of aircraft dropwindsonde program Global Weather Experiment SOP-II DATE t LU O oo ■z. a. o LU 1-1 CO oo s. ^ o LU a: 3: UJ (-) CO z Z —1 Q LU I— OS o LU LU CO — 1 < X — 1 1— =3 O < —I a 1— <-> 1/1 t— z o AA = Air Abort AMD = Aircraft Mechanical Difficulty EP = Equipment Problems (ODUS) NF = No flight TM = Track modified 10 May C-141 AllE 18 16 8 10.3 11 May C-141 AllE 15 12 7 10.7 12 May C-141 AllE 18 17 7 10.3 13 May C-141 AllE 22 19 8 11.0 14 May C-141 AllE 14 13 7 9.6 EP 15 May C-141 — -- — — -- AA-AMD 16 May C-141 AlOE 19 18 9 10.8 17 May C-141 AlOE 20 19 10 10.0 18 May C-141 -- — — -- -- NF-AMD 19 May C-141 AllE 21 19 8 9.9 20 May C-141 AllE 20 17 8 10.7 21 May C-141 AllE 21 18 9 11.0 22 May C-141 AllE 22 20 9 10.6 23 May C-141 AllE 21 19 12 10.3 24 May C-141 AllE 16 15 1 9.S 25 May C-141 AllE 19 16 8 10.0 26 May C-141 AllE 21 20 13 10.2 27 May C-141 A13E 21 16 8 10.0 28 May C-141 A13E 19 16 8 9.5 29 May C-141 A13E 20 19 11 10.0 30 May C-141 A13E 16 15 8 9. -8 31 May C-141 A14E 21 17 9 10.0 1 June C-141 A14E 19 16 10 9.5 2 June C-141 A14E 21 18 10 10.0 3 June C-141 A14E 20 17 10 9.6 4 June C-141 A14E 22 18 n 10.0 5 June C-141 A14E 21 18 9 10.0 5 June C-141 A14E 22 21 9 10.0 7 Jun-e C-141 A14E 15 13 7 7.0 TM NOTE: 2 RR 7 June 7 June C-141 A14W 14 11 5 5.7 TM 8 June C-141 A14E 22 19 9 9.7 TOTAL 29 560 492 254 287.0 140 I O 00 CU O) Q. X s- O) -M fO CD jQ O CD OJ -M S- o n- u s- ■l-> CD ^- o s- 03 I I 0) J3 iviol Aiwa U3 lO in LO m ^3- ^ •a- ID U3 «3 ^T tn c X X X X X >c X X X X X CO Z LU O o o H- Z <: -J 1— 317 IV X X X X X X X X X en CM 3eLV X X X >; •«r 31LY >; X X X X X X X X X X X 3< ^ 301V X X CM Ml7lV X r— PACIFIC OCEAN dOlJ L£d X <- 5 dOlJ 0£d ■ X X CM 2Sd >; - L€d X X X X X X VO 0£d X X X X >s X X X X X X >S X X X X X X X x: X X X CM 2Zd X X X X X X X r>. I2d X X X X X X X X X X X X X X ■^ Old X >; X X X :x X X X X O Sid X - ILd __. X 1 Old X X X >s X X X X X X X X X X X m DATE 1979 o 1— (T3 C\J en I— «^ UfJ U3 r»» CO a> o CVi CM ^3- CM CM «3 CM CM 03 CM en CM o en en CM en ■T LO <£) 1 CO — 1 (— o OCEAN TOTAL 141 \ V-. )/ ^^^-. ^^"^ ""'■ — »^ .'-^ .-4 ,— ^ ^" — ^ rTi^ — Q. o t/JE^ . — xico capu OL-1 ^ * ■ ■ ^ r CM ,,,.. Q. 1 ffi V „;, ., V \ ^ U. -M O U Q. fO U u s- CD o CO o CM 142 / ^/| / w ^ 1 \' \ r^ CO (U -a c: o to -o Q. O S- o I C\J cu S- u CD 143 o Ql. i- ■M C a. o CD s- o 4- 00 o ns s- +-> CD O) o (/) C i Q. O i- Q I 1 CO CD -D c o 00 X3 2 Q. O S- I OJ C71 146 s- o Q. CD OO O 4-> to •4-> Ol •r- 0) Q. 00 I c ro (U o o c to i a. o 00 Ul o V) o S- 4-> CD e o to ■a c CL o s- Q I I U3 O) s- :3 cr> 147 o ■ O 3 ' 5 Q.. ■S IB < X o 7^ ^ - ^" £ ^ 1 M |\ 1 \ 1 s o CM CM Q. 1 CM 1 ^ Q.I Q. ';■. ' Q. o -J u. Q. 1 a 1 1 -1 IL Q. ^ 1 1 i 1 ^ 1 * y r #»*^ ,; ickam AFB OL-2 ^ 1 J ^^. - z ^ ^ \ ^ ^ ' „- o c or > CM Q o o o o CO o 4^ CD U o o CO o o in o o 1^ o GO uj o S- c o s- E 03 o o 03 O 3 Q <: CD o CO o CM 14R o o> o CO I cn I Q. C to ; ' 2. OBSERVATIONS PROVIDED BY THE NAVAID SYSTEM The FGGE NAVAID System is an integrated array of electronic and mechan- ical equipment that measures certain upper air thermodynamic parameters (temper- ature, humidity, barometric pressure) and utilizes the worldwide Omega Navigation System to determine wind speed and direction. The system has two major subsystems: (a) an expendable balloon-borne radiosonde launched from a ship, and (b) a deck unit to receive the signals telemetered from the sonde and to preprocess and record the data on magnetic tape. Data are recorded as the sonde rises from sea level to the stratosphere - a period of approximately two hours. During FGGE, the mag- netic tapes containing the NAVAID observations were forwarded for processing to a special center established by the Finnish Meteorological Institute. NAVAID: Following are the design specifications for parameters measured by Pressure +_ 1 mb over the range 1040 mb to 5 mb Temperature + 0.5°C over the range +50°C to -90°C Relative +_ 5% up to first tropopause or 300 mb, whichever Humidity is lower Wind j^ 2 m/sec in areas of good Omega reception (minimum of 5 vertical layers in the troposphere and 3 in the stratosphere). 3. SYSTEM DEVELOPMENT The Secretariat of the World Meteorological Organization (WMO), with contributions from the United Nations Environment Program and the governments of Saudi Arabia and the United States, undertook a coordinated procurement of the NAVAID system in March 1976 when appropriate industries around the world were invited to respond to a WMO technical solicitation. Industry proposals for 155 systems were analyzed and evaluated by an "international expert group" convened by the WMO in Geneva, Switzerland. This group selected VAISALA OY of Helsinki, Finland, as the major system contractor, with Tracor, Inc., of Austin, Texas, as the primary subcontractor to VAISALA. The procurement plan for the NAVAID system was designed to meet two objectives. The first objective was to equip 28 ships with NAVAID deck hardware, defined as modules A, C, D, and E. The modules were defined as follows: Module A. This is the data acquisition subsystem. It is composed of the receiving antenna and preamplifiers, an operator control panel, a sonde telemetry receiver, a microprocessor, and dual re- cording electronics for placing semi-processed data on digital magnetic tape cassettes for subsequent processing at a central location. Module C. This contains shipboard support equipment, including a power supply for Module A, a radiosonde balloon launcher, a helium gas flowmeter, and an air conditioner for Module D. Module D. This is the shipboard equipment shelter. It is a "knocked-down" prefabricated shelter for housing Modules A and C. This module was designed for ship deck mounting and engineered to withstand the anticipated marine environment of the Global Weather Experiment. Module E. Module E contains ancillary equipment (spare parts, tools, expendable supplies) necessary for the operator to maintain and operate the system at sea during the experiment. The second objective was to acquire a number of Modules B, the NAVAID data processing subsystem. This module, intended for use after the experiment, could be combined with the other NAVAID modules used in the experiment to form a "stand-alone" system which could be used to obtain meteorological soundings and to process the received signals, producing standard coded data on site. phases: To meet these objectives, the procurement was divided into three Phase I (February-October 1977) ' ■> Design, build, and test prototype NAVAID units A, C, D, and E. Phase II (March-November 1977) Design, build, and test a prototype Module B processing unit. Identifiy spare parts Test Modules A, C, D, and E at sea Retrofit tested Modules A, C, D, and E and mate them to Module B 156 Phase III (February 1978-June 1979) Manufacture a number of Modules A, C, D, and E for the experiment, and a number of processing units (Module B) for post-experiment use Train sounding operators Install NAVAID equipment aboard ship Develop the capability to maintain the equipment during the experi- ment. 4. NAVAID WINDFINDING METHOD The NAVAID sounding system made use of transmissions from the inter- national Omega navigation network (Figure 1) to determine winds. The pattern of radio waves transmitted by two Omega stations is illustrated by the concen- tric rings in Figure 2. The hyperbolas in the figure represent curves on which the phase difference between these two transmissions is constant. A set of hyperbolas remains stationary geographically and thus can be considered as a set of coordinates for geographic location. In order to define a location in this new coordinate system, another set of hyperbolas is required. This is achieved by reference to a third Omega station, as shown in Figure 3. Transmitter pair AB generates one set of hyper- bolic coordinates, pair AC the other set. The resulting coordinate grid is seen in the center portion of the figure. These phase measurements from two pairs of Omega stations allow the determination of the exact position of each point on a horizontal plane. The ascending NAVAID radiosonde receives and relays the Omega signals to the ship -- or (for later use) to the land-based observatory -- where the phases of the Omega signals are recorded for calculating the location of the sonde during its ascent. The successive positions of the sonde reveal the winds. The height of the radiosonde is computed from the three measured thermo- dynamic parameters. 5. SYSTEM DESIGN Figure 4 is a block diagram of the NAVAID Sounding System. The radio- sonde, which after preparation was attached to a balloon (Totex 650 g) and launched, was a VAISALA type RS2-1 with specially treated and selected meteoro- logical sensors. The sonde received all 13.6kHz signals from the worldwide Omega network and generated signals from its own temperature, humidity, and pressure sensors. Both the Omega and the thermodynamic data were transmitted to the ship via a specially designed, circular-polarized sonde antenna manu- factured by Synergetics International, Inc., Boulder, Colorado (403 MHz signal). The composite signal was received aboard ship with a high-performance receiving antenna, also manufactured by Synergetics. The antenna design was selected for optimum performance in the anticipated wind conditions (high elevation angles 157 • REUNION ^^^mmjim. ^V^ lAROENTlNA Figure 1 .--International Omega network. Figure 2.--Hyperbol ic curves generated by a station pair TRANSMITTER C TRANSMITTE TRANSMITTER B Figure 3. --Hyperbolic grid generated by three stations 158 403 MHz ANTENNA (CIRCULAR POLARIZATION! 403 MHz ANTENNAS (VERTICAL POLARIZATION! _A_ OMEGA WHIP ANTENNA ^ 7 \ 7\7\7\ 7 y MASTER OSCILLATOR OMEGA PREAMP — J! f y ANTEIMNA SELECTOR f 1 ^ TIME OF YEAR CLOCK TIME SINCE LAUNCH TAPE RECORDER NO. 1 REMOTE START y ' r UHF PREAMP AUDIO MONITOR y ' " ik i k OMEGA RECEIVER/ PROCESSOR UNIT A k 1 ^ ' 1 403 MHz DEMODULATED . MET DATA ^ RECEIVER \ OMEGA SIGNALS ^ r ' f 4 k i i i i i V i k i 1 ' AFC TUNING METER SIGNAL OMEGA AUDIO MONITOR TAPE RECORDER NO. 2 MONITOR SIGNAL LEVEL FLAG SONDE 1 D. ROSEMOUNT PRESSURE TRANSDUCER LATITUDE/ LONGITUDE AMBIENT PRESSURE 1 IF USED SYSTEM STATU VOLTAGE CONVERSION SYSTEM POWER AUTOMOTIVE BATTERIES ( IN USE ) AUTOMOTIVE BATTERIES (CHARGING I BATTERY CHARGING & SWITCHING 230/115 VAC TRANSFORMER SHIPS POWER 100 130 VAC OR 200-250 VAC 45 64 Hz Figure 4.--NAVAID sounding system block diagram, 159 for the sonde) in the tropics during the experiment. The antenna system was lightweight and small (1.5 kg and 1 m tall) and was quickly clamped to the ship's superstructure. After reception, the data signal was amplified by low-noise preampli- fiers adjacent to the antenna. The signals were then routed to the Module A electronics via prefabricated vapor-block-type cables. The cables were built for external service and were quickly attached to external ship structures. The signal cables entered Module D and were routed to the Module A electronics. The first element in the electronics was the telemetry receiver. The receiver had a signal-level monitor meter which displayed the received telemetry signal strength. This enabled the operator to tune the receiver to obtain the maximum sonde signal. The telemetry receiver had two demodulated outputs. One output contained the sonde thermodynamic data (pressure, temperature, humidity) in the form of square-wave signals within the frequency range of 46 to 52 kHz. The second output contained the 13.6-kHz Omega signals retransmitted from the radio- sonde. In addition to Omega signals received from the radiosonde, local Omega signals were received aboard the ship via a standard Omega ship antenna. These signals also were amplified and recorded by the Module A electronics. Local Omega signals were subsequently used in data processing. Omega signals from the telemetry receiver were routed to the NAVAID receiver processor unit (RPU). This receiver contained a microprocessor which was the heart of the electronics aboard ship. The program for the processor was stored in ROM (read-only memory) and organized in a modular form. Omega signals from the telemetry receiver entered the RPU and were synchronized auto- matically to the Omega transmission format. In addition, the RPU performed amplification, filtration, noise suppression, signal compression, phasing, and digitization of the sonde's Omega signal, which was comprised of the signals from all Omega stations within receiving range of the sonde. Meteorological (thermodynamic) signals from the telemetry receiver were also routed to the RPU for digitizing and formatting. The RPU compares frequency samples to establish and maintain synchronization with the sonde. In addition, the RPU filters and averages the meteorological data. The Omega and meteorological data were separately buffered in the RPU data memory. When- ever either buffer was filled with data {ewery 10 seconds for the Omega data and 6 seconds for the meteorological data), the buffers were read out to the Module A recorders. For data security, the NAVAID system recorded data in parallel on two Memodyne recorders using certified data-tape cassettes. 6. OPERATOR CONTROLS AND DISPLAYS Figure 5 depicts the NAVAID operator's control and display panel, part of Module A. The panel cues the operator through a launch sequence with the use of colored indicator lights. In addition, the panel provides thumbwheel switches which enabled the operator to enter meteorological surface data as well as an error-correcting sonde identification code. Specific sonde identification pro- vided the postprocessing center (the TWOS NAVAID Data Center of the Finnish Meteorological Institute, Helsinki, Finland) with the capability of reducing sonde bias errors. 160 n n PRESSURE 10 9 9 ' 50 LATITUDE N 1 3 o 3 2 ' LONGITUDE E 1 4 7 o 54 ' SONDE NUMBER 1 - 2 3 II 4 5 6 7 { 2 3 5 9 5 9 PRESSURE TIME FLIGHT TIME TAPE 1 TAPE 2 OMEGA NOT NOT NOT SYNCHRONIZED READY READY NO BATTERY TELEMETRY LOW DATE CARTON NUMBER TIME DATE SET SET RESTART LAUNCH IDLE READY TRACKING ON MANUAL NEXT START ENTRY ENTRY o ^ STANDBY Figure 5. --Operator control data entry and display panel 161 The NAVAID Sounding System had the ability to measure and record surface (shipboard) atmospheric pressure automatically. This was accomplished with a Rosemount pressure transducer. This measurement was made as a backup to the operator-entered pressure measurement, which was obtained from an aneroid barom- eter provided with the system and mounted within Module D. At the time of the radiosonde launch, the Module A electronics were activated and the recorders began to record when a "launch signal" was received from the radiosonde launcher located on deck. The launch signal was obtained from a "nail plate" switch attached to the sonde and routed to the RPU by a weatherproof signal cable. 6.1 Modules C, D, and E 6.1.1 Module C The components of Module C were the following: Power System. The power source for the NAVAID consisted of two banks of sealed lead-acid storage batteries, each containing two batteries. One of the battery banks supplied power for operation of all the electronics. During this time, the other bank was being charged. The charger, then, was the only device that needed to interface directly with the ship's power system. ' As such, it was designed to operate over a wide range of power frequencies and voltages. Also, because no electronics were connected to the charger, power-line noise and voltage or frequency fluctuations did not affect the system's operation. Battery capacity was such that the system would operate for roughly three launches before it needed recharging. Air conditioner. This operated directly from the ship's service power; it too operated on a wide range of ships' voltages and fre- quencies. Balloon inflation and release device. This was a new design for NAVAID. The device was a portable aluminum- framed canopy. Release of the balloon from the canopy was accomplished by a single-handed pull. The sonde, after being prepared for launch, was held in a special container attached to the release device. The sonde was held firmly in place in the container with a "nail plate" attach- ment. As the sonde was pulled away from the launcher, the nail plate rose a short distance with the sonde, triggering a balloon release switch which activated the start of the Module A recording system. Helium inflation gas regulator and flowmeter. During inflation of the balloon, the precise helium gas flow was monitored. When a prescribed amount of gas had been delivered to the balloon, an audible alarm and a flashing light were activated. 6.1.2 Module D The NAVAID Sounding System was installed on a great variety of ships for the Global Weather Experiment. Many of the ships were not specially out- fitted for environmental research. Consequently, the system was adaptable and 162 easily installed. The shelter, Module D, facilitated this. Figure 6 is a cutaway sketch showing the organization of the interior. The module was manufactured with a foam-core sandwich construction for high strength and low weight. It was shipped in a "knocked-down" config- uration and erected in approximately 2.5 hours with a minimum of tools. The total weight of the module, including the power and air conditioning systems, was approximately 300 kg. The unit had adjustable legs for leveling on a canted ship's deck and was constructed from materials able to withstand the anticipated marine environment. Within the module all the electronics were supported by shock-isolating equipment which reduced the shock and vibration transmitted to the module while the ship was under way. 6.1.3 Module E Module E contained the ancillary equipment necessary for the operator to maintain and operate the system at sea during the Global Weather Experiment. Many of the ships were at extended distances from their home bases and were on station for as long as 35 days. The design concept for the maintenance at sea program was to avoid having the operator repair anything within the electronic chassis. It was felt that the levels of training, test equipment, and spare parts required to repair the equipment at sea would necessitate an unaccept- ably large amount of training for the operator and place the system costs sub- stantially over budget. System reliability was created with the use of highly reliable components, independent processing, and substantial burn-in time for the units prior to shipment. Therefore, Module E contains only those items which a reasonably skilled operator would replace at sea with a minimum of sup- port equipment. 7. TESTING AND RESULTS The testing of the NAVAID was carried out during the period 15 November 1977 through 13 January 1978, when 34 soundings with both Module A and Module B were performed in Helsinki under the supervision of the Finnish Meteorological Institute. In six of these soundings the Helsinki airport radar (Selenia Meteor 200 M RMT-2S) was used as reference. The comparison between the various systems gave yery satisfactory results as shown below, where winds determined by the Finnish Meteorological Institute from Module A recordings, winds computed in real-time by Module B, and radar winds (using in all cases a four-minute smooth- ing scheme) are compared for six test soundings. The RMS differences are of the same order as the windfinding accuracy of the reference radar. TABLE 1 .--Root-mean-square differences in wind direction (dd, degrees) and wind speed (ff, m/s) for Module A / Radar and Module B / Radar based on six pairs of test soundings. MODULE A / RADAR MODULE B / RADAR dd(°) ff(m/s) dd(°) ff(m/s) 1.45 0.89 1.84 1.01 163 Figure 6. -NAVAID module D cu taway, 164 8. NAVAID WIND DATA QUALITY DURING FGGE The most important atmospheric feature to be defined by the NAVAID is the vertical wind profile. Assessing the NAVAID wind accuracy is not straightforward, because it varies with time and geographical location of the sounding, is dependent on the software used for the Omega data reduction, and is directly related to the quality of the Onega signals available. How- ever, by monitoring the behavior of the signals one can estimate the accuracy of the wind profile. In the data processing it is possible to vary the signal integration time so that a desired wind accuracy will be achieved. If the quality of Omega signals is good, an integration time of one minute is enough to provide accuracy of, say, 2 m/s. In poorer Omega conditions, it would be necessary to increase the integration time to attain the same accuracy. Each selected integration time corresponds of course to a certain vertical layer through which the balloon ascends during the integration time. There is thus the opportunity for a trade-off between accuracy and vertical resolution. Table 2 below, which was prepared at the TWOS NAVAID Data Center in Helsinski, shows the combined statistics of the NAVAID wind accuracy and the corresponding vertical resolution for all NAVAID ships during the two FGGE Special Observing Periods. From the table it can be concluded Table 2. — Root-mean square wind vector error for various vertical integration layers according to Lanqe [ 1 ] Estimated RMS- ■error of vector wind less than (mps) Total number of 0.5 1.6 3.3 5.5 8.1 11.1 14.5 18 .3 22.4 soundings E 100 1 1 c 330 17 68 1 86 670 162 530 5 697 to 1100 234 1641 21 1 1897 03 > 1620 65 526 7 598 (V ■M C 2227 22 184 4 210 (J 2909 8 51 1 60 3666 1 17 5 3 1 27 > 4495 2 1 3 6 Total number of 509 3019 46 4 1 3 3582 soundings 165 500 1000 1500 2000 2500 Metres 3000 3500 4000 4500 5000 Figure 7. --Frequency distribution of vertical integration layer for NAVAID data given in form of percentage of soundings having integration thickness less than the value indicated on the abscissa. UV: 10 20 30 40 50 60 70 80 90 100 Figure 8, --Frequency distribution of the pressure altitudes attained in the form of the percentage of soundings exceeding given pressure altitudes. 166 that 98.5% of the soundings had RMS-error of vector wind less than 1.6 m/s dis- regarding the integration interval. The frequency distribution of the integra- tion thickness for the data having wind accuracy better than 1.6 m/s is given in Figure 7 which shows that in three quarters of the soundings the integration interval was 1100 metres or less. In 90% of the cases the integration interval was less than 1600 metres. Please note that in the shipboard NAVAID system, the relation between time and layer thickness is approximately linear (1 minute equals about 300 meters). In contrast, the parachute-borne dropwindsonde shows a more linear relationship between time and pressure difference (1 minute equals about 25 millibars). Figure 8 shows the frequency distribution of the pressure-altitudes attained in the form of the percentage of soundings exceeding given pressure altitudes. About 90% of the soundings reached 150 mb or higher whilst two thirds penetrated beyond the 50-mb level. 9. PERFORMANCE OF U.S. TWOS During FGGE, nine U.S. ships were equipped with NAVAID. Of these nine ships, six were operational during each SOP. During the first SOP, systems were installed on the RESEARCHER, DISCOVERER, GYRE, TOWNSEND CROMWELL, DAVID STARR JORDAN, and the WILKES. The DISCOVERER, after 19 days of successful operation during January, experienced a failure in the NAVAID system electronics. As a result, they were unable to make soundings in February. The WILKES experienced battery problems early in the first SOP and was able to make only one sounding per day in January and after five days of operation in February, the system failed completely. The RESEARCHER and GYRE were able to maintain their sounding schedule; however, some soundings failed to reach minimum altitude due to prob- lems with the balloon inflation system. The following NAVAID system shifts were made for the second SOP: FROM TO DISCOVERER OCEANOGRAPHER DAVID STARR JORDAN COLUMBUS ISELIN TOWNSEND CROMWELL KNORR No major NAVAID system failures occurred during the second SOP; hov/ever, some observations were lost due to changes in ship schedule and operational require- ments. The Operations Plan for U.S. TWOS scheduled approximately 370 total at-sea ship days between 10°N and 10°S latitude. Hence, at a rate of two soundings per day per ship, approximately 740 soundings should have been made. As shown in Table 3, U.S. ships made a total for both periods of 653 soundings or almost 90% of the planned total number. The DISCOVERER and WILKES had prob- lems with their NAVAID systems, while the GYRE and OCEANOGRAPHER had problems maintaining their schedule. 167 The following table summarizes the NAVAID soundings taken by U.S. ships during FGGE. Table 3.— Soundings made by U.S. TWOS SOP I SHIP WILKES RESEARCHER JORDAN CROMWELL GYRE DISCOVERER LOCATION (OCEAN) ^ Indian Atlantic Pacific Pacific Pacific Pacific SOUNDINGS COMPLETED ( INCLUDING SUCCESSFULLY PROCESSED DATA) 22 55 71 89 31 18 TOTAL 367 SOP II RESEARCHER Indian COLUMBUS ISELIN Indian KNORR Pacific OCEANOGRAPHER Pacific GYRE Pacific WILKES Indian TOTAL 69 50 26 120 79 23 286 168 While U.S. ships did not accomplish 100% of their planned soundings, they did provide approximately 1/5 of the total soundings (3,582) made by all ships equipped with the WMO FGGE NAVAID systems. 10. LOGISTICS NAVAID, in order to operate successfully, had to be installed on re- search ships all over the world. This installation was compressed into a three- month time frame prior to the start of SOP-1 because of equipment manufacturing lead times and research ship availability for installation. It was felt at the beginning of the program that both the logistics effort (shipping, scheduling, clearing through customs, etc.) and the installation effort would be the most difficult part of the program and be fraught with unforeseen problems. The concepts for logistics were reasonably simple and straightforward. The following are the highlights of the plan: A detailed plan for logistics was developed which pinpointed in- stallation ports, ship schedules, and installation dates. All NAVAID equipment was to be shipped on a worldwide basis by a single shipping agent. The agent was selected by competitive procurement against a specification of performance and price. All NAVAID equipment would be installed by an experienced, well rehearsed installation team(s). The teams would be flexible, totally familiar with shipping and customs requirements, multi- lingual, and have total responsibility for the installations when in the field. A central coordinating center would be set up at WMO Headquarters in Geneva, Switzerland. This center would be aware of up-to-the- minute changes in any program requirement, shipping schedule, or ship availability and be able to quickly communicate with the installation teams or vice versa on a 24-hour-per-day, 7-day--per- week basis. Figure 9 is an overview of the NAVAID worldwide installation schedule for SOP-I. It is broken down into geographic teams and specific installation times for each of the NAVAID ships. NOAA recognized the importance of the logistics effort for the FGGE NAVAID and seconded a logistics specialist to the WMO for the duration of the program. He was involved in the generation of the logistics plan, the selection of the shipping contractor and served as leader on one of the installation teams. 169 ■S %:»-::::; c •?? k II •X'co :•:•: c (0 ■o If ID ra |S^ o -> « •v. CO ,v.' « isi 8^ •D >« i n •:%c>4 ¥: Pi CD C « CD >« 1 q) Q. CO k 5 s III 3 3 O C O is! 1-9 c (0 ■o 2 E "3 CO CO » Q. CD z Z u (D ■5 C •'•'•••••••••V 0> ::S (0 '••Xv«*t%*t' 3 J£ i^-in I;:;; id > O -I C n c (0 -J '•••••••••••••V CL a O 3 - >. if! 3 Si (0 3 a •:•:•& :-i^: a c E ■o :':-:^- ■:•:' . s ■:•:•: *^ •:•:• o ■ (0 i-^i-ct >::; £ S I'X'X'X'I*! CD n :*:-:eo ¥:' J3 c >::::«:' % E - CD S ^o»g ? ~ !>; c !'l"I*I%'lvi IS E o Si® $5 4> c >X • 'X o oc > O CD o X'lO 'X i'X'" 'X' St z k. a> c c l.r E (0 e oo |i«|j o « oc z i « *•••••••••••%• oc CO :¥ o ¥: :¥ CO cr> C7) So. (0 CO |E M UJ S CD 170 In addition NOAA provided additional personnel to other installation teams and supported the overall logistics effort. A program like NAVAID succeeded logistically only because extremely dedicated people, working with. a well-organized plan, make the shipping arrange- ments, customs formal ities, and ship installations happen. The highest praise must be given to the NAVAID contractor VAISALA OY of Helsinki, Finland, and especially to their dedicated field service engineers that worked under extreme conditions to make the installations successfully. 11. PROBLEMS AND SOLUTIONS After evaluation of the NAVAID data, discussions with the NAVAID oper- ators and evaluation of the ship reports, the following were judged to be the major system problems. Balloon inflation volume difficult to accurately measure with the NAVAID flow meter. Balloon launcher too fragile and balloon tie-off too long and com- plicated a procedure. It was possible to overcharge system batteries, and with heavy system use, the charging system did not maintain battery charge. Unexplained "jumps" in recorded signal data were observed on some of the signal records. 12. PROPOSED SOLUTIONS Redesign the balloon inflation system based on an automatic volu- metric measurement rather than an operator-observed digital flow- meter system. Redesign balloon launcher to be simpler to assemble and made from stronger materials, even with a sacrifice in weight. Redesign battery charging system for higher charge capacity and complete shut-off when the NAVAID system is inactive for long amounts of time. Data jumps appear to be caused by the operator retuning the re- ceiver at too rapid a rate. Additional operator training may be the only reasonable cure for this problem. 13. CONCLUSIONS The NAVAID system contributed approximately 3500 soundings of the Equatorial Tropical Atmosphere during the FGGE Special Observing Periods. The system performed, for a high percentage of cases, within the design requirements. The system was deployed on a worldwide basis and installed on approximately 30 ships. The delivery and installation dates were accomplished according to the required schedule and total system was produced within budget. 171 CHAPTER 9 AIRCRAFT TO SATELLITE DATA RELAY (ASDAR) By James K. Sparkman, Jr. (NOAA) James Giraytys (NOAA) George J. Smidt (NOAA) 1. INTRODUCTION Wide-bodied jet aircraft were used as data collection platforms for in-flight meteorological reports for research projects, almost from the moment B-747s were introduced as commercial carriers in 1970. Data were collected on board using the airplanes' own sensors for wind and temperature. For many years, meteorologists had urged the development of a fully automated system for data collection from aircraft in flight. An automatic system would be free of the human errors found in voice-radioed aircraft reports, and of the prohibitive manpower costs for collection at major terminals of hard-copy reports made by aircrews during flights. Initially, a tape system using hardware already installed aboard many aircraft seemed a low-cost approach to an improved operational data collection system. However, removal and carrying of tapes from arriving aircraft to processing centers for transcription proved to have many of the costs and delays experienced in collection of hard-copy reports. The delays proved unacceptable. As preparations for the First GARP Global Experiment (FGGE) took shape, it became evident to planners that all the elements needed for a worldwide auto- mated data collection system using aircraft were in place. An automated data collection system for aircraft reports was at last possible. Collection system elements included: (1) Quality sensors for winds and temperature installed aboard air- craft; (2) Digital processing of sensor-signals aboard aircraft, yielding digital values in engineering units for winds, temperature, altitude, and air- craft location; (3) Satellites, with a proven capability to relay reports (from ground-located "Data Collection Platforms" (DCPs)) to ground receiving stations; (4) Established communications between satellite, ground receiving stations, and meteorological data processing centers; (5) Worldwide circuits for meteorological data, called the Global Telecommunications System, through which aircraft reports can be sent to data users as required. In January of 1975, NASA and NOAA agreed on a plan to develop ASDAR as a possible operational data system and for an initial test of the system during FGGE. Key events in the ASDAR development are listed chronologically in Table 1 at the end of this chapter. The goal of the ASDAR program was to develop an electronic package for installation aboard aircraft which simultaneously (a) could collect meteoro- logical reports, formulate them into an appropriate message, and radio the message to relay satellites, and (b) would be acceptable to airlines and air- safety regulatory agencies. There were a number of technological uncertainties 175 to face to achieve this goal. What the ASDAR program provided for FGGE and some of the difficulties that had to be overcome to meet the above goal are outlined in the sections which follow. 2. PROGRAM DEVELOPMENT When NASA's Lewis Research Center began to work jointly with NOAA to develop ASDAR, it was clear that data could be collected from aircraft sources and that digital messages from such avionics units as Inertial Navigation Systems (winds and location) and Flight Data Acquisition Units (temperature and altitude) could be selected and re-assembled into suitable messages, and stored until a specified time for transmission to a satellite. The main area of uncertainty was the radio link between the aircraft and the geostationary relay satellites, located 22,000 miles overhead. Relay satellites are operated by the U.S., Japan^and the European Space Agency. By their overlapping fields of view, they provide virtual global relay coverage up to 80° latitudes. NASA's engineers made a comparison between the needs for an ASDAR radio system, and the hardware used for tiny DCPs to report rainfall rates, river heights, snow depths, and many other environmental parameters. DCPs, they noted, require no more power than a handi-talkie (6 watts), but are given a high-gain narrow beam antenna that points precisely at its relay satel- lite. Since a five- foot-long helical antenna cannot be mounted on top of a B-747, the challenge for ASDAR was to build a large transmitter to feed a small, flat, low-gain antenna mounted flush atop an aircraft, radiating its signals in all directions upward. The need for aiming antennas would be obviated. The 80-watt transmitter developed by a NASA contractor for the program, operating at 402 MHz, represented state-of-the-art technology. After three years of deployment, the transmitter has a mean-time-between- failure of six months. This failure rate is still too great for an operational program. However, be- tween failures, the transmitter provides an unusually "clean" signal (free of unwanted harmonics and other off-frequency components) that delivers ASDAR messages to the satellites at appropriate signal levels. Drift of the trans- mitter frequency has not been brought wholly within suitable bounds, largely because of the uncertainty as to how many hours each day a given ASDAR unit will be provided with pov/er. (Transmitters generally drift-* that is. change frequency-r to lower frequencies when off power, but to higher frequencies while operating. Current drift rates carried most units out of band within 10 months.) The ASDAR antenna, while not a new design, was new to its application. The working part of the antenna, a block of plastic with embedded wires, (8 inches square by 1/2 inch thick) was installed into a machined-aluminum "picture frame" milled to fit flush on the cylindrical top of a B-747. Occasional failures (five of seventeen failed in two years) resulted from seepage of water into the plastic, driven by the force of flight- speed winds. Given the overwhelming success of the satellite DCP program (platforms now number into the thousands), it was not likely that the ASDAR antenna system would totally fail. However, it was considered possible that the ASDAR unit 176 might operate successfully only when near a satellite's sub-point. Originally, NASA promised data relay only to distances no more than 45° away from beneath the satellite, in any direction. It is to the credit of the Lewis engineers that they evolved a design which did not overload the satellite (with too strong a signal) when aircraft were just below the satellite, yet provided a sufficiently strong signal to relay messages until aircraft disappeared from sight of the geostationary satel- lite. Reports from flights to or across the Earth's poles were generally relayed to about 82° North or South. 3. DATA PROVIDED BY ASDAR The ASDAR system obtains winds from the host aircraft's Inertial Navi- gation System (INS). In an INS, the calculation of wind values is part of the operation necessary to guide an aircraft on an "automatic pilot" course compen- sating for cross winds. "Wind" is the vector remaining, after aircraft heading is compared with the actual course followed. ASDAR temperatures are obtained in digital form from the Flight Data Acquisition Unit (FDAU). Temperatures are sensed by a platinum wire probe mounted outside the fuselage near the aircraft's nose. The electrical resist- ance of the probe (which varies with temperature) is sensed on B-747s by the Central Air Data Computer, which converts resistance- value to an analog value of temperature, while also compensating for the heating-effect of rapid motion of the probe through the atmosphere. Output of the CADC is digitized by the FDAU. This value is then ready for digital recording or for use by the ASDAR unit. As designed by NASA engineers, the FGGE ASDAR unit can be set for data collection rates varying between eight samples per hour (112 km apart), and 128 samples per hour (7.0 km between data points). For deployment in FGGE, all units were set for the slowest data collection rate, except the Air Force C-141. Its faster data collection rate (16 samples per hour), permitted tests related to flight tracking, and possible use of an ASDAR-type system to speed aircraft location in event of aircraft ditching at sea. The normal FGGE operating mode was to take eight samples per hour and transmit these as a group once per hour. 4. OPERATIONAL DEPLOYMENT The first ASDAR test flight, on February 4, 1977, and Pan Am's routine flights in the weeks that followed demonstrated the success of the ASDAR design as a data collection platform. What remained, in addition to providing data for FGGE, was to show the success of ASDAR as a data collection system worth its cost in the world of operational meteorology. This latter step entailed flights of ASDAR units in many test environments, in sufficient quantities to permit an evaluation of ASDAR as a source of upper air data from remote regions not other- wise reported. Toward this goal, the focus of the program turned to increasing the number of operational ASDAR units (eventually 17), and deploying them in as many different operational situations as possible. Figure 1 shows the number 177 0} ex: Q C o +J 03 S- Q} Q. O I CD :uo!jej9do uj ubjojiv 178 of operating ASDAR equipped aircraft as a function of time during the FGGE Operational Year. The ultimate FGGE ASDAR fleet proved to have remarkable diversity. Home bases ranged from Copenhagen to Sydney (Australia). Typical daily routes (Figure 2) traversed almost every major ocean and continent except South America and its adjacent oceans, and some 30,000 hourly ASDAR reports were logged, yielding 240,000 data points to the FGGE data base. During FGGE the ASDAR data were available around the world via the Global Telecommunications System (see Figure 3). ' 5. PROBLEMS ENCOUNTERED 5.1 FGGE ASDAR aircraft, while all B-747s, were found to have large differ- ences in the kinds of avionics to which ASDAR would interface. Indeed, in two instances a "standard" interface for ASDAR had to be installed in addition to the aircraft's own processor-boxes for altitude and temperature, in order to permit reporting of these parameters by ASDAR. 5.2 A finite budget for ASDAR during FGGE led to some compromises. No changes were permitted to be made in the flight hardware without the approval of FAA safety inspectors. In some instances proposed modifications were dropped, because they could be approved only after additional flight tests to demonstrate that the package remained safe for commercial use. For a B-747, flight test time cost $14,000/ hour. Other surprises were related to the differences found within B-747s, and between B-747s and DC-lOs. The latter might have been accommodated by changes made within ASDAR units. However, with those changes, new flight tests for safety would have been required. These proved too costly for the FGGE budget for ASDAR. DC-lOs were not used, as a result. 5.3 Deployment on foreign airlines showed the need for extensive documen- tation of the initial tests for the ASDAR units. In many instances, safety tests were repeated in other countries, to supplement information supplied from the U.S. ASDAR manufacturer. 5.4 Shipment of failed units back to NASA for repair proved far more costly and slower than was anticipated. This lesson tends to confirm an early suspicion that repair depots located at several spots around the globe might prove cost effective in a test program such as ASDAR. For an operational pro- gram, airlines would themselves undertake many repairs, and would subsidize shipping of failed units. The costs to local weather services and to NOAA and NASA would then remain far less than was true in the FGGE deployment. 6. SUMMARY 6.1 The ASDAR program was a good example of the development of useful hardware for environmental sensing, carried out under the umbrella of a large research program. ASDAR' s contribution to FGGE, despite the program's modest size, was extremely important. ASDAR and its counterpart. Aircraft Integrated Data System (AIDS), which utilize on-board tape recordings of meteorological reports, were major suppliers of new upper air data. They played a key role in filling data gaps in remote areas. 179 6.2 By the end of the Global Weather Experiment, endorsements for ASDAR were proffered by three operational forecasting centers within the U.S., and by several major weather service centers abroad. However, the full potential of ASDAR messages, from an operational ASDAR fleet, can be realized only with modifications to the present system for delivery of weather products to avia- tion. More timely analyses are needed, at daily times keyed to the needs of the aviation industry. 6.3 While ASDAR reports are intended primarily for meteorological uses, they also proved of interest to planners for air safety programs. ASDAR reports permit flight-following, and, in event of an over-water accident, would more closely pinpoint areas for rescue searches than is possible with the current mandatory voice-radioed position reports. It appears likely that ASDAR will lead to more automated position reporting from aircraft, whether or not the ASDAR system is continued. . . 6.4 Now that FGGE operations have terminated, ASDAR remains. The system has a potential for fuel savings to airlines of perhaps $100 million a year or more, if fully exploited for improved flight level forecasts and aircraft flight track planning. Through FGGE, ASDAR has answered the pleas of fore- casters, heard for many years, that commercial aviation provide data from regions where costs or location will not permit the construction of additional surface weather stations. 180 Table 1.--ASDAR development: Chronology 1975 (January) - NASA and NOAA agree on a plan to develop ASDAR as an operational data system, using satellite data relay, for initial test in FGGE (May) - NASA's Lewis Research Center, Cleveland, begins ASDAR develop- ment program. Staff, approximately 15 persons (September) - First ASDAR specifications proposed by Lewis Research Center 1976 (March) - Pre-FGGE testing of ASDAR to be scheduled through U.S. Domestic GOES Channel 91, pending operational status of "International Data Collection System" channels (September) - NOAA-NASA team solicits participation from international air- lines in ASDAR program (October) - Lewis Research Center begins C-47 tests of ASDAR unit (November) - NMC begins programming for use of ASDAR transmissions 1977 (January) - ASDAR transmissions from Pan Am's "Clipper Arctic" successfully received by satellite while on the ground at JFK Airport, New York (February) - Australia and Sweden agree to carry ASDARs; Australia decides to buy several (March) - Initial antenna design proves unserviceable. Manufacturer begins again (April) - ASDAR broadcasts from PAA heard from beyond the geometric horizon (April) - Shift of ASDAR units to International DCS Channel 17 (= U.S. Channel 234) begins (June) - ASDAR messages sent out over the GTS (September) - Contract signed for commercial fabrication of additional ASDAR units 181 November) 978 [January) 'January) 'February) - [March) [March) [July) [August) [October) 1978 [December 1 ) [December 4) [December 10) 1979 [January 12) [January 19) [February 26) [March 13) [March 14) [March 17) [April 8) [April 14) [May 5) ;Mary 24) [June 22) - ASDAR tested via International DCS Channel 17 ASDAR now reporting via two satellites, GOES East and West First installation kit completed ASDAR unit shipped to QANTAS KLM's first flight, JFK to Amsterdam, tracked to 3°29'N, 50°31.7'E ESA reports first ASDAR relay ASDAR operational aboard USAF C-141 SAS's ASDAR installed, operational First relay of ASDAR reports by Japanese satellite FGGE begins ;., USAF- (I I) installed ,.,^^., ,, QANTAS (QF7) installed QANTAS (8) installed QANTAS (10) installed ,., ^ ,, , Lufthansa ASDAR installed - "■ ' ^ Singapore (13) installed by NASA team Singapore (14) installed ., ;.,..■. . Singapore (15) installed .. :.r . QANTAS (9) installed "■ : -^'^ British Airways (16) installed South Africa (18) installed British Airways (17) installed South Africa (19) installed ■; -if.' .' tV:if*,„ > AO:; ; '''vjinvv,/ .;.i i^; .H\i>l'' 182 1 183 ASDAR Aircraft GOES West 135W GOES East 75W Wallops I. GCA World Wx BIdg. Marlow Hts., MD. \ PLAR-1 DCP data outlet NMC KAWN Carswell AFGWC Offutt Meteosat OW MIchelstadt Darmstadt DWO Offenbach Service FNWC Monterrey Sunflower 140E Met. Sat. Cntr. Tokyo JMA, Tokyo Satellite RFIink Airlines Ground Station Processing Center GTS link to worldwide users Figure 3.— ASDAR data flow, during the FGGE. 184 CHAPTER 10 AUGMENTATION OF THE WORLD WEATHER WATCH (U.S. PARTICIPATION) By Terry Bryan (NOAA) a 1 . INTRODUCTION The operational surface and space-based components of the existing World Weather Watch (WWW) Global Observing System (GOS) were the primary source of observational data for FGGE - The Global Weather Experiment. However, there were serious data gaps in this network, particularly in the tropics and Southern Hemis- phere where many stations were not implemented or were conducting only partial programs and where vast oceanic areas exist. The Congress of the World Meteoro- logical Organization (WMO) appealed to Members to maximize implementation of WWW stations in these areas prior to FGGE, and the Intergovernmental Panel on FGGE recommended that those stations for which permanent installations were not planned or were difficult to implement should, as a minimum, be considered for temporary installation during the Special Observing Periods (SOPs). Several of these sta- tions were identified as "indispensable" to FGGE. ■ ■' ' In response to these requests for the improvement of the surface-based WWW, the United States temporarily implemented five new stations, and augmented seven existing stations in U.S. territorial areas and at locations operated by the U.S., and provided assistance to 30 upper air stations through the Voluntary Assistance Program (VAP). 2. STATIONS AND OBSERVING SCHEDULE 2.1 Temporary Implementation of New Stations Rawinsonde observing stations were temporarily implemented for FGGE - The Global Weather Experiment at the locations listed below. These stations operated from 5 January to 5 March 1979 (SOP-I), and from 1 May to 30 June 1979 (SOP-II). Rawinsonde and surface synoptic observations were scheduled each day at OOOOZ and 1200Z. Station Index Number Latitude Longitude Type of Equipment Enewetak 91250 • 11 21N 162 21E AN/GMD-1 Woleai 91317 07 23N 143 55E AN/GMD-2 Kapingamarangi 91434 01 05N 154 46E AN/GMD-2 *Fanning 91487 03 54N 159 23W AN/GMD-1 Canton 91700 02 46S 171 43W AN/GMD-1 *Fanning had been temporarily implemented for NORPAX on a one per day, six day per week schedule. It continued temporary operations for FGGE with augmentation to two per day, seven days per week during the SOPs. 2.2 Schedule Augmentation of Existing Stations The following existing upper air stations augmented their schedule of observations to provide two rawinsonde observations per day for the Special Observing Periods (SOP) indicated above. 187 ■' station Ind ex Number Latitude Longitude Type of Equipment Ascension 61902 07 58S 14 24W AN/GMD-4 Diego Garcia 61967 07 21S 72 29E AN/GMD-1 Truk 91334 07 28N 151 51E AN/GMD-1 Ponape 91348 06 58N 158 13E AN/GMD-1 Majuro 91376 07 05N 171 23E AN/GMD-1 I Koror 91408 07 20N 134 29E AN/GMD-1 Yap 91413 09 29N 138 05E AN/GMD-1 3. OBSERVING SYSTEM 3.1 Descript ion « ■■( ' * ■ The observing systems for these locations consisted of AN/GMD rawin- sonde systems which include a balloon-borne radiosonde (1680 MHz), rawin set AN/GMD, and a radiosonde recorder. By use of appropriate evaluation methods, the observer computed values of pressure, temperature, relative humidity, and wind speed and direction. 3.2 Expected Accuracy The present standard of accuracy for the AN/GMD ( ) rawinsonde system is: Temperature: j^ 1°C Relative Humidity: +_ 5% within the temperature range of +40°C to -40°C and a relative humidity range of 10% to 100% Pressure: +_ 2mb within a range of 1050mb to 5mb Wind Direction: +_ 5° Wind Speed: (a) AN/GMD-1: Mean wind vector from surface {+) Altitude 5-30 Kts 10,000 ft +2 kts 20,000 ft 3 kts 40,000 ft 4 kts 60,000 ft 6 kts 80,000 ft 8 kts 100,000 ft 10 kts - * beyond capability of equipment (b) AN/GMD-2-4: + 5 kts 4. OBSERVATIONAL ACCOMPLISHMENTS , ■ The temporary and augmented stations listed in the previous section made 908 extra upper air observations for FGGE during SOP-1 and 1022 extra 188 30-60 Kts 5 kts 7 kts 14 kts 21 kts 28 kts 60 -90 Kts 10 kts 15 kts 30 kts * • * observations during SOP-II. Planned and accomplished observations are listed for individual stations in Table 1. It should be noted that these observations were made under difficult conditions in extremely remote locations. The overall figure of successfully making 92% of all planned observations is an impressive testimony to the hard work and professionalism exhibited by the involved personnel. Examples of specific problems encountered are: On Enewetak only two observations were made during SOP-I before Typhoon Alice struck, knocking out power. The GMD main cable was severed in three places by flying debris. Eleven observations were missed before repairs were made and operations resumed on 11 January. On Canton Island, SOP-I observations commenced three days late due to last minute site and equipment problems. An additional fifteen observa- tions were missed during the remainder of the SOP due primarily to flight and/or ground equipment failure caused by the torrential rains which occurred almost continuously. Woleai SOP-I operations were terminated on February 4 due to the failure of two generators and the absence of logistical replacement/repair sup- port due to mechanical problems on the re-supply ship. 4.1 U.S. VAP Support to WWW Upper Air Network New windfinding radars (equipment, installation, training,, and expend- ables). 82022 Boa Vista, Brazil (in cooperation with VAP-F) - 82825 Porto Velho, Brazil 84008 *San Cristobal (Galapagos), Ecuador 64500 *Libreville, Gabon 97723 Ambon, Indonesia (in cooperation with VAP-F, France) 97900 Saumlaki, Indonesia 65660 Roberts Field, Liberia 61415 *Nouadhibou, Mauritania 78741 Managua, Nicaragua 98753 Davao, Philippines 63832 *Tabora, Tanzania 65503 *Ouagadougou, Upper Volta 43395 Male, Maldives 91800 Penrhyn, Cook Islands 94910 Douala, Cameroon Radiosonde/radiowind equipment (equipment, installation, training and expendables) . ^. . 78073 Nassau, Bahamas ■ 80222 *Bogota, Colombia , 189 o O) •r-J to O c s- O Q- •r- +J UJ fO o > o S- U- &- . Q.JD a. ZD T3 O) r— -o 03 c C 3 O LL. •r— 4-> -O C -o rc •o < S- o H- M- O ■a >, r*v. cTi coloo «^ t^ CO CO CM r^ CM o CM I— CM CM CO CM CM CM «v^ «vi- "^ «* CM CM CM CM CM CO CO CO O O O O O O O O O O CM CM CM CM CM O O cr> CM ko en cr>|cr> ^ I— CO CM CTi en CT>| en tn ID 1— o o o o o o o o o o o o o o o VO UD eo ^£} vo o o o o o VO ^ *^ ^ VO LO VOn- r>. CM eo CO 00 O LD CM r-. sf LD en ^ O r->. (£> CO t— CVJ fO CO 1 — r— O i- '^— o p>* VD CO CM LD CO .- r— O . LD CM ^— O CO r— r— en -o a; 4-> (0 , 1 ■ . . c •1— E o 00 o 00 o 00 s- o h*. vD CO CM CO a; CO r— ^— 1 o +j ' •- 4-> «* OJ CO v^ • '■ s- •vt- P>. (U ^— vo Q. en en CM 1— O O -r- o yD 1— < LO CD en r-^ T- CD c en CD o ^^ in C O) fl3 (U O) fO c o 1 — a. s_ s- cn 1 — c o 1 — •r- O) 4J _J r— S 3 c •>-> s- a. 4-> u ^ o t— 1— Q. ^ :i^ >- H- • National Climatic Center Federal Building ~« ;. Asheville, North Carolina 28801 USA The catalogue provides the scientific community with a list of FGGE data sets that are available at WDC-A and suggests a standard procedure to expe- dite acquisition of the data. The catalogue is designed in a loose leaf format so that supplements can be easily inserted to keep the catalogue current as the data are archived. The supplements are mailed at approxi- mately three-month intervals. 203 3. CONCLUSIONS AND RECOMMENDATIONS 3.1 Magnetic Tape Specifications and Formats Magnetic tapes produced for international exchange for the Global Weather Experiment were required to be built in accordance with tape record- ing characteristics and formats approved by the WMO Commission for Basic Systems. For observational data the magnetic tapes were to be recorded at 800 BPI density, the tapes were to be 9-track tapes, and the recording code was EBCDIC. The data were to be placed on the tapes according to prescribed formats for data recording. Analyses data were to be placed on 800 BPI 9- track tapes, and recorded in binary code, according to prescribed formats. The choice of 800 BPI density was dictated by political considera- tions (some countries could not read higher density tapes). Recording on magnetic tapes at 1600 BPI density is superior to 800 BPI density for economy of storage and reliability of recording purposes. Bilateral arrangements to use 1600 BPI tapes were made by some of the shipping and receiving countries. The recording density adopted as standard for tape exchanges should be kept under review to ensure optimum reliability and efficiency. The formats used for data recording did not include specification of the originating data center in the data reports. This caused some diffi- culty later in dealing with the satellite data, when there was a problem with data from one satellite data producer and the data from all producers had been merged. A similar problem occurred with regard to area sub-center data, but here no indication of the method of data collection was stored. Indications of both the originating data center and the method of data col- lection should form part of the format used for storing data. 3.2 Checking and Correcting Data Planning for FGGE provided that each data producer had primary responsibility for ensuring that its data products were stored on magnetic tapes according to agreed upon formats, and that the data have quality con- trol indicators attached. Some of the data were shipped by data producers to data centers without sufficient checking of the data and formats to en- sure that the exacting standards required for international exchange were met. Data producers and centers should always perform detailed evalua- tions of their own products according to a predetermined plan, before sub- mitting them to another center for evaluation. However, in spite of considerable efforts by data processing cen- ters to develop perfect software, it is inevitable that occasional small errors of content or format will occur. Normally, the originating center should be asked to correct its software and reprocess the data set con- taining the errors. However, in certain cases reprocessing and redelivery of data sets by originating centers may cause considerable delays in the data production. Therefore, receiving centers should plan sufficient re- sources to handle and correct for such minor errors. 204 FGGE planning required that complete replacement tapes be sent by data producers to data receivers whenever there were errors in the data content or format that had to be corrected. This was true even though in many cases there were only a small number of errors on the tape, and they may have been contained in the same file on the tape. For greater economy and efficiency in data transfers, receiving centers should allow correc- tions of errors to be sent in the form of single corrected files, rather than completely rewritten tapes. Another problem which occurred in the many FGGE data centers concerned the elimination of duplicate and near-duplicate reports. Al- though duplicate and near-duplicate reports were supposed to be eliminated from the FGGE data collection, efforts to do so were met with many diffi- culties and uncertainties, and different methods were used by the many data producers and data centers. For future efforts, a study should be made of methods of eliminating duplicate and near-duplicate reports, so that a suitable scheme can be developed in advance of the operations of the experiments. 3.3 Communications Between Centers Planning for FGGE included exchanges of telex, cable, and GTS ad- ministrative messages as the primary communications media between data shipping and receiving centers. Initial planning required the inclusion of only the shipping or receiving date of a magnetic tape shipment, with identification of tape designators for tapes included in the shipment. Receivers of the tapes also included comments regarding readability of the tapes, descriptions of errors found on the tapes, and a statement as to whether any tapes had to be redone. Later it was found useful to in- clude the data period contained in the shipment, and a sequence number for the message to assist in monitoring the exchange of messages between cen- ters. This plan for communicating between the international centers for FGGE was "^ery successful. Most of the messages sent by U.S. data producers were sent via telex. Messages received by the U.S. were sent via telex and GTS. A number of the GTS messages did not arrive and had to be retransmitted to the U.S. Telex was far superior, but more expen- sive, especially for some of the non-U. S. centers. 3.4 Tracking During FGGE it became wery clear that a key ingredient to the success of each data management component was an active system for track- ing the flow of information into, within, and out of the center. This was particularly noted in regard to the operations of the U.S. Area Sub-Center. In this case data which were used to produce the data set were obtained in real-time and non-real-time. The real-time received data arrived at the area sub-center on magnetic tapes from NMC and the Air Force Global Weather Cen- tral, in different formats. The non-real-time received data were sent to the area sub-center from 37 foreign countries within the area of responsibility 205 of the area sub-center. These mailed-in data were placed on several differ- ent kinds of forms, even though unique forms were recommended for upper air and surface reports by the area sub-center. In addition, teletype pages were also used to obtain delayed data, as were other sources of data routinely kept by the National Climatic Center. These data had to be taken and processed using different programs into a common intermediate format. The data then had to be quality con- trolled and sorted, duplicates were then eliminated, and the data were then formatted to produce a final output data set. It took around 75 days to receive and prepare each shipment of data, and each shipment contained ten days of data. The operations were staggered so that at any given time data for several ten-day periods were being processed at once, each increment being in a different stage of processing. The monitoring and tracking of input, processing, and output oper- ations for such an effort is complex, and needs to be supported by well thought out plans and procedures. It is important that each center partici- pating in an experiment such as FGGE, in early stages of its preparation, prepare the detailed forms and mechanisms necessary to ensure that the functions of the center are properly carried out. While the above discussion concentrates on problems, the overwhelm- ing result of the efforts of the eleven United States data management centers was the accomplishment of a wery complex, year-long, international data ex- change in an exemplary fashion. ■ ;,(;' 206 REFERENCES Acheson, D.T. 1970: Loran-C Windfinding Capabilities: Wallops Island Experi- ments. Equipment Development Laboratory, Applied Research and Exploratory Report, ESSA Project Task No. 21 A336602, Report to Systems Development Office, U.S. Weather Bureau. Acheson, D.T. 1978: The Application of Radio Frequency Transmitters to Radio- sonde Tracking and Windfinding. Bull. Am. Meteor. Soc. , Vol 55, No. 5 pp. 385-398. Bauer, B. and J. Lienesch, 1975: VISSR Data Calibration, NOAA Technical Memo- randum NESS 64 , Washington, D.C., pp. 59-65. Beukers, J.M. 1967: Windfinding Using the Loran-C and Omega Long Range Navi- gational Systems. Suppl . , IEEE Trans. Aerosp. Electron. Syst. Beukers, J.M. 1972: Accuracy Limitations of the Omega Navigation System Employed in the Differential Mode. Paper Presented at the Institute of Navigation Conference. Beukers, J.M. 1975: Windfinding Using Navigational Aids. Paper Presented at the 3rd Symposium on Meteorological Observations and Instrumentation of the A. M.S., February 10-13, Washington, D.C. Bristor, C.L. (Editor), 1975: Central processing and analysis of geostationary satellite data. NOAA Technical Memorandum NESS 64 , National Oceanic and Atmospheric Administration, U.S. Dept. of Commerce, Washington, D.C, 155 pp. Brower, Robert L., Hilda S. Gohrband, William J. Pichel, T.L. Signore, and Charles C. Walton, 1976: Satellite-derived sea-surface temperatures from NOAA spacecraft. NOAA Technical Memorandum NESS 78 , National Oceanic and Atmospheric Administration, U.S. Dept. of Commerce, Washington, D.C, 74 pp. Chatters, G.C and V.E. Suomi , 1975: The applications of McIDAS. IEEE Trans. Geosci, Electron. GE-13 (3), 137-146. Cole, H.L., S. Rossby and P.K. Govind 1973: The NCAR Wind-Finding Dropsonde. Atmospheric Technology , No. 2 pp. 19-24. Fleming, R.J., T.M. Kaneshige, and W.E. McGovern 1979a: The Global Weather Experiment, I. The Observational Phase Through the First Special Observing Period. Bull. Am. Meteor. Soc , Vol 60, No. 6 pp. 649-659. Fleming, R.J., T.M. Kaneshige, W.E. McGovern, and T.E. Bryan 1979b: The Global Weather Experiment, II. The Second Special Observing Period. Bull. Am. Meteor. Soc. , Vol 60, No. 11 pp. 1316-1322. Govind, P.K. 1973: Dropsonde Velocity Measurements Based on Omega Signals. Atmospheric Technology , No. 2 pp. 39-42. 207 Hasler, A.F., W.E. Shank, and W.C. Skillman, 1977: Wind estimates from cloud motions: Results of phases I, II, and III of an in situ aircraft verifica- tion experiment. J. App. Meteoro. 16 (8) , 812-815. Herkert, J., B. Remondi, B. Goddard, and W. Calliott, 1975: An overview of GOES Data Flow and Processing Facilities, NOAA Technical Memorandum NESS 64 , Washington, D.C. pp. 2-20. Hubert, L.F., 1975: Note on Jet Cirrus. Emissivity W.J.R. Meteorol. Soc. 101, 1017-1019. Jessie, L, L. Thackuray, B. Sharts, and H. Sparks, 1975: Data Processing Sup- port for Spacecraft Operations, NOAA Technical Memorandum NESS 64 , Washington, D.C. pp. 33-36. McMillin, Larry M. , D.Q. Wark, J.M. Siomkajlo, P.G. Abel, A. Werbowetzki, L.A. Lauritson, J. A. Pritchard, D.S. Crosby, H.M. Woolf, R.C. Luebbe, M.P. Weinreb, H.E. Fleming, and CM. Hayden, 1973: Satellite infrared soundings from NOAA spacecraft. NOAA Technical Report NESS 65 , National Oceanic and Atmospheric Administration, U.S. Dept. of Commerce, Washington, D.C, 112 pp. Menzel , W.P., W.L. Smith, and H.M. Woolf, 1978: A Man-Interactive Technique for specifying cloud heights from sounding radiance data. Preprints of the Third Conference on Atmospheric Radiation, Davis, Calif., American Meteoro- logical Society , 154-157. Mosher, Frederick, 1975: SMS Cloud Heights; Man-Computer Data Access System (McIDAS) final report NASA contract, NAS-5-23296, 3-1, 3-29. Mosher, Frederick R., 1978: Cloud Drift Winds from Geostationary Satellites; Atmospheric Technology, No. 10. National Center for Atmospheric Research, Boulder, Colo., 53-60. Norris, K.D. 1973: Dropsonde Data Reception and Processing. Atmospheric Tech- nology , No. 2 pp. 33-34. Olson, M.L. 1977: Central Pacific VLF Signal Survey and Omega Error Predictions. Technical Note, NCAR/TN 120+EDD, NCAR, Boulder, CO. Passi, R.M. 1973: Errors in Wind Measurements Derived from Omega Signals. Tech- nical Note, NCAR/TN/STR-88, NCAR, Boulder, CO 39 pp. Phillips, Dennis, 1974: Geosynchronous Satellite Navigation Model, Internal report. Space Science and Engineering Center, Univ. of Wisconsin, Madison. Pike, J.M. , and B.R. Lee 1973: The Dropsonde Sensors and Data Transmission. Atmospheric Technology , No. 2 pp. 29-32. 208 Poppe, M.C., Jr. 1971: LO-CATE III - The Application of Retransmitted Omega to the Tracking of Remote Objects. Paper Presented at ION Omega Symposium. Poppe, M.C., Jr. 1973: A Survey of Omega Windfinding Computation Techniques. Final Report CE-4002, Cambridge Engineering, Cambridge, VT. Poppe, M.C., Jr. 1974: Omega Signal Measurement Algorithms. Final Report CE-4004, Cambridge Engineering, Cambridge, VT. Poppe, M.C., Jr. 1979: A Review of Omega Windfinding Processing Techniques Used in FGGE. Technical Report CE-4036, Cambridge Engineering, Cambridge, VT 23 pp (and appendixes). Saum, G. 1973: The Dropsonde Receiver-Transmitter. Atmospheric Technology , No. 2 pp. 25-28. Schwalb, Arthur, 1978: The TIROS-N/NOAA-A-G satellite series. NOAA Technical Memorandum NESS 95 , National Oceanic and Atmospheric Administration, U.S. Dept. of Commerce, Washington, D.C., 75 pp. Smalley, J.H. 1978/79: Aircraft Dropwindsonde System. Atmospheric Technology , No. 10 pp. 24-28. Smith, E.A., 1975: The McIDAS System. IEEE Trns. Geosci. Electron. GE-13 (3), 123-136. Smith, W.L. and H.M. Woolf, 1976: The use of eignevectors of statistical CO- variance matrices for interpreting satellite sounding radiometer observa- tions. Journal of Atmospheric Sciences , 33 , _7, 1127-1140. Smith, W.L., H.M. Woolf, CM. Hayden, D.Q. Wark, and L.M. McMillin, 1979: The TIROS-N operational vertical sounder. Bull. Am. Meteorol. Soc , 60 , 1177-1187. U.S. FGGE Project Office 1978: The U.S. FGGE Coordinating Center. Annex 8, Vol 1, Plan for U.S. Participation in the First GARP Global Experiment (FGGE). U.S. FGGE Project Office, NOAA, Rockville, MD 65 pp. U.S. FGGE Project Office 1978: Operations Plan for the University of Wisconsin Space Science and Engineering Center Satellite Data Producer. Annex 3, Vol 2, Plan for U.S. Participation in the First GARP Global Experiment (FGGE) - Data Management. U.S. FGGE Project Office, NOAA, Rockville, MD 35 pp. World Meteorological Organization 1978: The FGGE Special Observing Systems, Part A: Tropical Wind Observing Ships and Other FGGE Ship Operations. Global Weather Experiment Implementation/Operations Plan, Vol 5. GARP Activities Office, World Meteorological Organization, Geneva, Switzerland 270 pp. World Meteorological Organization 1978: Formats for the International Exchange of Level II data sets during the FGGE. Appendix 10, FGGE Data Management Plan. Global Weather Experiment Implementation Operations Plan, Vol 3, GARP Activi- ties Office, World Meteorological Organization, Geneva, Switzerland 34 pp. 209 BIBLIOGRAPHY This bibliography is a listing of publications which are considered perti- nent to the planning and implementation of the field phase of the Global Weather Experiment. As such it is not intended to be a comprehensive listing of scien- tific publications. Instead, these documents outline the evolution of the field phase of this experiment which goes back nearly two decades. GARP PUBLICATIONS SERIES - International Council of Scientific Unions/World Meteorological Organization, Geneva No. 1 An Introduction to GARP - October 1969 No. 3 The Planning of the First GARP Global Experiment - October 1969 No. 11 The First GARP Global Experiment - Objectives and Plans - March 1973 No. 18 The Monsoon Experiment (MONEX) - October 1976 No. 19 The Polar Sub-programme - March 1978 No. 21 The West African Monsoon Experiment (WAMEX) - June 1978 GARP SPECIAL REPORTS - International Council of Scientific Unions/World Meteoro- logical Organization, Geneva Report of Planning Conference on GARP - Brussels, March 1970 Report of the Planning Conference on the First GARP Global Experiment - Geneva, September 1972 Report on Special Observing Systems for the First GARP Global Experi- ment - Geneva, February 1973 Report of the Meeting on Drifting Buoys for the First GARP Global Experiment - Geneva, March 1974 Report of the First Session of WMO Executive Committee Inter-Governmental Panel on the First GARP Global Experiment - Geneva, October 1974 Report of the Meeting of Experts for the Development of a Data Management Plan for the FGGE - Washington, April 1975 Report of the Second Session of WMO Executive Committee Inter-Governmental Panel on the First GARP Global Experiment - Geneva, September 1975 Report of the Inter-Governmental Planning Meeting for the First GARP Global Experiment - Geneva, February 1976 Report of the Extraordinary Session of WMO Executive Committee Inter- Governmental Panel on the First GARP Global Experiment - Geneva, February 1976 Report of the Planning Meeting for the Monsoon-77 Experiment - Colombo, Sri Lanka, May 1976 Report of the Third Session of WMO Executive Committee Inter-Governmental Panel on the First GARP Global Experiment - Geneva, July 1976 Report of the First Planning Meeting for the West African Monsoon Experi- ment (WAMEX) - Dakar, November-December 1976 Report of the Fourth Session of WMO Executive Committee Inter-Governmental Panel on the First GARP Global Experiment - Geneva, February 1977 Report of the Third Planning Meeting for the Monsoon Experiment (MONEX) - New Delhi, February-March 1977 Report of the Fifth Session of WMO Executive Committee Inter-Governmental Panel on the First GARP Global Experiment - Geneva, December 1977 210 No. No. 1 8 No. 10 No. 13 No. 14 No. 16 No. 17 No. 18 No. 19 No. 21 No. 22 No. 23 No. 24 No. 25 No. 26 No. 27 Report of the Second Planning Meeting for the West African Monsoon Experiment (WAMEX) and Report of the Preparatory Meeting of the WAMEX Scientific and Management Regional Committee (WSMRC) - Geneva, October 1977 No. 28 Report of the Fourth Planning Meeting for the Monsoon Experiment (MONEX) - Kuala Lumpur, February 1978 No. 29 Report of the Sixth Session of WMO Executive Committee Inter-Governmental Panel on the First GARP Global Experiment - Geneva, June 1978 No. 30 Report of the Fifth Planning Meeting for the Monsoon Experiment (MONEX) Part I - Winter MONEX - Manila, October 1978 No. 31 Report of the First Session of the West African Monsoon Experiment (WAMEX) Scientific and Management Regional Committee - Ibadan, November 1978 (Eng/Fr) No. 32 Report of the Fifth Planning Meeting for the Monsoon Experiment (MONEX) Part II - Summer MONEX - New Delhi, February 1979 No. 33 Report of the Second Session of the West African Monsoon Experiment (WAMEX) Scientific and Management Regional Committee - Douala, March 1979 (Eng/Fr) No. 34 Report of the Sixth Planning Meeting for the Monsoon Experiment (MONEX) - Singapore, 5-9 November 1979 No. 35 Report of the Seventh Session of WMO Executive Committee Inter-Govern- mental Panel on the First GARP Global Experiment - Geneva, November 1979 FGGE IMPLEMENTATION/OPERATIONS PLAN Vol. 1 Summary of Implementation/Operations Plan Vol. 2 Operational Direction of the FGGE and the Related Regional Experiments Vol. 3 FGGE Data Management Plan Vol. 4 Implementation and Operations Plan for the World Weather Watch in FGGE: PART A PART B PART C Global Observing System Global Telecommunications System Global Data Processing System Vol. 5 Implementation and Operations Plan for the FGGE Special Observing Systems PART A PART B PART C PART D PART E Tropical Wind Observing Ships and Other FGGE Ship Operations Aircraft Dropwindsonde System Tropical Constant Level Balloon System Southern Hemisphere Drifting Buoy System Aircraft Integrated Data System Vol. 6 Implementation and Operations Plans for the Regional Experiments: PART A PART B PART C PART D MONEX - Winter MONEX - Summer WAMEX POLEX Vol. 7 Oceanographic Programme for the FGGE NATIONAL ACADEMY OF SCIENCES DOCUMENTS 1. Committee on Atmospheric Sciences, 1966. The Feasibility of a Global Obser- vation and Analysis Experiment, NAS-NRC Publ. 1290, Washington, D.C., 172 pp. ?21 2. U.S. Committee for the Global Atmospheric Research Program, 1969. Plan for U.S. Participation in the Global Atmospheric Research Program , National Academy of Sciences, Washington, D.C., 79 pp. 3. Committee on Polar Research, 1974a. U.S. Contribution to the Polar Ex- periment (POLEX), Part I, POLEX-GARP (North) , National Academy of Sciences, Washington, D.C., 119 pp. 4. Committee on Polar Research, 1974b. U.S. Contribution to the Polar Ex- periment (POLEX), Part II, POLEX-GARP (South) , National Academy of Sciences, Washington, D.C., 33 pp. 5. The Global Weather Experiment - Perspectives on its Implementation and Exploitation . U.S. Committee for the Global Atmospheric Research Program; Assembly of Mathematical and Physical Sciences, National Research Council, National Academy of Sciences; 179 pp. MISCELLANEOUS DOCUMENTS i 1. ICSU/IUGG/CAS and COSPAR, 1967. The Global Atmospheric Research Programme (GARP), Report of the Study Conference in Stockholm, 28 June-11 July 1967 , World Meteorological Organization, Geneva, 144 pp. 2. COSPAR Working Group VI, 1971: The Feasibility of the First GARP Global Experiment (FGGE) and the Critical ity of Initiating Systems Planning, Report to JOC, Geneva, 42 pp. 3. U.S. Plan for Participation in FGGE (First GARP Global Experiment), 1975: Goddard Space Flight Center, Greenbelt, Maryland, 20771. 4. The First GARP Global Experiment - Review of Objectives and Plans and Opportunities for Participation, 1975. Secretariat of the World Meteoro- logical Organization, Geneva, 36 pp. (First Revision January 1976). 5. Joint Planning Staff for GARP, 1980. FGGE Operations Report, Vol. 1, Observing System Operations 1 December-30 June 1979, World Meteorological Organization, Geneva, 77 pp. (Vol. 2 through 11 are planned.) 212 APPENDIX The following United States Government Agencies had a role in funding the Global Weather Experiment. U.S. Department of Commerce U.S. Department of Defense U.S. Department of Energy U.S. Department of Interior U.S. Department of State U.S. Department of Transportation National Academy of Sciences National Aeronautics and Space Administration National Center for Atmospheric Research National Science Foundation The following Foreign Governments, Research Institutions and Private Companies participated in the implementation of the United States activities associated with the Global Weather Experiment. American Airlines British Airways Cambridge Engineering Defense Nuclear Agency Directorate, Defense Research and Engineering European Space Agency Federal Aviation Administration, DOT Finnish Meteorological Institute General Electric/MATSCO Global Associates Goddard Space Flight Center, NASA Government of Argentina Government of Australia Government of Mexico Government of the United Kingdom Holmes and Narver, Inc. Japan Meteorological Agency (Meteorological Satellite Center) KLM Royal Dutch Airlines Kuhne and Nagel Lewis Research Center, NASA Lockheed Georgia Company Lufthansa German Airlines National Oceanic and Atmospheric Administration Data Buoy Office Envi Envi Nati Nati Nati Nati Offi ronmental Data and Information Service ronmental Research Laboratories onal Earth Satellite Service onal Marine Fisheries Service onal Ocean Survey onal Weather Service ce of Management and Budget Office of the NOAA Corps 213 Pan-American World Airways Quanta Systems Corporation Qantas Airlines Scandinavian Airlines System Service ARGOS Singapore Airlines South African Airways Synergetics International, Inc. Texas A&M University ^ : TRACOR Aerospace Trans World Airlines Trust Territory of the Pacific Islands University of Hawaii University of Miami University of Washington University of Wisconsin, Madison (Space Science and Engineering Center) U.S. Air Force Military Airlift Command Systems Command , • '^- ,' ' ! V Logistics Command ' J::^>t ;? Communications Service U.S. Army ^' ' • \ '' ' Electronics Research and Development Command U.S. Bureau of Mines U.S. Bureau of Standards ^ ,■ ■ U.S. Coast Guard ' /• ' ■ ' ■• U.S. Navy '. ' - CINCPACFLT .- Naval Oceanography Command Vaisala Oy VIZ Manufacturing Company ''_' ' ' Westinghouse Electric Corporation *" Woods Hole Oceanographic Institution ^ U.S. GOVERNMENT PRINTING OFFICE : 1981 - 345-999 : QL 3 214 ■\^ PENN STATE UNIVERSITY LIBRARIES ADDDD7D ^S NOAA~S/T 81-105