c CSS; /3 : £*'£ Vo£ ~s?p<^{- W NOAA Technical Report ERL 406-APCL 44 J- c s *res c* * Comparison of Spectroscopic and Radiometric Measurements of Upper Atmosphere Water Vapor Ira G. Nolt James V. Radostitz Russell J. Donnelly Lois P. Stearns May 1979 / ,**>■* $r c,; U. S. Department of Commerce National Oceanic and Atmospheric Administration Environmental Research Laboratories rfggjgS& ^AJENT Of ' NOAA Technical Report ERL 406-APCL 44 Comparison of Spectroscopic and Radiometric Measurements of Upper Atmosphere Water Vapor Ira G. Nolt James V. Radostitz Russell J. Donnelly Lois P. Stearns Atmospheric Physics and Chemistry Laboratory Boulder, Colorado May 1979 U. S. Department of Commerce Juanita M. Kreps, Secretary National Oceanic and Atmospheric Administration Richard A. Frank, Administrator Environmental Research Laboratories Wilmot Hess, Director NOTICE The NOAA Environmental Research Laboratories do not approve, recommend, or endorse any proprietary product or proprietary material mentioned in this publication. No reference shall be made to the NOAA Environmental Research Laboratories, or to this publication furnished by the NOAA Environmental Research Laboratories, in any advertising or sales promotion which would indicate or imply that the NOAA Environmental Research Laboratories approve, recommend, or endorse any proprietary product or proprietary material mentioned herein, or which has as its purpose an intent to cause directly or in- directly the advertised product to be used or purchased because of this NOAA Environmental Research Laboratories publication. For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 (Order by SD Stock No. 003-017-001 57-1) CONTENTS Page ABSTRACT 1 1. BACKGROUND AND OBJECTIVES 1 2. INSTRUMENT DEVELOPMENT 2 2.1 Spectroscopic Method 2 2.2 Radiometric Method 4 3. ACCUMULATED WATER VAPOR STATISTICS 5 4. CORRELATION WITH RADIOMETRIC MEASUREMENTS 5 5. ANALYSIS OF MEASUREMENTS OBTAINED IN CV-990 PROGRAM 6 6. CONCLUSIONS 6 7. ACKNOWLEDGMENTS 7 8. REFERENCES 7 APPENDIX I. Catalog of C-141A Flights During 1974 9 APPENDIX II. Data Acquisition, Reduction, and Error Analysis 11 APPENDIX III. Catalog of Object-Leg-Average Water Vapor Measurements 15 APPENDIX IV. Tabulation of Radiometric and Spectroscopic H 2 Values at 41,000 Feet 19 COMPARISON OF SPECTROSCOPIC AND RADIOMETRIC MEASUREMENTS OF UPPER ATMOSPHERE WATER VAPOR Ira G. Nolt, James V. Radostitz,* Russell J. Donnelly,* and Lois P. Stearns ABSTRACT. This report compares simultaneous spectroscopic and radiometric measurements of atmospheric water vapor above aircraft flight levels and deter- mines the water vapor overburden. The measurements taken during 15 flights indi- cate that (1) both techniques give the same water vapor overburden to within 1 /un of precipitable water vapor, and (2) the median water vapor overburden is 6.5 /*m precipitable water vapor at 41,000 ft ( - 180 mbar pressure) in the mid-latitude west- ern United States during the summer/fall season, with a range from 4 to 11 /xm. 1. BACKGROUND AND OBJECTIVES During CY 1974 a joint program was con- ducted by the National Oceanic and Atmospheric Administration (NOAA) Atmospheric Physics and Chemistry Laboratory and the University of Ore- gon Department of Physics to measure atmos- pheric emissions at high altitudes. Two unique techniques for determining water vapor burden above aircraft flight level were used. They inferred total water vapor (H 2 0) burden by the use of far infrared (IR) spectra and middle IR radiances. To our knowledge, this is the first intercomparison of the spectroscopic and radiometric techniques used for this purpose. Most of the data were acquired on the National Aeronautics and Space Adminis- tration (NASA) Ames Research Center Kuiper Air- borne Observatory C-141A at a flight altitude of 41,000 ft (12.5 km). Appendix I catalogs the data flights. The specific objective of the program was to investigate the calibration of water vapor meas- urements by a comparison of the two methods. The IR radiometric method measures the zenith (downward) absolute radiance in the water vapor rotational band of 270 to 520 cm" 1 (37 to 19 /*m). By matching the measured radiance in an iterated procedure to a calculated radiance, the H : overburden is derived for an assumed tem- perature profile and water vapor transmission function. In these experiments the temperature profile was fitted at flight level to the corrected ram air temperature. The water vapor transmis- sion functions were those given by Elsasser and Culbertson (1960). Additional details of the radio- metric method can be found in Kuhn, Stearns, and Lojko(1975). The spectroscopic method measures the rela- tive equivalent widths of adjacent oxygen (0 2 ) and water vapor rotational transitions in the re- gion of 10 to 18 cm"' (1000 to 600 ^.m). Following the general technique first proposed by Burroughs and Harries (1970), the oxygen rotational transi- tions provide the basic normalization for calibrat- ing the absolute strength of the water vapor transi- tion. This method is relatively insensitive to un- certainties in the assumed standard atmosphere 'Department Oregon. it Physics, University of Oregon, Eugene. Figure 1. Overall view of instrumentation for measuring the submillimeter sky emission spectrum. temperature profile (ARDC Model; see, for example, Goody, p. 391). Operationally, at the beginning of a flight we calculate a grid of relative emission strengths for the oxygen and water vapor line pairs over the expected range of altitudes and water vapor values. The results are stored as a "look-up" table for the operating program and serve to invert the measured relative emission line ratio to a water vapor overburden. Further details of the program are given in Appendix II. At this stage, we have obtained a broad base of comparative data using the two methods, and the main conclusions are as follows: 1) The median water vapor overburden above a nominal altitude of 41,000 ft ( ~ 180 mbar) and at mid-latitudes is approxi- mately 6.5 /.im of precipitable water vapor from spectroscopic measurements. This result applies to 12 flights that were all conducted out of Ames Re- search Center in the period betwen August and December 1974. 2) The comparison of spectroscopic and radiometric measurements confirms the basic cali- bration of atmospheric water vapor determina- tions using the \'OAA radiometer. This conclusion is now supported by results in two flight pro- grams. The C-141A observations were all at high altitudes of 37,000 to 41,000 ft, and are comple- mented by earlier CV-990 flights at largely lower altitudes. Thus, the study extends over a fairly wide range of water vapor values. 3) Analysis of the C-141A results shows that real-time monitoring of the outside air tempera- ture by an air temperature radiometer should im- prove the accuracy of both techniques. This seems to be especially important when large and rapid changes in air temperature are occurring at flight level. 2. INSTRUMENT DEVELOPMENT 2.1. Spectroscopic Method In figure 1, we see the submillimeter sky spec- trum apparatus. Dr. Nolt's hand is on the HP 2100 A computer; to the right is the sextant port control knob. Below the computer is the lock- in amplifier and above it are seen the interferom- eter-computer interface unit, chopper speed con- trol and chart recorder, and the real-time spec- trum display scope. The white dewar is the LN 2 chopper reference. On top is the light-pipe which enters the chopper housing (not seen) coupled to the Michelson interferometer with its stepping- motor drive. The detector is the 3 He bolometer cryostat connected to the interferometer; it re- quires no external pump installation. The detector temperature is 0.3 K. The total weight of this equipment is =280 lb and has a peak power con- sumption, including the computer, of -800 watts. The left-hand bay of the rack contains the equip- ment used with Dr. Kuhn's radiometer. The optical system uses a light pipe which re- quires a relatively small (~2 1/ 4-in diameter) fuse- lage port and, being sealed against the cabin pres- sure, eliminates the need for a cabin pressure window. Transmission losses and radiative contri- butions from the cabin polyethylene window in early Convair 990 (CV-990) experiments required difficult corrections, and for this reason conver- sion was made to the above light-pipe arrangement. In the CV-990 work, we recorded our inter- ferograms on an FM instrument tape recorder and analyzed the data at the university after each flight series had terminated. For the C-141A experi- ments, we developed a real-time spectrum calcula- tion and display system. In figure 1, we can see SCANS: 331- 1. SKY BACKGROUND EMISSION n A " ^ j*\ /% j*y j j-i /\/\ 1\ . g o '^ W ^""V WV\ I o m + o o » 2 V r / o I XI - m CO * r~ co j> -i s: i— i CO < ^ m < I ct m r— ^ £2) CO LU CO Ct Ct I < LU LL. —1 -< 1 O 1 O 1 o i on: o 1 1 I 1 I 1 1 1 1 1 1 1 I 1 1 1 1 I II, I i i I 10.00 20.0 .OOOE+00 NUMBER (CM-1) 25.0 SCANS: 331- 1, DATE: 26/27 NOV '4 START END TIME- 0:10 0:20 LAT- 33°43'8.0" 33°57'0.0" LONG- 1 1 6 ° 56 " 115°22' OBJECT LEG: OUTBOUND LEG AT 37000 FT CLOUD OR TR0P0SPHERIC HAZE LEVEL AT AND SLIGHTLY ABOVE FLIGHT LEVEL (200 PT SCAN PARAMETER) JAT :n VAPor? DISTRIBUTION MODEL LINE PAIR RATIO UNIFORM MONOLAYER 12.7/12.3 1.2916 17.66 14.45 (MICRONS) 14.95/14.2 .3761 15.64 11.70 AVERAGE- 16.65 13.07 AVE KUHNS H20 (RADI0METER)= 15.947 MICRONS AVE OUTSIDE AIR TEMP= -36.08 DEG CENT AVE FROST POINT = 244.31 DEG CENT UNIVERSITY OF OREGON Figure 2. A 10-min scan of sky emission obtained at a flight level of 37,000 ft. the movable mirror (lead screw) arm of the inter- ferometer with its digitally-controlled stepping motor control. A computer (HP 2100 A) and coupler interface give us real-time scanning con- trol and data inversion. Each time the computer actuates the stepping motor to move the inter- ferometer mirror, the interferogram is advanced, and an up-dated Fourier transform is displayed on the oscilloscope for instant inspection. This fea- ture has been important in obtaining a high ratio of successful data-yielding flights. Once the inter- ferogram is complete, the data are transferred over the CPU-ADAMS (Acquisition Data And Man- agement System) link, which provides a digital tape of all data for post-flight analysis. Upon the completion of each interferogram scan and strip chart plot of the spectrum, an im- mediate calculation of water vapor is carried out by the computer system. Calculations are based upon a multi-layer atmospheric model, described in Appendix II, which contains no free param- eters. The accuracy of this method for determining atmospheric water depends primarily upon the measurement accuracy of the relative 2 and H 2 equivalent widths, and of the quantum mechani- cal matrix elements of observed spectral transi- tions. Since the oxygen lines provide the basic cali- bration, there is no dependence on any absolute calibration of the system. Figure 2 is an example of the spectrum from a single 10-minute scan. The spectral resolution is = 0.3 cm" 1 . The normalization assumes the peak of the strong 18.6 cm" 1 line represents an atmos- pheric emissivity of 1, or zero transmission, and utilizes the "envelope" spectrum of the sky-refer- ence difference to correct for the relative system response. This normalization increases the ap- parent amplitude of spectral noise as one moves away from the peak system response region of 12-18 cm" 1 . Comparing this with figure 3 shows the typical difference in sky emission (or transmis- sion) for the last 4,000-ft altitude increment. With longer time averaging, better sensitivity is achieved. Figure 3 shows a flight-averaged spec- trum. It is also possible to produce an entire mis- sion-average. 3 a O uj + u_ a 5 WIS: 331- 4, 5, 6, 7, G, 9, 10, 11, 12, 13, 14, 15, 16, 17 SKY BACKGROUND EMISSION c J I L J I L i o l o I Ox Ox O 1 J ! L_ _l___i 1 J I I L .000E+00 10.00 20.0 WAVE NUMBER (CM-1) 25.0 SCANS: 331- 4, 5, 6, 7, 8, 9, 10,11,12,13,14,15,0, 1 DATE: 26/27 NOV '74 START END TIME- 1:16 6:25 LAT- 39°46'9.0" 35°39 , 5.0" LONG- OBJECT LEG: JUPITER, MOON, AND W3 COMPOSITE FLIGHT AVERAGE ERICKS0N FLIGHT 2 ,15,0, 1 WATER VAPOR DISTRIBUTION MODEL LINE PAIR RATIO UNIFORM MONOLAYER 12.7/12.3 .6367 5.85 5.41 (MICRONS) 14.95/14.2 .1942 5.44 4.75 AVERAGE- 5.64 5.08 AVE KUHNS H20 (RADI0METER)= 6.617 MICRONS AVE OUTSIDE AIR TEMP= -42.05 DEG CENT AVE FROST POINT = -44.34 DEG CENT UNIVERSITY OF OREGON Figure 3. A flight-average sky emission spectrum at a flight level of 41,000 ft. Figure 3 illustrates the high sensitivity that can be achieved in flight-averaged results for the detection of weak emission features. In this flight- average spectrum (~3 hours), the weak features agree in position and relative strength with those expected for ozone emission between 12 and 18 cm" 1 . These features exhibit effective peak brightness temperatures in the range of 2-5 K, corresponding to emission changes in the atmos- phere at the 1% level. This result is the most accurate detail of this part of the sky emission spectrum yet obtained in a zenith sky measurement at aircraft altitude. If desired, the detection of such weak emission fea- tures could be enhanced an order of magnitude by observations at low elevation angles. The error analysis discussed in Appendix 11(E) shows that the standard measurement error of the H 2 quantity derived from a single ~15-min scan is 1.5 /v.m. This is derived from a comparison of the two separate line pair results in 21 single scan reductions and does not include possible system- atic effects. Thus, for a typical object leg of ~1 h, we have a measurement uncertainty of —0.7 /xm. 2.2 Radiometric Method The radiometer consists of an optical unit with necessary electronics which continuously compares the amount of energy emitted by a tar- get to that of an internal, controlled temperature reference cavity. The sensor used is a lithium tan- talate, hermetically-sealed, pyroelectric detector. The comparison is translated into a voltage which is fed directly into a computer which in turn calcu- lates effective radiance. The careful calibration of the radiometer is a necessity since radiances meas- ured by this process range from 6.0 X 10" 6 to 14.0 x 10" 6 W cm" 2 sr" 1 and the noise equivalent radiance is 2.1 x 10~ 7 W cm" 2 sr~\ In order to infer the water vapor overburden, the radiometer is mounted to "look" upward. The radiative transfer equation (RTE) is employed for the computations. Since the RTE requires a tem- perature profile above the aircraft, this profile is obtained by using the nearest upper air station soundings and corrected ram air observed at flight level. The computer iterates the RTE. stepping the water vapor burden until the condition of the dif- ference between the calculated radiance and the observed radiance is less than or equal to the noise equivalent radiance. In these calculations, the transmission functions of Elsasser and Culbertson (1960) were used. 3. ACCUMULATED WATER VAPOR STATISTICS Figure 4 is a histogram showing the results of water-vapor determinations on 12 flights which took place between August and December 1974. The preliminary results represent 44 flight-leg average values at 41,000 ft altitude. The median water vapor corresponded to a 6.5 fim precipi- table layer of H 2 above the aircraft. Values ranged from 4 to 11 ^m. A complete data tabula- tion is given in Appendix III. At slightly lower altitudes a few results, shown in table 1, indicate a steep gradient of water vapor with altitude near 41,000 ft. Table 1 also shows the advantage that accrues to the air- craft observatory in operating at the highest pos- sible altitude. Figures 2 and 3 illustrate the relative spectral difference with altitude. 4. CORRELATION WITH RADIOMETRIC MEASUREMENTS A preliminary comparison of the spectro- scopic data (tabulated in Appendix IV) with the simultaneous in-flight radiometric values is pre- sented in figure 5. The two methods agree best when the outside air temperature readings indicate stable ambient conditions. During periods of ex- treme temperature excursions, more scatter is seen 25 20 > £15 10 r "i r i — i — r No. Flights 12 (Aug. - Dec. 1974) No. Leg - Averages 44 Median Value 6.5 fj.m Range 3.9 to 1 1 .4 5 10 Leg - Average Spectroscopic H 2 (/Am) Figure 4. Statistical variation of leg-average spectroscopic water vapor at a flight level of 41,000 ft. 15 E i i i i | • - AOAT < 5°C o -5 < AOAT < 8 + - 8 < AOAT I 10 a CO i i i r / / V / • / ° / • o ° • / o f • + / / / / L / oL^_i i i i I i i l o 5 . 10 Leg - Average Radiometric H2O (u.m) Figure 5. Comparison of spectroscopically and radiometrically determined amounts of water vapor measured at 41,000 ft on C-141A flights. Table 1. Water vapor overburden at different altitudes Flight (Day) Water Vapor Overburden 1,1 ml % Reduction/ Date 37,000 ft 39,000 ft 41,000 ft* Altitude Increment 22/23 Aug. (235) 12.9 9.6 26% /2000 ft 25/26 Nov. (330) 10.8 7.1 34% /4000 ft 26/27 Nov. (331) 13.3** 6.6 50% /4000 ft 26/27 Nov. (331) 9.2 6.6 28% /2000 ft "Values for leg immediately following that at lower altitude. "Average of two consecutive scans. Table 2. Comparison of results from two methods of obtaining water vapor overburden Outside Air Temperature Change During Measurement Spectroscopic Minus Radiometric H : values (nm) Mean Difference Standard Deviation Number of Comparisons AT < 5 + 0.2 0.9 9 5 < AT < 8 + 0.1 2.4 10 AT> 8 -1.9 3.2 6 in the correlation. However, on the basis of the present limited statistics, which are shown in table 2, there is no significant mean difference be- tween the two methods. The larger scatter under changing conditions is probably due to a number of factors. Large tem- perature changes generally imply large water vapor changes, and the two methods do not repre- sent identical averaging over time. Both methods also have intrinsic dependencies on temperature conditions, which at this preliminary stage of the analysis could account for additional uncertainties of - 1 /tm. More refined analysis of the data with regard to temperature effects and the subsequent installa- tion of a free-air temperature radiometer will sig- nificantly improve the results. At the time of these measurements, it is very encouraging to find a cor- respondence within approximately 1 ^uri under "stable" conditions between the two funda- mentally different modes of measurement. different in some respects. The results of one flight that has been re-analyzed are shown in table 3. The comparison of the two methods, based on the data in table 3, is illustrated in figure 6. The agreement between the two methods is excel- lent, except for the two observations at altitudes below 30,000 ft. This agreement, together with the C-141A results discussed previously, is good evi- dence that the calibrations of the two methods are in good agreement over a water vapor range of -5 to 30 ftm and altitudes between 30,000 and 41,000 ft. 6. CONCLUSIONS Considerable data have been obtained which support the basic absolute calibration of water vapor measurements provided by the NOAA-radi- ometer instrument. We are also investigating the 5. ANALYSIS OF MEASUREMENTS OBTAINED IN CV-990 PROGRAM The current C-141A flight program is largely restricted to observations at altitudes between 39,000 and 41,000 ft. For the most meaningful cross-correlation study, it is important to have a comparison of the radiometric and spectroscopic results over the widest possible range of water vapor and altitudes. Lower altitude data were obtained on five CV-990 flights in 1972. The measurements are of excellent quality, and these results are being re- analyzed by means of the same multilayer atmos- phere model described in Appendix II. The instru- mentation was basically the same as that currently employed in the C-141A program, except that the data acquisition was less sophisticated without an on-line computer and the optical arrangement was 1000 E 100 u- 10- 1 1 1 1 | 1 ! 1 1 I 1 i i - • Uniform Mixing Model - - + Monolayer at Aircraft Altitude - / - / / _ ■H- / - / - / V .A ■ . <* _ %/ - / - / / - / / - / - / ' i i > i i . i i i i i i i 10 100 Kuhn Radiometer PWV (^.m) 1000 Figure 6. Comparison of spectroscopically and radiometrically determined amounts of water vapor measured between 23,800 ft and 39,800 ft on CV-990 flights. Table 3. Comparison of radiometric and spectroscopic results, Convair 990, 1 September 1972 Scan Identification Altitude (ft) Precipitable Water Vapor (^(im) Spectroscopic* Radiometric** Uniform Monolayer Average ' 9.1.446 23,800 114 72 38 9.1.537 27,700 101 72 33 9.1.854 31,000 31 27 29 26-33 9.1.590 31.000 22 17 20 22 9.1.921 32.500 22 17 20 25 9.1.660 36,400 14 11 12 10-14 9.1. 1028 27,100-38,600 12 9.0 10 11-13 9.1.755 39,800 8.7 7.8 8.3 5.9-7 'Values derived from average of Oakland and Salt Lake City radiosonde temperature profiles. 'Values from strip chart record. No computer printout on flight. implications which these findings may have for further refinements and improvements of water vapor measurement techniques. For example, analysis of the initial results indicates that the in- stallation of a free-air temperature radiometer in the C-141A program has significantly improved the accuracy of measurements when ambient con- ditions are undergoing rapid changes at the flight level. Improvements in the stability and tempera- ture of the radiometer reference source are also being employed. Analysis of the comparative data, however, indicates that outside air tempera- ture monitoring may be the most important factor to pursue at this stage. staff, in particular B. Kelley, D. Olson, B. Horita, and D. Oishi of NASA, and E. Morales of Sterling Engineering in the C-141 KAO program. We ac- knowledge also the contributions in 1972 by the CV-990 staff, especially our deceased colleagues, H. Cross and D. Wilson. Support for this program was provided by NASA Grant NGR 39-003-034 and NOAA Grant 04-5-022-4. The unique 3 He-cooled detector was developed in 1971 with NASA support under Grant NGR 38-003-021. 8. REFERENCES 7. ACKNOWLEDGMENTS Important, and in many cases essential, con- tributions to this work have been made by many people. R. M. Cameron has provided support and encouragement to the concept of auxiliary meas- urements to document the sky characteristics. C. M. Gillespie has furnished close liaison for engineering and flight assistance throughout this effort, and along with J. Bull (Northrup Engineer- ing) and J. McClenahan (NASA), contributed much to the efficient system installation. J. Masten- bfook of the Naval Research Laboratory has pro- vided frost-point measurements which are valu- able in the data analysis. The design of the soft- ware and ADAMS interface benefited from the ad- vice and help of R. Munoz (NASA) and the staff of Informatics (D. Wilson, J. Panteleo, T. Mathe- son). J. Nosier (Univ. of Oregon) designed the in- terferometer-computer interface and wrote the operating programs, and J. Gibbons (Univ. of Oregon) provided the efficient set of data recovery and post-flight analysis programs. Finally, we were given unstinting help by all the flight support Burch, D. E. (1968). Absorption of infrared radiant energy by C0 2 and H 2 0. III. Absorption by H 2 be- tween 0.5 and 36 cm' 1 (278 /x-2 cm). /. Opt. Soc. Amer. 58:1383-1394. Burroughs, W. J., and J. E. Harries (1970). Direct method of measuring stratospheric water vapour mixing ratios. Nature 227:824-825. Elsasser, W. M., and M. F. Culbertson (1960). Atmos- pheric radiation tables. Meteorological Monographs, Vol. 4, No. 23, Amer. Meteorol. Soc, Boston, Massachusetts, 43 pp. Gebbie, H. A., W. J. Burroughs, and G. R. Bird (1969). Magnetic dipole rotation spectrum of oxygen. Proc. Roy. Soc. A:310:579-590. Goody, R. M. (1964). Atmospheric Radiation: I. Theo- retical Basis. Oxford Univ. Press, London, 436 pp. Kuhn, P. M., L. P. Stearns, and M. S. Lojko (1975). Latitudinal profiles of stratospheric water vapor. Geophys. Res. Lett. 2:227-230. Schulze, A. E., and C. W. Tolbert (1963). Shape, inten- sity and pressure broadening of the 2.53 mm wave- length oxygen absorption line. Nature 200:747-750. Appendix I Catalog of C-141A Flights During 1974 Date (Day) Principal Investigator Comments 1 . 16 May (136) Engineering flight 2. 22 May (143) Erickson (NASA Ames) 3. 23/24 May (144) Erickson (NASA Ames) 4. 4/5 May (156) Hoffman (Univ. of Arizona) 5. 6/7 June (158) Hoffman (Univ. of Arizona) 6. 16/17 July (198) Harper (Univ. of Chicago) 7. 17/18 July (199) Harper (Univ. of Chicago) 8. 18 19 July (200) Harper (Univ. of Chicago) 9. 23/24 July (205) Harper (Univ. of Chicago) 10. 24 '25 July (206) Harper (Univ. of Chicago) 11. 29/30 July (211) Harper (Univ. of Chicago) 12. 14 15 Aug (235) Ney (Univ. of Minnesota) 13. 22 '23 Aug (235) Ney (Univ. of Minnesota) 14. 5/6 Sept (249) Soifer (Univ. of Calif., San Diego) 15. 18 19 Sept (262) Gautier (Obs. de Paris) 16. 20/21 Sept (264) 17. 23/24 Sept (267) 18. 24 '25 Sept (268) 19. 25/26 Sept (269) 20. 24'25Oct(208) 21.6/7Nov(311) 22. 7 8 Nov (312) 23. 11/12 Nov (315) 24. 25/26 Nov (330) 25. 26 27 Nov (331) 26. 2/3 Dec (338) Gautier (Obs. de Paris) Gautier (Obs. de Paris) Gautier (Obs. de Paris) Gautier (Obs. de Paris) Larson (Univ. of Arizona) Harper (Univ. of Chicago) Harper (Univ. of Chicago) Harper (Univ. of Chicago) Erickson (NASA Ames) Erickson (NASA Ames) Erickson (NASA Ames) First flight tests of installation strip chart records only; no digital data. Tests of various beam-splitter combinations and digital data recovery program. Limited in-flight results obtained. In-flight data generally. Limited data recovery due to intermittent electronic problem during this period. Problem finally traced to loose component in interferometer/CPU interface power supply. Complete flight data and recovery. Complete flight data but housekeeping data DPU interference problem prevented full digital recovery. Complete flight data. Strip chart records only, no house- keeping data. Complete flight data and recovery. Results useful during series in eliminating atmosphere as source of signal changes in telescope observations. Complete flight data and recovery. No results because of interferometer CPU parity-check problem. Complete flight data and recovery. Complete flight data and recovery. Included driest conditions to date. 27. 4 '5 Dec (340) Erickson (NASA Ames) No flight data. Appendix II Data Acquisition, Reduction, and Error Analysis A. Data Acquisition The computer control/data acquisition sys- tem for the sky interferometer consists of the fol- lowing components. (See fig. 1): 1) HP 2100 A (16,000 core) CPU with a CPU-CPU link for data transmission to the ADAMS. 2) An interferometer-computer interface unit which provides the following system functions: a) 8-channel 12-bit analog-to-digital input b) Programmable time base (resolution = 100 lis) c) Stepper motor power supply, translator and limit controls d) Two 4-digit LED display for parameter display 3) A dual-trace HP 1228 scope for monitoring the detector output signal and simultaneously dis- playing the developing spectrum. 4) An HP 680 6-in chart recorder for recording the interferogram scan and an immediate hard copy plot. The latter is obtained by simple switching of the CPU D/A output from the scope to recorder, whereupon the program automati- cally scales the plot rate to provide 2, 4, 10, or 20 cm"7in plot scales. Basic operator control and feedback is by the CPU switch register and LED display. B. Operating Program Upon reaching flight altitude, the operating computer program is sequenced through the fol- lowing subset of programmed functions where all operator control and parameter entry is by means of the 15-bit switch register. a) Scan parameter entry (number increments, in- tegration time, etc.). b) Manual positioning control of interferometer is necessary. c) Step-wise generation of interferogram with automatic adjustment of zero phase position and rejection of any large transient signal per- turbations. The spectrum is computed concur- rently with the scan and displayed on the scope. d) Upon scan completion and manual switch to "plot," the spectrum is plotted on the chart re- corder to provide real-time hard copy output. During this period, the mirror drive automati- cally resets for the next scan. e) Digital integration of the appropriate 2 /H 2 line pairs is performed and an interpolated table "look-up" provides the water vapor value for each line pair and the two water vapor distribu- tion models. These values are displayed in the 1/0 data registers for log entry. f) The complete set of scan and derived values is transferred by the data line to the ADAMS sys- tem for magnetic tape storage. At this point, water vapor values can be accessed for display, if desired, with other flight parameters. g) Default mode recycle to (c) above. An example of the strip chart records ob- tained in steps c-e is shown in figure A. Ratio Uniform Monolayer 1.06 13.4 1 1 .5 fivn 0.32 12.4 9.7 12.9 10.6 Figure A. Example of strip-chart record (Scan 331-2) showing (a) measured interferogram, (b) spectrum plot and imme- diate in-flight water vapor result. 11 SCANS i p- 331- 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 UN-NORMALIZED SPECTRUM Figure B. Un-normalized measured spectrum and the computer-calculated system response background spectrum for normaliz- ing I he relative instrument response. C. Model Atmosphere Calculation On the basis of previous exploratory studies in the CV-990 program, the following procedure for deriving the water vapor overburden from the measured far-infrared sky emission spectrum was developed. A multilayer emission model of the at- mosphere is solved and a tabular relation between the relative emission strengths for two oxygen- water vapor emission line pairs (12.312.7 and 14.2-14.9 cm" 1 ) on a grid of altitudes and water vapor values is stored. The calculation can accept current radiosonde temperature sounding values if available, and generally is computed during the climb to operating altitude. A maximum of 35 min is required for a full table generation spanning 30,000 to 53,000 ft in altitude and 0.5 to 64 /xm of precipitable water vapor. The results are stored as an accessible "look-up" table for the operating program and serve to invert the measured relative emission line ratios to water vapor values. The program utilizes two different models for the assumed water vapor distribution with respect to altitude. It does the computation for a uniform mixing model and a monolayer concentration of the water vapor at the flight altitude. These two results are taken to indicate the range of resultant water vapor for possible mixing distributions be- tween these two limiting cases. The temperature profile of the atmosphere is approximated by assigning mean temperatures (Ti) on a grid of 20-mbar pressure increments (P,). Then, for each layer (/) and a given transition, the central absorption A, is computed where A, o _ S(TiPi)qi voAJuPx) and S(TjPi) = line strength parameter for (T,P, a(TjPi) = linewidth parameter for (T,P,) q, = amount of 2 or H 2 in i' h layer. (1) 12 The emission from ;' ,h layer is l„i(v)dv = B v (Td {l-exp[- A? L(p)]}dp (2) where L{v) = line shape function B„{T,) = Planck function. Equation (2) is basically a statement of the Kir- choff relation between emission and transmission. Finally, the summation of layer emission contribu- tions less the absorption by the intervening layers gives the theoretical emission profile /„ [v) for a given line, i.e., N N /„(") = £ ZwdOexpI- E A°L(u)]dv. i' = l ;'=i + l The summation runs from the level of interest (e.g., flight level, 1=1) to the uppermost atmos- phere layer, i = N having a mean pressure of 10 mbar. The above summation is calculated as a func- tion of frequency for each transition from which the ratio of 2 /H 2 line emission can be com- pared to the experimental result for deducing the integrated column density of water vapor. For the transitions at 12.70 and 14.95 cm" 1 , we employ the two FFO line strength parameters, temperature dependence, and Van Vleck-Weiss- kopf lineshape parameters as given by Burch (1968). For the 2 magnetic dipole transitions at 12.3 and 14.2 cm" 1 , we use line strength values as given by Gebbie (1969), and the linewidth values as measured by Schulze and Tolbert (1963), with an inverse temperature dependence and a Lorentz lineshape function. D. Spectrum Normalization and Post-Flight Analysis For the in-flight water vapor calculation, a simple linear interpolation is made across the base of the emission features in order to define the equivalent width ratios for the two 2 /H 2 line pairs. This ratio is inverted by the "core-stored" tabular relation computed from the model atmos- phere for the immediate in-flight H 2 estimate. In post-flight analysis, the ADAMS-logged spectroscopic data and other selected "housekeep- ing" information are copied to another tape upon the conclusion of the flight. At this point, it is pos- sible to analyze the data in any number of ways. One useful way is to reduce it by object-leg aver- ages, wherein all scans on a given leg and/or simi- lar atmospheric conditions are averaged into one result. The system response function is defined by a computer fit to the spectrum exclusive of the emis- sion features, which essentially represents the dif- ference spectrum between the cold sky and the 77 K reference black-body. The system response is basically defined by the 125 jtim beamsplitter thickness and the detector cold filter. Scaling the integrated signal level to reference-closed port measurements yields a mean sky brightness at 41,000 ft in this passband of -10 K. The meas- ured spectrum as shown in figure B is then divided by the amplitude of the system response function to give a normalized representation as illustrated by figure 3 in the text. At this same time, supple- mentary housekeeping information is also re- trieved and included in the output format, includ- ing the 15-min segments of radiometric H 2 and outside air temperature simultaneous with each scan as illustrated in figure C. These latter plots are very useful for identifying anomalous or changed atmospheric conditions for more detailed analysis. Figure C. Records of the simultaneous radiometric water and outside air temperatures retrieved from the ADAMS house- keeping system. 13 E. Accuracy of Spectroscopic H : Measurements The accuracy of the spectroscopic H 2 meas- urements should be considered from two aspects: spectral measurement uncertainties and systematic uncertainties. A comparison of the results for the two line pair ratios computed for each spectrum provides a useful measure of non-systematic errors, since under ideal conditions both results should be identical. Taking 21 spectra now re- duced as single scans, each equal to a — 15-minute measurement, we find the mean difference be- tween the 12.3/12.7 and 14.2/14.95 line pair re- sults is 0.03 (tm of H 2 0, with a derived standard deviation of 1.5 /xm for a given spectrum. This result is extremely useful since it suggests (a) that there is no systematic difference between the two independent line pair constants or their normalization, and (b) that the standard measure- ment error is 1.5 ^m per 10 minutes of observa- tion. For a typical object leg of - 1 hour, the spec- troscopic water vapor measurement uncertainty thus becomes - 1 /im, exclusive of any systematic effects. From the tentative convergence of the spec- troscopic mean H 2 results with the radiometer results, there is no reason to suspect any syste- matic errors in the model constants. The only other likely source of systematic errors could be uncertainties in the knowledge of the atmospheric temperature profile above the flight level. Since the derived H 2 quantity depends on the relative ratio of 2 to H 2 equivalent widths, the tem- perature enters only as a second-order effect to the degree it affects the 2 and H 2 strength dif- ferently. For example, a gross shift of the entire temperature profile by 10°C results in <10% change in the derived result, or <0.6 ^m for the typical value at 41,000 ft. All results given in this report have been derived assuming a common "standard" temperature profile (Goody, 1964). 14 Appendix III Catalog of Object- Leg- Average Water Vapor Measurements This appendix contains an object-leg-average listing of water vapor results obtained for all 12 flights between 22/23 August and 2/3 December of 1974. Definitions of column-head terms follow. Scan Sequence Number: sequence number for each scan during flight. Object Leg: astronomical object of telescope observation by principal investigator. Time: the universal time interval retrieved from housekeeping records by the post-flight analysis program. In a few cases strip chart records were used. Average Temperature (°C): uncorrected average air temperature. This value reads approxi- mately 24 °C higher than the true air tempera- ture (true air temperature = uncorrected air temperature -24 °C for a true air speed of 425 kn). It is obtained, along with a tempera- ture-time plot, from the housekeeping records by the post-flight analysis program. FIR Spectroscopic: the number of sky emission scans averaged together, the water vapor over- burden based on the observed emission strength ratios of two 2 /H 2 line pairs, and the range of overburden given by two distribution models. The average value is the mean of the two values defining the model range. The model range values are derived for uniform mixing with altitude (high model limit) and for all the water vapor concentrated in a monolayer at flight level (low model limit). Radiometric: average equals the arithmetic mean of NOAA radiometer water vapor values during the time interval. The range represents the ex- treme values during the same time interval. Date: 22/23 August 1974 (Day 235) Principal Investigator: Ney (Univ. of Minn.) Water Vapor Overburden (/ 5.6 1 8-11 -38 8.3 3.8 7.0 -3.2 12 -44 1.3(6)** 6.0 4.3 1.7 13,14 -47 2.9 5.9 5.4 0.5 15-17 -43 5.9 6 1 9.0 -2.9 338 4,5 -40 11 5.7 8.3 -2.6 8-11 -23 5.1 5.3 6.6 -1.3 12 -26 5.1 3.0 2.7 1.2 14 -25 2.0 4.6 2.9 1.7 15 -23 2.5 3.9 5.0 -1.1 'Uncorrected air temperature. See definition in Appendix III. 'Large change in temperature immediately preceding period of measurement. 19 'U.S. GOVERNMENT PRINTING OFFICE: 19 7 9-0-677-034/4 r LABOR AT O RIES The mission of the Environmental Research Laboratories (ERL) is to conduct an integrated program of fundamental research, related technology development, and services to improve. understanding and prediction of the geo- physical environment comprising the oceans and inland waters, the lower and upper atmosphere, the space envi- ronment, and the Earth. The following participate in the ERL missions: W/M 'lans a NHEML National Hurricane and Experimental Meteor- ology Laboratory. Conducts research on trc RFC meteorology such a systems, to imi casting accuracy ai beneficial modificat systems. irncanes and cumu'us terstanding and ) develop methods for weat! Research Fac<> os instru- mented aircraft for environmental research pro- grams, and pre. measurement tools, logged data, and associated information fc airborne and selected surface rese programs AOML Atlantic Oceanogra?- Laboratories. Stud geological characteristic ocean waters, the sea U APCL Atmospheric Physics and Chemistry Laboratory. Conducts research on mesoscale meteorology related to atmospheric pollution and inadvertent weather modification, and to the prediction of mesoscale weather phenomena. Research subjects include cloud and precipitation physics, diagnosis and numerical models of mesoscale weather events, secular changes in atmospheric chemical and particulate composition, and radia- tive heat transfer applications. WPL Wave Propagation Laboratory Studies the propa- gation of sound waves at millimeter, infrared, and optica quencies to develop new methods for remote irinq of the qec environment. ARL P-MEL Pacific Marine Envin AL the deep-ocear; of the Pac Aeronomy Labc dynamics, and chemis grounding upper ie st NSSL ■GFDL Geophysical Fluid Dynamics Laboratory. Studies the dynamics of geoph, systems (the atmosphere, the hydrosphere, and the cryc- sphere) through theoretical analysis and numer- simulafion using powerful, high-speed digital computers. SEL \tory. ■MHH^MHMMHHBUHRH^HBHHaHHHHHSaBMmaHHBMI ^ U.S. DEPARTMENT OF COMMERCE National Oceanic and Atmospheric Administration BOULDER, COLORADO 80302 PENN STATE UNIVERSITY LIBRARIES AQD0D7ED2V