THE UNIVERSITY OF MICHIGAN COLLEGE OF ENGINEERING High Altitude Engineering Laboratory Departments of Aerospace Engineering Meteorology and Oceanography Quarterly Report HIGH ALTITUDE RADIATION MEASUREMENTS 1 January 1971 - 31 March 1971 Fred L. Bartman ORA Project 03635 under contract with: NATIONAL AERONAUTICS AND SPACE ADMINISTRATION CONTRACT No. NSR 23-005-376 WASHINGTON, D. C. administered through OFFICE OF RE S EARCH ADMINISTRATION ANN ARBOR April 1971

QOsS 0t\ltq QnO'+T~~~~~~~I f J t X _ ~ ~ ~

Table of Contents Page List of Illustrations v List of Tables vii Abstract ix I. Introduction II. High Resolution Measurements of the 15pm Absorption Band of CO2 (L. W. Chaney). 1 A. Long Path (20 meter Measurements) 1 B. Isotope Measurements 2 C. Detector Investigation 2 III. Analysis of High Resolution 15pm C02 Spectra (S. R. Drayson). 3 IV. Medium Resolution Measurements of the 9. 6pm Absorption Band of CO2 5 A. Experimental Details (L. T. Loh) 5 B. Computer Processing of the Infrared Data (P. A. Titus)l 7 V. Report Writing 20 VI. References 20 iii

List of Illustrations Figure Page 1. Spectrum of mercury vapor lamp showing six lines that will be used for determination of 03 amounts. 21 2. Amount of 03(in Atm-cm) in 3200 cm. absorption cell vs. I/Io (%) for six Hg lines. 22

List of Tables Table Page la Band center, rotational constants and wave pbers of lines in the P and R branches of the 12C 02 V 6 fundamental (010:1)-(000:0). 6 lb Rotational constants and wavenumbers of lines in the Q branch of the 12C1602 fundamental (0100:0)-(000:0). 7 2a Band center, rotational constants and wavenumber of lines in the P and R branches of the 12C1602 (020:2)(010:1) band (c-c). 8 2b Band center, rotational constants and wavenumbers of lines in the P and R branches of the' 12C160 (020:2)(010:1) band of (d-d). 9 3a Band center, rotational constants and wavenumbers of lines in the P and R branches of the 12C1602 (030:3)(020:2) band (c-c). 10 3b Band center, rotational constants and wavenumbers of lines in the P and R branches of the 12C1602 (030:3)(020:2) band (d-d). 11 4 Band center, rotational constants and wavenumbers of lines in the P and R branches of the 13C160 fundamental (010:1)-(000:0). 12 5 Band center, rotational constants and wavenumbers of lines in the P and R branches of the 12C160180 v2 fundamental (010:1)-(000:0). 13 6 Ozone absorption coefficients in the ultraviolet for six Hg lines 16 vii

Abstract This report summarizes project activity during the period 1 January 1971 to 31 March 1971. Experimental measurements of 15p/m CO2 data are described. Preliminary results of the analysis of some of the data is given. Progress in experimental measurements of 03 spectra and the method of processing these medium resolution data are described.

1. Introduction This is the 6th Quarterly Progress Report on Contract No. NSR 23-005-376, covering the period 1 January 1971 to 31 March 1971. The project effort during this time was divided among the following tasks. A. High resolution measurements of the 15pm absorption band of CO2. (L. W. Chaney). B. Analysis of high resolution 15gm absorption band of CO2 spectra. (S. R. Drayson). C. Medium resolution measurements of the 9. 6 pm absorption band of 03 (L. T. Loh and P. A. Titus). D. Report Writing II. High Resolution measurements of the 15 gm Absorption Bana of CO2. L. W. Chaney. A. Long Path (20 meter) lVIeasurements During the previous reporting period a program was initiated to cover the entire 15n band at high resolution using the approximate 20 meter path through the complete instrument. These measurements were continued and completed by Feb. 1, 1971. The only problem encountered was due to the wavelength shift resulting from a change in the index of refraction of the gas as the pressure was increased. Fortunately, the solution was quite simple. Known calibration lines were included in every scan. Supplementary data was taken with the 8. 74 cm cell at low pressure near the main Q-branch lines. This was done to identify individual lines in the Q-branch. 1

B. Isotope Measurements Three isotope enriched carbon dioxide samples were obtained. (1) C13016016' (2) C12016018, and (3) C12016017. The measurements were made in the 8. 74 cm sample cell. The gas manifold was modified to receive the sample and a provision was made to trap the sample after the measurement using liquid nitrogen. The measurements on the C13 sample were carried out satisfactorily in about two weeks with no complications. The 018 measurements were delayed slightly and not completed due to a leak in the sample bottle. A portion of the sample was lost, but has now been replaced by the supplier (Isomet Corp. ). The leak was difficult to find. The system was not leaking to the outside atmosphere but from the sample to the gas manifold. The helium leak detector was not usable for this condition. Data was taken on the third sample enriched in 017 The measurements were approximately 90% completed when the spectrometer SIN ratio began to deteriorate. An investigation has shown that the problem was due to a grating location screw which had become loose and thus had slightly tipped the beam. The spectrometer must now be re-aligned and re-assembled. C. Detector Investigation Two infra red detectors were received from Texas Instruments (1) The Cu-Ge detector is intended for the 15p study. (2) The Indium-Antimonide detector is intended for the 4. 4l study. 2

The sensitivity of the detectors was checked with a set-up as close as possible to that reported by Texas Instrument. The check indicated that the sensitivity of the Cu: Ge was low and when used in the spectrometer the output voltage was identical to a unit belonging to the Willow Run Laboratories, but the noise was twice as large. After re-checking the detector several times it was returned to the manufacturer. The unit has now been returned to Michigan and will be checked in the spectrometer. The other detector seems to be satisfactory. During the next quarterly work period it is planned to re-assemble the spectrometer and complete the isotope CO2 data. The new indium antimonide detector and a new grating will be installed in order to take data in the CO2 4. 31/ band and the CO 4. 6ij band. III. Analysis of high resolution 15M m CO2 spectra. (S. R. Drayson) The analysis was continued to determine accurate values for the vibrational and rotational constants for CO2 in the 15/gm region. Some of the preliminary results are shown for some of the bands in Table 1-5. It is anticipated that only minor modifications will be necessary for Tables 1-3, but more major changes will be made to Table 5 when the analysis of the isotope enriched samples is completed. The following is a brief discussion of the results and the methods used to obtain them. A detailed description will be given in a Technical Report as soon as the analysis of 12C16 02 is complete. The basic calibration for the work was obtained from the CO2 measurements of Gordon and McCubbin (1965). Lines of the v2 fundamental were used as calibration for the (020:2)-(010:1) band. These were then used 3

to recalibrate lines of the v2 fundamental and an iterative procedure was thus established. Convergence was very rapid (2 iterations). Note that the band center of the v2 fundamental has been shifted slightly from Gordon and McCubbin's value (667. 379 cm ) to 66'r. 381 cm. The rotational constants for the ground state were provided by Benedict 1970 (private communication). Table 1 b gives results for the v2 fundamental, based on measurements from J= 32 to 68. The band center cannot be accurately determined from these data, so that the value from the P- and R- branches was used. For the (020:2) - (010:1) transition (Tables 2 a and b) a band center is obtained for the (c-c, odd J) and (d-d, even J) transitions. It is a very good check on accuracy to obtain the band center from each of these and to compare -1 the result. For this band the difference was less than 0. 001 cm. The values shown in the tables were obtained by solving simultaneously both sets of data. The accuracy of the data for the (030:3) - (020:2) band is considerably less since lines are much weaker and tend to become blended with other weak lines or influenced by random noise in the spectra. They represent the first -1 such measurement for the band. The band centers differed by 0. 002 cm when each of the even and odd set of J-numbers was treated separately. In Tables la, 2a and 2b the results were obtained from a least squares fit of the sum of P(J) and R(J). A least squares fit of the P(J) and R(J) without adding them gave results that agreed very closely. For Table 3a and b this later technique was used since there were very few J-values for which the P- and R- branch line positions could both be accurately measured. This also applies to Table 5. 4

Tables 4 and 5 are for the isotopic molecules 13C 02 and 12 16 18 12C160180 respectively. The former seems to be in a very satisfactory state, while the later will be considerably improved when the analysis of the isotopic enriched samples is complete. IV. Medium resolution measurements of the 9. 6um absorption band of 0 A. Experimental Details (L. T. Loh). During the measurement of the infrared transmissivity of ozone, the quantity of ozone in the White cell will be determined by ultraviolet spectroscopy. The U. V. measurement system consists of a mercury vapor lamp, a quartz lens system in the White cell, a monochromator, and a 1P28 photomultiplier tube with amplifier and power supply. All components except the RCA 1P28 photomultiplier and lenses were purchased from Jarrell-Ash. They worked together qualitatively very well, with plenty of signal and good resolution. The signals, however, drifted. A series of tests were performed to determine the cause of the drift. Measurements showed that the output of the mercury vapor lamp changed with the line voltage and also with the time after being turned on. When operated through a constant voltage transformer, the lamp output slowly decreased to a stable value in about an hour. Upon turning off and then turning on again, the output approached a stable value in minutes. Even though the strength of the mercury lines was kept constant, the signal from the photomultiplier tube still was not reproducible. It was suspected that this was due to the extreme sensitivity of the photomultiplier tube to the operating voltage and poor reproducibility of the operating voltage. 5

Table la Band center, rotational constants and waveljmbers of lines in the P and R branches -of the 12C O 2 2 2 fundamental (010:1)-(000:0). BAND CENTER= 667.381 8LK 0.3902180 BU= 0.3906369 BU-BL= 0.0004189 OL= 13.31D-08 U= 13.470D-08 DU-DL= 16.21D-10 J P CALC O-C R CALC O-C P+R CALC O-C, 2 665.821 0.002 669.727 -0.003 1335.548 -0.001 4 664.264 -0.001 671.295 0.002 1335.559 0.00-1 6 662.711 -0.001 672.867 0.002 1335.578 0.001 8 661.161 -0.005 674.442 0.000 1335.603 -0.005 10 659.615 -0.002 676.020 0.002 1335.635 0.000 12 658.0 72 0. 000 677.601 0. 002 1335. 673 - 0.002 14 65~6.532 -0.003 679.186 0.003 1335.718 -0.000 16 654.996 -0.000 680.774 -0.001 1335.770 -0.001 18 653.464 0.003 682.364 -0.001 - 1335.828 0.002 20 651. 935 -0.003 683.95.8 -0.001 1335.893 -0.004 22 650.410 -0.001 685.555 0.001 1335.965 -0.000 24 648.888 0.003 687*15.5 1336.043 0.003 26 647.370 0.001 688.758 0..003 1336.128 0.004 28 645.856 -0.002 690.364 0.001 1336.220 -0.001 30 644.345 -0.000 691.972 0.001 1336.318 0.000 32 642.838 -0.002 693.584 0.002 1336.422 -0.000 34 641.335 0.002 695.198 0. 001 1336.533 0*003 36 639.835 0.002 696.816 -0.003 1336.651 -0.001 38 638.339 -0.004 698.436 -0.001 1336.775 -0.005 40 636.847 0.004 700.059 -0.003 1336.905 0.002 -42.635*358 0.002 701.684 0.000 1337.042 0.002 44 633.874 -0.001 703.312 0.002 1337.186 0.001 46 632.393 -0.004 704.943 0.003 1337.336 -0.001 48 630.916 0.006 706.576 0.001 1337.492 0.007 50 629.442 -0.000 708.212 0.002 1337.654 0*002 52 627.973 0.000 709.850 0.001 1337.823 0.001 54 626*5C7 -0.003 711.491 -0.007 1337.998 -0.010 56 625.045 713.134 O.001 1338.179 -0.006 58 623.587 0.000 714.780 0.003 1338.366 0.004 60 622.132 0*001 716.428 0.003 1338.560 0.004 62 620.682 0.006 718*078 -0.006 1338.760 0.000 64 619*235 0.006 719.730 1338.965 66 617.792 721.385 -0.004 1339.177 68 616.353 723.042 -0. 003 1339.395 -0.006 70 614*918 724.700 -0.005 1339.619 -0.008 72 613.487 -0.009 726.361 -0.002 1339.848 -0.011 74 612.060 0.004 728.024 0.006 1340.084 0.010 76 6106 37 0.000 729.689 0.003 1340. 325 O.004 78 609.217 0.000: 731.3 56 1340.573 RMS O-C = 0.0030 WAVENUMBERS 6

Table lb Rotational constants and wavenumbers of lines in the Q branch of the 12C1602 fundamental (0100:0)-(000:0). BAND CENTER= 667*.381 81= 0.3902180 BU= 0.3912529 BU-BL= 0.0010349 OL= 13.310-08 DU= 13.540-08 DU-DL= 22*59D-10 J Q CALC O-C 32 668.471 -0. 001 34 668.609 0.002 36 668.755 -0.000 38 668.910 -0.002 40 669.072 -0.005 42 669.243 -0.004 44 669*421 0.001 46 669.*608 0.00.4 50 670.005 0. 004 52 670.216 0.001 54 670.435 0.001 56 670.661 -0*001 60 671. 138 -0.000 68 672* 187 -0.001 RMS O-C = 0.0025 WAVENUMBERS 7

Table 2a Band center, rotational constants and wavenumber of lines in the P and R branches of the 12C1602 (020:2)(010:1) band (c-c). BAND CENTER= 667.751 BL= 0.3906375 RU= 0.3916697 BU-BL= 0.00!C322 DL= 13.48D-08 DU= 13.940-08 DU-DL= 45.64D-10 J P CALC 0-C R CALC 0-C P+R CALC O-C 3 665.413 -0.003 670.897 1336.310 -0.002 5 663.865 0.001 672.482 -0O.L04 1336.347 -0.003 7 662.325 0.001 674.075 -0. 002 1336.401 -0.002 9 660.794 0.005 675.677 -0.001 1336.471 0.004 11 659.271 -0.000 677.286 -0.003 -1336.557 -0.003 13 657.756 0.002 678.904 -0.003 1336.660 -0.001 15 656.250 0.002 680.529 0.00.1 1336.780 O.002 17 654.752 -0.002 682.163 -0.000 1336.915 -0.002 19 653.263 -0.004 683.805 0.003 1337.068 -0.001 21 651.782 0.004 685.454 1337.236 29 645.942 692.130 -.o001 1338.073 31 644.503 O,.01 693.819 -0.004 1338.322 -0.003 33 643.073 )0.004 695. 515 0.000 1338*588 0.004 39 638.833 -0,.003 700.649 0.000 1339.481 -0.003 41 637.436 0.003 702.374 0.001 1339.810 O.004 43 636.048 7C4.1 07 0.003 1340.155 0.005 47 633.297 -0.006 707.595 -0.002 -1340.892 -0.008 49 631.934 0.005 709.350 -0.001 1341.284 0.004 51 630.580 -0.004 711.111 0.002 1341.691 -0.002 53 629.234 0.000 712.880 0.001 1342. 114 0. 001 RMS O-C = 0.0027 WAVENUMBERS 8

Table 2b Band center, rotational constants and wavenumbers of lines in the P and R branches of the 12C1602 (030:2)(010:1) band of (d-d). BAND CENTER= 667.751 BL= 0.3912529 BU= 0.3916621 BU-BL= 0.0004C92 DL= 13.54D-08 DU= 13.60D-08 DU-DL= 57.40D-11 J P CALC O-C R CALC 0-C P+R CALC O-C 2 666.187 670,103 0.001 1336.290 4 664.626 0.002 671.676 -0.004 1336.301 -0.001 6 663.068 -0,.000 673.251 0,.001 1336.319 0.001 8 661.514 0.005 674.830 -0.001 1336.344 0.004 10 659.963 0. 005 676.412 -0.003 1336.375 0.002 12 658.416 -0.001 677.997 0.000 1336.412 -0.000 14 656.872 0.001 679,585 -0.002 1336,456 -0. 000 16 655.331 0.004 681,176 0.000 1.336.507 0.004 18 653,794 -0.000 682.770 0.001 1336.564 0,.001 20 652.260 0.001 684.367 -0.000 1336.628 0.000 22 650.730 0.001 685.968 -0.003 1336,698 -0.002 24 649,204 -0.003 687.571 -0.002 1336.775 -0. 005 26 647.681 -0.0 02 689.177 -0.002 1336.858 -0.004 28 646.162 0.001 690.786 0.000 1336.947 0.002 30 644.646 0,.002 692.398 0.002 1337.044 0.004 32 643.134 0.000 694.012 0.001 1337.146 0.001 34 641.625 695.630 0.001 1337.255 0.002 38 638.620 0.004 698.873 1337.493 0.001 40 637.122 -0.003 700,499 0.001 1337.621 -0002 42 635.629 -0.001 702.128 0.001 1337.756 0.001 44 634.139 0.002 703.759 -0.003 1337,898 -0.001 46 632.653 0.004 705.393 -0,.003 1338.045 0.002 48 631.171 -0.006 707.029 0.003 1338.200 -0.003 50 629.692 0.001 70C8.668 0.000 1338.360 0.001 52 628.218 0.000 710.309 0.001 1338.527 0. 001 RMS. O-C = 0.0023 WAVENUMBERS 9

Table 3a Band center, rotational constants and wavenumbers of lines in the P and R branches of the 12C1602 (030:3)(02 0:2) band (c-c). BAND CENTER= 668.105 BL= 0.3916621 BU= C.,3923806 BU-BL- 0,0007185 DL= 13*60D-08 DU= 13.91D-08 DU-DL= 30.9CD-10 J P CALC 0 —C R CALC 0-C 23 6b50458 687. 327 -0. 002 25 648*961 688,965 0.012 27 647.469 60SC608 0,000 29 645.983 692.256 0* 04 31 644. 504 693.909 -0.011 39 638.646 -0. 020 700. 573 41 637.196 -0*003 702.252 O. 005 43 635.753 703.936 0*005 45 634.316 705 624 0,007 51 630.040 710.718 -0o004 53 628.627 712.426 0*014 55 627*220 714,138 0.007 57 625 818 715.854 -0, 015 RMS O-C = 00097 WAVENUMBERS STOP 0 EXECUTION TERMINATED 10

Table 3b Band center, rotational constants and wavenumbers of lines in the P and R branches of the 12C1602 (030:3)(02 0:2) band (d-d). BAND CENTER= 668.105 BL= 0.3916697 BU= 0.3923889 BU-BL= 0.0007192 DL= 13.940-08 DU= 13*98D-08 DU-DL= 40.97D-11 J P CALC O-C R CALC 0-C 4 664.980 67 2043 -0. 009 6 663,427 673.629 0*005 8 661.879 675.220 0.002 10 660.33 7 676. 816 0.002 12 658.801 678.418 0.003 14 657.271 -0. 009 680.026 0*004 16 655.747 681,639 0,001 8 6 54.228 -0.008 683.258 0.010 20 652. 716 0.003 684 882 -0.009 22 651.210 0.012 686*512 28 646. 727 -0.012 691.434.0000 30 645*245 -0.003 693*085. 006 32 643.770 694.742 0.012 34 642.300 696404 0. 007 36 640*836 698.071 -0*007 38 639.379 -0. 006 699 743 0.002 40 637*928 701*421 -0*004 42 636.483 703.104 -0,006 44 635.045 704. 792 0*004 58 625*153 716.748 0.001 RMS O-C = 0.0068 WAVENUMBERS 11

Table 4 Band center, rotational constants and wavenumbers of lines in the P and R branches of the 13C1602 2 fundamental (010:1)-(000:0). BAND CENTER= 648.481 BL= 0.3902350 BU= 0.3906002 BU-BL= 0.0003652 DL= 13.100-08 DU= 12.94D-08 DU-OL=-16.270-10 J P CALC 0-C R CALC O-C P+R CALC 0-C 2 646. 921 -0.005 650.827 0.001 1297.748 -0.004 4 645. 364 -0.003 652. 394 -0.001 1297. 758 -0.004 6 643.809 0.002 653.965 0o001 1297.774 0. 003 8 642.258 0.003 655.538 1297.796 0.003 10 640.710 0.000 0 657.114 0.002 1297.823 0.003 12 639.164 -0.000 658.692 0.002 1297.857 0.0001 14 637.622 -0.000 660.274 0.001 1297.896 0.001 16 636.083 661.858 0.003 1297.942 0. 004 18 634.547 -0.000 663.445 -0*000 1297.993 -0.001 20 633.015 665.035 -0.004 1298.050 -0. 004 22 631.485 666.628 0.002 1298.113 -0. 002 24 629.959 -0.000 668.223 1298.182 30 625.400 673.024 -0. 001 1298.424 50 610.428 0.003 689.195 1299.623 0. 002 52 608.950 -0.001 690.826 1299.777 -0.002 54 607.477 -0.001 692.460 1299. 937 RMS O-C = 0.0021 WAVENUMBERS 12

Table 5 Band center, rotational constants and wavenumbers of lines in the P and R branches of the 12C160180 2 fundamental (010:1)-(000:0). BAND CENTER= 662. 367 BL= 0.3681640 BU= 0.3686018 BU-BL= 0.0004378 DL= 11.lOD-08 DU= 12.45D-08 DU-DL= 13.46D-09 J' P CALC 0-C R CALC O-C 3 660.161 0.019 665.321 4 659.427 666 062 0.003 5 658.694 666~803 0.009 9 655. 772 -0*013 669.778 1 1 654*316 -0.012 671. 270 12 653*589 -0.008 672.018 13 6 52 864 0.003 672. 766 14 652*139 -0.009 673.515 -0. 001 15 651.415 674.264 0.011 16 650.692 675. 015 -0.003 17 649.970 675,766 -0.004 18 649. 248 676.5 19 0.004 21 647*090 678*780 0.009 22 646.372 0*004 679.535 23 645.655 -0.010 680.291 -0.006 24 644.939 -0.004 681.047 25 644.224 0o013 681.804 0.009 26 643. 509 -0.003 682.562 -0 * 007 28 642.083 0. 003 684. 080 31 639.950 -0.004 686.362 -0O 009 33 638. 531 -0.008 687.887 -0.009 34 637.824 00 12 688.650 36 636.410 -0.001 690.178 38 635*000 0.005 691,708 39 634.296 0.000 692*474 41 632.890 0*002 694.007 42 632.188 -0.003 694. 774 45 630*087 0.001 697.079 51 625.902 O007 701.698 65 616.216 712.510 -0.002 RMS O-C = 0.0077 WAVENUMBERS

A change of voltage from 400 to 520 volts changed the output by a factor of 7. 75. This amounted to 30% change in voltage and a 675% change in output. The high voltage meter on the power supply had a full scale of 2000 volts. The span of 400-520 volts covered only 6 small divisions on the meter. To make the high voltage setting reproducible, a test point for an external digital voltmeter was installed on the power supply. The reading of 520 volts was made to show as 5. 200 volts on the digital meter with a resolution of + 1 digit or. 02%. This addition drastically improved the resettability of the high voltage. The UV spectrum of the mercury vapor lamp was recorded on a Brush recorder (fig. 1). All 6 lines metioned in McCaa and Shaw (1967) were recorded with low bakcground and good resolution. The very strong line 0 at 3121. 5 Aon p. 30 of McCaa and Shaw was adeuqately resolved into 2 peaks. A chart (see fig. 2) showing the amount of ozone in atm-cm in the 3200 cm absorption cell vs (I/Io) x 100 was prepared, based upon absorption coefficients given in McCaa and Shaw (1967) reproduced below for reference. This table shows 2 series of coefficients, one from Hearn (1961) and the other Walshaw (1957). The differences between them were not insignificant! o The data of Hearn were used, except for the 3131. 5 A line. The net accuracy of thesedeterminations of 03 has yet to be determined. The chart records of the infrared output of the Perkin Elmer Model 221 are not corrected for random variations in 0 % and 100 % transmittance. Because of the large amount of data that will be taken and the corrections needed, digital computer processing of the absorption data is desirable. The computer not only will make the 0% and 100 % correction but also will provide the data 14

in a form suitable for further computer processing. After considering various possible ways of doing this, it was decided to use the laboratory's PDP-8 to record the raw data, to make the corrections, and to store the data for future use. The work is carried out as follows: 1. Wavelength Data The wavelength signal is generated in the form of pulses by a pentagon actuated microswitch plus power supply and logic circuits. The pentagon is mounted on a gear, rotated synchronously with the shaft of the wavelength counter. The shaft rotates at 0. 1 /a m per revolution, corresponding to 5 actuations of the microswitch, or 0. 02 p m per actuation. 2. Spectrum Data The signal is an analog voltage corresponding to the pen deflection on the recorder of the spectrophotometer. The voltage swing is limited to + 5 volts to conform to the requirement of the A to D converter of the computer. Each data channel is sent from the spectrometer to the computer through a direct wire line. Each spectrum is started and ended between 2 selected wavelength limits, running 50 data points per,u m. The 0 % and 100% lines are sampled in the same way. A purge solenoid valve for the McIntire air drying system was purchased and installed. This valve will save the air during the idling period. A survey of room temperature was made at 2 locations in the spectrometer room. At each location, a thermistor was used, and its signal monitored with a Brush recorder. 15

Table 6 OZONE ABSORPTION COEFFICIENTS IN THE ULTRAVIOLET FOR SIX Hg LINES (after McCaa and Shaw) Line No. Wavelength (A) Absorption Coefficient k(x) (cmNTP) Hearn _Walshaw (adopted) 6 2536.5 133.9 + 19 115. 7 5 2893.6 17.2 + 0.3 16.0 4 2967.-3 6.969+ 0. 030 3 3021.5 3. 340+ 0. 014 3.29 2 3131.5 0. 76 1 3341.5 0.0498+0.0007 0.0504 16

B. Computer Processing of the Infrared Data (P. A. Titus) Computer programs and interface hardware were developed during this work period for the PDP-8 data processing system. The data is processed in 3 steps. a) accept the voltage analog of transmissivity from the ozone system b) reduce it to transmissivity vs wavelength data c) punch the data on tapes for permanent storage and for further processing, plotting and analysis on an IBM 360/67 system Analog voltages representing the transmissivities of 0 %o, 100 T7 and of the gas sample (E) are generated by a precision potentiometer which is coupled to the optical attenuator of the spectrophotometer. Since the reference beam is attenuated to become equal to the intensity of the sample, the attenuator's position at null is a direct measure of the transmissivity. These voltages are connected by cable to the multiplexer on the computer. The wavelength range of the prism is represented by a set of voltage pulses generated by a cam-switch unit. This unit is coupled to the wavelength cam of the monochromator scan motor. At each switch closure, the pulse is sent by cable to the computer where it is conditioned by the interface to the proper polarity, amplitude and duration, and connected to the computer interrupt line. The data processing system consists of a PDP-8 computer, a multiplexer, an analog-to-digital converter and a time code generator. When the computer detects an interrupt pulse, the transmissivity analog voltage is 17

channeled through the multiplexer, measured and digitized by the analogto-digital converter, and stored in the computer for processing. The general procedure for the determination of the transmissivity of an ozone sample as a function of wavelength is as follows: 1. The 40 m cells are evacuated to better than 1. 0 mm of Hg and the spectrophotometer is turned on and allowed at least an hour to stabilize. 2. Three sets of data are taken for steps 3 to 7, as follows. a. A black painted card is inserted between the source of illumination and the light source optics. b. The black painted card is removed. c. An ozone sample is pumped into the sample cell. 3. The wavelength drive is set to the minimum wavelength. 4. The computer is turned on and sits in an idler loop waiting for the first interrupt pulse. 5. The spectrophotometer is started and the analog of transmittance is sent to the computer together with its corresponding set of wavelength pulses. 6. Each time the data processing system detects a pulse, the analog voltage is measured, binarized and stored for further processing. 7. When the wavelength range of the prism has been scanned, the computer stops; a few seconds later the spectrophotometer stops and then automatically recycles itself. When the three successive runs of steps 2a-7, 2b-7, and 2c-7 have been completed, voltage level data for O %, 100 % and the gas sample 18

transmissivity will be stored in the computer. 8. The computer calculates, rounds and stores a transmissivity value of ~~~.value of = 1000 (E - 0) 100 - 0 for each set of 0, 100 E levels for each wavelength increment (0. 02 M-m) in the prism wavelength range; T(X) is a binary representation of a decimal integer. The set of T (k) data is punched on a paper tape which is marked for identification and stored as a permanent record. 9. A second tape is punched on which the T (X) data appear in binary-coded decimal form. These data are formatted for entry into the IBM 360/67 system for further processing. In the equation above, the numerator (E-0) is multiplied by 1000 to insure the greatest accuracy in T (X). The resulting double-precision product, when divided by a single-precision number, yields a singleprecision quotient and a remainder which is tested for rounding. And, since the rounded result is a binary integer, it can be converted exactly to its decimal equivalent. A cam-switch combination driven by a constant-speed motor usually produces pulses which vary in time between successive pulses. The error in wavelength produced by the pulses will be checked during the next work period. The spectrophotometer will be run through its wavelength range and, each time the computer detects a pulse, the seconds-milliseconds register of a time code generator will be read and printed. The time information will be analyzed and converted to probable error in incremented wavelength. 19

Preliminary operation of the spectrophotometer has shown that the 0 % and 100 %l voltage levels not only vary over the wavelength range but also that the levels are not repeatable for successive runs. This rules out the use of a single pair of 0 %, 100 % values for the reduction of a set of gas sample voltages; it also suggests that the use of one set of 0 %, 100 % values for the reduction of several successive sets of gas sample data could produce transmissivity data which will have appreciable error. A considerable effort will be made during the next work period to improve the stability of the ozone generating and handling system; and a number of runs using a polystyrene sample will be made. Analysis of the results should suggest methods for reducing the error in the transmissivity calculations. V. Report Writing The review paper "Radiation Processes and the Distribution of Radiative Sources and Sinks", by W. R. Kuhn, which was presented at the Symposium on the Dynamics of the Mesosphere and Lower Thermosphere of the AMS, June 15-18, 1970, Boulder, Colorado, by Prof. W. R. Kuhn is is being written as a project report. It should be completed in May 1971. VI. References 1. Gordon, H. R. and T. K. McCubbin, Jr., J. Mol. Spect. 18, 73 (1965). 2. Hearn, A. G., Proc. Phys. Soc. 78, 932 (1961). 3. McCaa, D. J., and J. H. Shaw, Sci. Rept. No. 2, Contract No. AF 19(628)-2633, The Ohio State University Research Foundation Columbus, Ohio, AFCRL-67-0237 (1967). 4. Walshaw, C. D., Quart. J. Roy. Met. Soc. 83, 315 (1957). 20

253'6 A 2893. A 2961A 3021A 3126A 3132A 3341A - _!........ ~.............:!-:.......;.... _ =___=____ —,~-, 1 11-, If.,.____ _________- r.r~ + 1l11==l- =====~======- - I a.1 11_ 1. 1,..-IIII.-1... -l —i3 J __sEB...,.,. I..... __.________________________* r-*. * lrlErlErrZ.,, __.iTt- _ _ _ _ _ _ _ _ _.~... _.... - r -: - TrrI__T r~~~~~~~~~ ~ ~~~~~~. r r X X X; ~:r;;* t _ _ a _ _ _ _ _ Fiur. Spcru of mrcr vapo.r lam shwn si *x __________________________II -r- r-r-r ~ine tha wri-rl l be use fo deemnto of __====_== r ~ ~ ~ ~ 0 amount;- |III__| _______ _ __=*gs.

#5 # #1 #2 14 35 70 12 30 60 ~,~ - 3oo30 1.2 12 30 60 1. -J ~ w ) 5025 50 5000 (") 40~20 o 6 15 30 — 30000150 B z 4 I0 20 -- -- 2000' I0.4 2~I --- 000 2 02)- 5C~~~~~~I 0' 0 01 - 01 01 i, I I -I I I I I I - I I I I I~ I ~o I% 2 3 4 5 6 7 8 910% 20 30 40 50 60 7080 100% 1/1o IN % Figure 2. Amount of 03(in Atm-cm) in 3200 cm. absorption cell vs. I/I0 (%) for six Hg lines.

UNIVERSITY OF MICHIGAN 3 9015 02223 2337