NESS 69 An Evaluation of May 1971 Satellite- Derived Sea Surface Temperatures for the Southern Hemisphere P. KRISHNA RAO WASHINGTON, D.C. APRIL 1974 noaa NATIONAL OCEANIC AND / NATIONAL ENVIRONMENTAL, ATMOSPHERIC ADMINISTRATION / SATELLITE SERVICE NOAA TECHNICAL REPORTS National Environmental Satellite Service Series The National Environmental Satellite Service (NESS) is responsible for the establishment and operation of the National Operational Meteorological Satellite System and of the environmental satellite systems of NOAA. The three principal offices of NESS are Operations, Systems Engineering, and Research. The NOAA Technical Report NESS series is used by these offices to facilitate early distribution of research results, data handling procedures, systems analyses, and other information of interest to NOAA organiza- tions. Publication of a report in NOAA Technical Report NESS series will not preclude later publication in an expanded or modified form in scientific journals. NESS series of NOAA Technical Reports is a continua- tion of, and retains the consecutive numbering sequence of, the former series, ESSA Technical Report National Environmental Satellite Center (NESC) , and of the earlier series, Weather Bureau Meteorological Satellite Laboratory (MSL) Report. Reports 1 through 37 are listed in publication NESC 56 of this ser- ies . Reports 1 through 50 in the series are available from the National Technical Information Service (NTIS), U.S. Department of Commerce, Sills Bldg. , 5285 Port Royal Road, Springfield, Va. 22151. Price $3.00 paper copy; $1.45 microfiche. Order by accession number, when given, in parentheses. Beginning with 51, printed copies of the reports are available through the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402. Price as indicated. Microfiche available from NTIS (use accession number when available). Price $1.45. ESSA Technical Reports NESC 38 Angular Distribution of Solar Radiation Reflected From Clouds as Determined From TIROS IV Radi- ometer Measurements. I. Ruff, R. Koffler, S. Fritz, J. S. Winston, and P. K. Rao, March 1967. (PB-174-729) NESC 39 Motions in the Upper Troposphere as Revealed by Satellite Observed Cirrus Formations. H. McClure Johnson, October 1966. (PB-173-996) NESC 40 Cloud Measurements Using Aircraft Time-Lapse Photography. Linwood F. Whitney, Jr., and E. Paul McClain, April 1967. (PB-174-728) NESC 41 The SINAP Problem: Present Status and Future Prospects; Proceedings of a Conference Held at the National Environmental Satellite Center, Suitland, Maryland, January 18-20, 1967. E. Paul McClain, October 1967. (PB--176-570) NESC 42 Operational Processing of Low Resolution Infrared (LRIR) Data From ESSA Satellites. Louis Rubin, February 1968. (PB-178-123) NESC 43 Atlas of World Maps of Long-Wave Radiation and Albedo — for Seasons and Months Based on Measure- ments From TIROS IV and TIROS VII. J. S. Winston and V. Ray Taylor, September 1967. (PB-176- 569) NESC 44 Processing and Display Experiments Using Digitized ATS-1 Spin Scan Camera Data. M. B. Whitney, R. C. Doolittle, and B. Goddard, April 1968. (PB-178-424) NESC 45 The Nature of Intermediate-Scale Cloud Spirals. Linwood F. Whitney, Jr., and Leroy D. Herman, May 1968. (AD-673-681) NESC 46 Monthly and Seasonal Mean Global Charts of Brightness From ESSA 3 and ESSA 5 Digitized Pic- tures, February 1967-February 1968. V. Ray Taylor and Jay S. Winston, November 1968. (PB-180- 717) NESC 47 A Polynomial Representation of Carbon Dioxide and Water Vapor Transmission. William L. Smith, February 1969. (PB-183-296) NESC 48 Statistical Estimation of the Atmosphere's Geopotential Height Distribution From Satellite Radiation Measurements. William L. Smith, February 1969. (PB-183-297) NESC 49 Synoptic/Dynamic Diagnosis of a Developing Low-Level Cyclone and Its Satellite-Viewed Cloud Patterns. Harold J. Brodrick and E. Paul McClain, May 1969. (PB-184-612) NESC 50 Estimating Maximum Wind Speed of Tropical Storms From High Resolution Infrared Data. L. F. Hubert, A. Timchalk, and S. Fritz, May 1969. (PB-184-611) (Continued on inside back cover) a a 3 NOAA Technical Report NESS 69 An Evaluation of May 1971 Satellite-Derived Sea Surface Temperature's for the Southern Hemisphere P. KRISHNA RAO ENVIRONMENTAL SCIENCES GROUP WASHINGTON, D.C. APRIL 1974 For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C, 20402. Price 55 cents -*° ftT ^ 0s ^- UNITED STATES / NATIONAL OCEANIC AND /NATIONAL ENVIRONMENTAL DEPARTMENT OF COMMERCE / ATMOSPHERIC ADMINISTRATION/ SATELLITE SERVICE Frederick B. Dent, Secretary / Robert M. White, Administrator / David S. Johnson, Director CONTENTS Abstract 1 1. Introduction 1 2. Data source 2 3. Results 2 4. Comments on the validation of satellite- derived sea surface temperatures 5 5. Summary and concluding remarks 6 Acknowledgments 6 References 6 Figures 1. Southern Hemisphere sea surface temperature anal- ysis derived from NOAA 1 scanning radiometer data for May 15, 1971 8 2. Southern Hemisphere mean monthly sea surface tem- perature analysis for May 1971, derived from NOAA 1 scanning radiometer data 9 3. Time-longitude section showing the brightness and temperature departures at 5°S derived from NOAA 1 satellite data for May 1971 10 4. Same as in figure 3 for 10°S 10 5. Same as in figure 3 for 15°S 11 6. Same as in figure 3 for 20° S 11 7. Comparison of mean monthly sea surface temperatures over the Southern Pacific Ocean for May 1971 derived from NOAA 1 scanning radiometer data ... 12 8. Comparison of mean monthly latitudinal sea surface temperature profiles at 130°W and 160°W over the Southern Hemisphere for May 1971 13 ii AN EVALUATION OF MAY 1971 SATELLITE-DERIVED SEA SURFACE TEMPERATURES FOR THE SOUTHERN HEMISPHERE P. Krishna Rao Environmental Sciences Group National Environmental Satellite Service NOAA, Hillcrest Heights, Md. ABSTRACT. An objective analysis program was used to derive sea surface temperature distribution over the Southern Hemisphere for May 1971. These observa- tions were obtained from the NOAA 1 satellite. The derived temperatures were subjected to an analysis program and daily sea surface temperature charts were generated. Examples of a daily and a monthly mean sea surface temperature chart are shown. Satellite-derived brightness values and sea surface temperature changes were used to construct time-longitude sections over the eastern part of the South Pacific for May 1971 to study the variations in these two parameters. The sea surface temperatures derived from NOAA 1 data showed good agreement with conventional ship data of the National Marine Fisheries Service. 1. INTRODUCTION Sea surface temperatures in selected regions of the oceans in both the Northern and Southern Hemispheres (Smith et al . 1970, Rao et al . 1972, Shenk and Salomonson 1972) have been derived from satellite infrared radia- tion (IR) measurements. In spite of system noise in both the high-resolution IR (HRIR) data from NIMBUS satellites and the medium- resolution IR (MRIR) measurements from the Improved TIROS Operational Satellite (ITOS) and NOAA 1 satellite, sea surface temperatures could be derived with an absolute accu- racy of 2°C or better. For most of the studies the information used was obtained in one infrared window channel, while for some other studies multi- channel information was used (Smith and Rao 1972, Shenk and Salomonson 1972) to minimize the influence of clouds on the surface temperature determination. All the above studies cited showed the feasibility of obtaining sea surface temperatures under relatively cloud-free conditions and the procedures used for these studies generally were objective. Operational environmental satellites now in orbit carry HRIR instruments and Very High Resolution Radiometers (VHRR) primarily designed for mapping the cloud cover by day and by night. The information obtained from these radiometers has shown that it is feasible to detect and monitor oceanic features such as thermal fronts, current boundaries, meanders and eddies, at least under relatively cloud-free conditions. When satellite radiation data over long periods of time become available, it will be possible to study the temporal and spatial variations of sea surface temperatures over many regions. This study shows an example of mean monthly sea surface temperature distribu- tion in the Southern Hemisphere derived from satellite radiation data for May 1971. Temperature changes over the eastern part of the South Pacific Ocean for this period are also discussed and comparisons are made with con- ventional data. 2. DATA SOURCE Data used in this study were obtained from the NOAA 1 satellite launched in December 1970. NOAA satellites are operational environmental satellites and were formerly known as the Improved TIROS Operational Satellites (ITOS); a full description of the system is given in the ITOS project report (God- dard Space Flight Center 1970). In brief, the satellite is a three-axis stabilized, Earth-oriented spacecraft designed to provide complete daily day and night coverage of the globe. It is a polar-orbiting satellite with an altitude of approximately 1500 km and carries two dual channel radiometers. One of the channels measures the radiation emitted from Earth and its atmo- sphere in the 10.5-12.5 ym region. The instantaneous field-of-view of the instrument results in a viewed spot at the earth's surface 8 km in diameter at the nadir. The global measurements are stored in a tape recorder on the satellite and are transmitted to the ground for computer processing. The infrared (IR) data obtained from the satellite are corrected for ab- sorption by water vapor in this spectral interval. The corrections vary with the viewing angle of observation, the atmospheric water vapor content, and the cloud conditions. Data presented in this paper have been corrected for atmospheric attenuation using a method developed by Smith et al . (1970). Sea surface temperature estimates were derived from the NOAA 1 IR measure- ments by using the histogram method developed by Smith et al . (1970). In this method a large number of observations over an area larger than that covered by most clouds is examined; by using the empirical rule mentioned in the above reference, sea surface temperature can be derived over most areas that are relatively cloud free. The method is objective and can be imple- mented by means of a digital computer. The derived sea surface temperatures were analyzed by the objective method developed by Holl et al . (1971) to pro- duce the complete sea surface temperature analyses over the Southern Hemi- sphere used in the present study. 3. RESULTS In studies relating to global or hemispheric distributions of sea surface temperature, a grid developed by the National Meteorological Center (NMC) is used. It consists of 64 x 64 squares over a polar stereographic projection of each hemisphere; the size of each grid square is approximately 2.5° x 2.5° (latitude-longitude) at mid-latitudes. In each grid square, approximately 1,024 satellite IR observations per observation time are used to define a temperature based on the objective technique. Temperatures cannot be obtained over some areas because of persistent cloud cover; to a certain extent, the effects of extensive cloud cover can be overcome by the objective analysis technique referred to earlier. The sea surface temperature distribution obtained over the Southern Hemi- sphere for May 15, 1971 is shown in figure 1; the isotherms are drawn at 2°C intervals. The strong thermal gradient at mid- latitudes is in good agree- ment with climatology. The warm and cold current regions along the coasts of South America and South Africa, and the warm regions along the Australian Coast are in reasonable agreement with historical observations over these areas. Some centers of low temperatures in the tropical latitudes disagree with the climatological values; this discrepancy can be attributed to cloud contamination of the satellite IR observations. Piatt (1972) compared the satellite-derived temperatures over a 3-day period with the sea surface temperatures prepared by the Bureau of Meteorology in Australia and found the differences between satellite values and ship temperatures to be between 1° and 2°C. Similar differences have been noted earlier by Smith et al . (1970) and Rao et al . (1972). A mean monthly Southern Hemisphere sea surface temperature chart for May 1971 was produced using the daily charts. The daily values at each grid point were averaged to obtain the mean monthly value. Figure 2 shows the mean monthly sea surface temperature obtained from NOAA 1 scanning radiometer infrared (SRIR) data; this is the first such map known to have been obtained by using satellite IR data exclusively. Even in this average monthly chart, the warm and cold regions along the coasts of South America, Africa, and Australia are noticeable. It is possible that the influence of clouds on the satellite observations might not have been removed completely, so some of the low -temperature areas in the tropics may be attributed to cloudiness. Mean monthly sea surface temperatures from this map will be compared with inde- pendent data and discussed later. Time- longitude sections prepared from the daily NOAA 1 sea surface tempera- ture charts for 5°, 10°, 15°, and 20°S are shown in figures 3-6. The daily charts show interesting temperature departures over the eastern South Pacific; the time- longitude sections portray some of these departures. Also shown in figures 3-6 are digitized brightness values for the corresponding period obtained from NOAA 1 satellite vidicon information. These relative bright- ness values are given on a scale of 1 to 10. The contours are drawn at inter- vals of 1, and brightness values greater than 2 are shaded to indicate gener- ally cloudy conditions. All the time sections represent a narrow longitude region between 90° and 120°W. Temperature departures rather than actual tem- peratures are given for each day at these locations. These departures are the daily departures from the mean monthly value at that longitude. Shaded areas indicate a positive temperature departure. In all these figures the emphasis is on the trend in the temperature departures rather than the abso- lute magnitude of the departure. No attempt will be made to relate the brightness variations to observed temperature departures. The purpose of providing the corresponding brightness is to show that the influence of clouds in the determination of sea surface temperatures has been minimized. Lack of correlation between the two fields is an indication that the derived temperatures represent relatively cloud-free conditions. Figure 5 shows the brightness distribution and sea surface temperature departures at 5°S. The brightness distribution during the first half of May shows a striped pattern, indicating cloudiness associated with weak tropical disturbances having a period of 2 to 3 days moving through the region. No such brightness pattern appears during the latter half of May. Surface temperature departures for the same period do not exhibit any par- ticular pattern except a general warming trend during the early part of May and again from May 18 to 30. Brightness distribution and sea surface temperature departures at 10°S, shown in figure 4, differ from those at 5°S. Noticeable cloudiness existed up to May 19 between 90° and 100°W, and scattered cloud conditions prevailed at other locations. The temperature departures show some cooling until May 18, and relative warming during the rest of the period. The region of warming is almost identical to the one shown at 5°S. Figure 5 shows cloud brightness distribution and sea surface temperature departures at 15°S. Cloudiness seems to have persisted except for the last 5 days of the period. Sea surface temperature departures show two periods of relative warming, one before May 14 and the other after May 20. The period between May 14 and 20 shows relative cooling, the cooling first taking place at 90°W and progressing westward with time. Similar temperature depar- tures are also shown in figure 6, at 20°S, where the relative cooling starts about May 15 at 90°W and seems to progress westward with time. The bright- ness data for 20°S show only a few cloudy periods compared with 15°S. Figure 7 shows a comparison of mean monthly sea surface temperatures ob- tained from satellite IR data with those published by the National Marine Fisheries Service (1971) (NMFS) . Between 100°W and 180°W there is good agreement between both sets of data, but between 100°W and the coast of South America there is a wide discrepancy. Because the NMFS analysis there is based on very sparse ship data, it probably does not correspond to the actual distribution. The satellite-derived sea surface temperature distribution shows a cool tongue of water extending from 30° to 10°S along the South Amer- ican Coast. Since there is good agreement where ship data are plentiful, perhaps more reliance can be placed on the satellite temperatures where ship data are missing as in this case. Another way to validate the satellite-derived sea surface temperatures is to construct latitudinal profiles and compare them with independent sets of data. One such comparison is shown in figure 8. Two profiles were construc- ted: one along 160°W and the other along 130°W. All the data are for the month of May 1971 except Wyrtki's (1964), which is climatology for May. The conventional data are profiles derived from the May 1971 sea surface tempera- ture chart produced by the National Marine Fisheries Service. The conven- tional data coverage extends from the Equator to 30°S, but Wyrtki's data extend to 40°S. The overall agreement is good at both longitudes, although there is 2° to 3°C difference between Wyrtki's values and the satellite data at 30° and 35°S. In the equatorial region the satellite profiles show lower temperatures than the other data. The gradients shown between 25° and 40°S are in good agreement. At least, by making comparisons of this kind whenever independent sets of data are available, the use of satellite observations can be extended to data-sparse areas where conventional information is almost nonexistent. 4. COMMENTS ON THE VALIDATION OF SATELLITE-DERIVED SEA SURFACE TEMPERATURES During the past few years a number of attempts have been made to derive sea surface temperatures from satellite window (8-12 urn) radiation data. Many of the studies (Smith et al. 1970, Rao et al . 1972) have shown that the root mean square (RMS) differences between the sea surface temperatures obtained from ship reports and those from satellites varied between 1.5° and 2.0°C. Similarly, a number of aircraft studies (Pickett 1966, Shaw and Irbe 1972) performed in the United States and Canada to determine the sea surface temperatures by remote sensing show results similar to the satellite studies, Table 1 summarizes some of the findings. The range of RMS differences Table 1 . --Comparison of root mean square (RMS) differences in in the determination of sea surface temperatures from various satellite and aircraft radiation data. RMS Difference °C Nimbus II HRIR ITOS SRIR NOAA 1 (S. Hemisphere) (with Australian data) 1966 1.7 1970 1.8 1971 1.6 AT = (T , . - T _ = 0.5°C ship sat) NOAA 2 (N. Hemisphere) 1973 1.6 AT = T , . - T J = 0.7°C ship sat Canadian studies 1.7 aircraft vs. ships U.S. Naval Oceanographic Office (Pickett) AT = (T , . - T ADT ) = 1°C ship ART' (range 0.3°C - 1.8°C) between the ship measurements and the remotely sensed values is about 1.5° to 2.0°C, the ship reports being higher by 0.5° to 1.0°C. Some of this varia- bility could be due to the different techniques used in measuring tempera- tures from ships and part could be attributed to the uncertainties in the atmospheric attenuation corrections used in this and the other studies (Rao et al . 1972, Maul and Sidran 1973). The attenuation corrections based on the present knowledge about atmospheric water vapor transmission and the effects of other particulates have not been considered. James and Fox (1972) have analyzed large amounts of extensive sea surface temperature data from ship reports and showed large variabilities in the data. They emphasized the need for adopting a standard technique to measure and define sea surface temperature. Until a well-defined standard is established, satellite- derived sea surface temperatures cannot be compared strictly with all the various types of ship reports (bucket temperatures, intake temperatures, etc.). 5. SUMMARY AND CONCLUDING REMARKS It has been shown that one can objectively derive sea surface temperatures from satellite IR information over large areas. The feasibility of gener- ating a mean monthly sea surface temperature chart using only the satellite information has also been demonstrated. A comparison between sea surface temperature analyses obtained from satel- lite IR data and an analysis based on conventional ship data showed good agreement. From recent comparisons of ship, aircraft, and satellite data, one can conclude that with the present state of the art it is possible to objectively derive sea surface temperatures from satellite IR data with an absolute accuracy of 1° to 2°C. ACKNOWLEDGMENTS I thank Julia Hart and Leonard Hatton for the analyses of the data and for the drafting of figures and Dr. E. P. McClain for his critical review of the manuscript. REFERENCES Goddard Space Flight Center, ITOS , National Aeronautics and Space Administra- tion, Greenbelt, Md. , 1970, 28 pp. Holl, M. M., Mendenhall, B. R. , and Tilden, C. E., "Technical Developments for Operational Sea Surface Temperature Analysis With Capability for Satel- lite Data Input," Prepared for Naval Weapons Engineering Support Activity Detachment (FAMOS) , 3737 Branch Avenue, Room 307, Hillcrest Heights, Md., 20031, under Contract No. N62306-70-C-0334, Sept. 1971, 73 pp. James, R. W., and Fox, P. T. , "Comparative Sea-Surface Temperature Measure- ments," WMO Reports on Marine Science Affairs, Report No. 5, WMO No. 336, Secretariat of World Meteorological Organization, Geneva, Switzerland, 1972, 27 pp. Maul, George A., and Sidran, Miriam, "Atmospheric Effects on Ocean Surface Temperature Sensing From the NOAA Satellite Scanning Radiometer," Journal of Geophysical Research , Vol. 78, No. 12, 1973, pp. 1909-1916. National Marine Fisheries Service, NOAA, U.S. Department of Commerce, Fishing Information , Fishery Oceanography Center, LaJolla, Calif., May 1971. 24 pp. Pickett, R. L., "Accuracy of an Airborne Infrared Radiation Thermometer," Informal Report 0-1-66, Naval Oceanographic Office, Suitland, Md. , 1966, 10 pp. Piatt, C. M. R., CSIRO Division of Atmospheric Physics, Aspendale, Victoria, Australia, private communication, 1972. Rao, P. K., Smith, W. L., and Koffler, R. , "Global Sea-Surface Temperature Distribution Determined From an Environmental Satellite," Monthly Weather Review , Vol. 100, No. 1, Jan. 1972, pp. 10-14. Shaw, R. W., and Irbe, J. C, "Environmental Adjustments for the Airborne Radiation Thermometer Measuring Water Surface Temperature," Water Resources Research , Vol. 8, No. 5, Oct. 1972, pp. 1214-1225. Shenk, W. E., and Salomonson, V. V., "A Multispectral Technique to Determine Sea Surface Temperature Using Nimbus 2 Data," Journal of Physical Oceanog- raphy , Vol. 2, April 1972, pp. 157-167. Smith, W. L., Rao, P. K. , Koffler, R., and Curtis, W. R. , "The Determination of Sea-Surface Temperature From Satellite High Resolution Infrared Window Radiation Measurements," Monthly Weather Review , Vol. 98, No. 8, Aug. 1970, pp. 604-611. Smith, W. L., and Rao, P. K., "The Determination of Surface Temperature From Satellite "Window" Radiation Measurements," Proceedings of the Fifth Symposium on Temperature, Washington, D.C., June 21-24, 1971 , Instrument Society of America, Pittsburgh, Pa., 1972, pp. 2251-2257. Wyrtki, Klaus, "The Thermal Structure of the Eastern Pacific Ocean," Deutschen Hydrographischen Zeitschrift, Erganzungsheft , Reihe A(8), No. 6, Deutsche Hydrographische Institut, Hamburg, 1964, 84 pp. ' W5 c ■H G C o3 • O U !/) c O r-^ C $2 o +-> 2 Cfl E W o CD fH — 1 M-l t/) 13 CD CD 3 > i— 1 •H 03 S-l > CD T3 U O W 4-1 •H | t— 1 03 rt U c -o rt CD CD fH h 03 3 +-> in a E p. s CD CD O 4-> I/) •H CD o O 03 2 i— i r- 03 en CD i—i 3 h O CD CO •P i CD i E • O i—i •H T3 CD 03 Jh U 3 bO u. 6 o h tH -a a* > • H Sh >. cti t/i "7-~ !/] CD U 1— ( o <-H in CD in 3 • H rH W rt >, > r— | nJ Jh C O n 4-1 | 3 ri 4-' in rt T3 !-i CD P. f-l E ri 0) 4-> in E cu H D CD a X iw 4-> 5h O 3 t/1 W1 • H cd o CD 2 i/i >, . >— i Bj X +-> 4-> ri c -o o E ?H 0) C +J ri CD CD E E o ■H CD T3 H ri 0) J-h X P- UO 1/1 c • H -H E c 0) c X cd o PI to H CD 1— 1 X =3 s O o Cfl z CD [t, 10 SOUTHERN HEMISPHERE (5°S'MAY 1971 BRIGHTNESS I . I I LJ I l l SFC TEMP CHANGES J I 1_ 100 90 120 110 LONGITUDE (°W) Figure 3. --Time- longitude section showing ihe brightness and temperature departures at 5°S derived from N0AA 1 satellite data for May 1971. Brightness units vary from to 9; temperature departures are in degrees C. SOUTHERN HEMISPHERE (10°S) MAY 1971 BRIGHTNESS SFC TEMP CHANGES 1- I II I 100 90 LONGITUDE (°W) Figure 4. — Same as in figure 3 for 10°S. 11 SOUTHERN HEMISPHERE (15"S) MAY 197I BRIGHTNESS J LJ I I L_ SFC TEMP CHANGES Figure 5. --Same as in figure 3 for 15°S. SOUTHERN HEMISPHERE (20 S) MAY 1971 SFC TEMP. CHANGES J I ill BRIGHTNESS -1 — I — L_l 1.1,1 100 90 LONGITUDE (°W) Figure 6. — Same as in figure 3 for 20° S, 12 >, rt s: o> c U -H O f-< <4-i 03 s P 03 i— 1 •H 03 4h 2 • H O p X t-> ,o +J 3 o3 O T3 CO PH O to •H CO 10 o3 Ph +-> to CD 03 O 13 P P -H P o p 6 -H P ■H ^3 o p z o • oo 6 03 •H O O J-i u • H 03 4-1 > Ph Sh 6 -o 1 -H l/I i H CD •