THE UNIVER S I T Y OF M I C H I G A N COLLEGE OF ENGINEERING Department of Aeronautical and Astronautical Engineering High Altitude Engineering Laboratory Technical Report LONG-TERM INTEGRITY OF THE TIROS 5-CHANNEL RADIOMETER VISIBLE CHANNEL CHARACTERISTICS F. L. Bartman Mo To.Surh M. G. Whybra ORA Project 03615 -under contract with: NATIONAL AERONAUTICS AND SPACE ADMINISTRATION GODDARD SPACE FLIGHT CENTER CONTRACT NO. NASw-140 administered through: OFFICE OF RESEARCH ADMINISTRATION ANN ARBOR December 1963

:

TABLE OF CONTENTS Page UNIVERSITY OF MICHIGAN CALIBRATION PERSONNEL v LIST OF TABLES vii LIST OF FIGURES ix ABSTRACT xi 1o INTRODUCTION 1 2. DESCRIPTION OF THE RADIOMETER 2 3. METHOD OF CALIBRATION 3 4o THE EFFECTIVE SPECTRAL RESPONSE 7 5. THE RADIANT EMITTANCE OF THE SOURCE 8 6. THE "EFFECTIVE" RADIANT EMITTANCE OF THE SOURCE 10 7. RESULTS 11 8. DISCUSSION AND CONCLUSION 18 9. ACKNOWLEDGMENTS 33 10 REFERENCES 34 iii eDL

UNIVERSITY OF MICHIGAN CALIBRATION PERSONNEL F. L. Bartman, MoA., Research Engineer Lo W. Chaney, BoA., Research Engineer V. Bo Devlin, Assistant in Research M. T. Surh, MoA., Assistant Research Engineer M. G. Weimeister, Technician M. G. Whybra, M.Ao. Associate Research Engineer E. Ao Work, Jr, o B.S., Assistant Research Engineer *Now at Bendix Systems Division, Ann Arbor, Michigano v

LIST OF TABLES Table Page Io Nominal Characteristics of TIROS 5-Channel Radiometer Channel 3 and Channel 5o 2 IIo Channel 3 Calibration Characteristic Datao 12 III. Channel 5 Calibration Characteristic Data. 13 IVo Calibration Results for Channel 3 (Uncorrected for Temperature Dependence)o 14 Vo Calibration Results for Channel 5 (Uncorrected for Temperature Dependence)o 15 VI. Radiometer Temperature During Calibration Runs. 15 VIIo Calibration Results for Channel 3, Normalized to 30~Co 16 VIIIo Calibration Results for Channel 5, Normalized to 31~Co 17 IXo Uncertainty of the Calibration Datao 19 vii

LIST OF FIGURES Figure Page lo Diagram indicating principle of operation of a channel of the TIROS 5-channel radiometero 20 2. Set of calibration curves obtained at several radiometer temperatures 21 35 Schematic diagram of calibration arrangemento 21 4. Relative spectral radiance of GoE, 30A/T24/3 standard lamp Noo 62710 22 5o Relative luminosity curve for a standard observer. 23 6. Diffuse reflectivity of Kodak white paper. 24 7. Effective spectral response of channels 3 and 5 of TIROS 5-channel radiometero 25 8. White paper "effective" spectral radiant emittance, kr\krX., for TIROS 5-channel radiometer channels 3 and 5o 26 9. Temperature dependence of calibration data for channel 3, bulb voltage vso radiometer temperatureo 27 10o Temperature dependence of calibration data for channel 3, bulb voltage vs radiometer voltage 28 11o Temperature dependence of calibration data for channel 5, bulb voltage vso radiometer temperature 29 12. Temperature dependence of calibration data for channel 5, bulb voltage vs radiometer voltage 50 135 Channel 3 calibration data, normalized to 300C. 31 14. Channel 5 calibration data, normalized to 310~C 32 ix

ABSTRACT The integrity of the visible channel calibrations of the TIROS 5-channel radiometer unit No. 103A is discussedo During a 16-month period, the radiometer experienced approximately 75 hours of near-space environment on balloon flights of 22-miles altitude and in environmental test chambers simulating conditions at 22 mileso Data from three sets of calibrations made during this time are shown to agree within the error of the measurements. The theory and method of making the calibrations are discussedo xi

1. INTRODUCTION The meteorological satellites TIROS II, III, IV, and VII have each used a Barnes Engineering Company 5-channel Radiometer1 for the measurement of the thermal radiation emitted by the earth and its atmosphere and of the solar radiation reflected and scattered by the same agentso2 A radiometer of the same type has been used by The University of Michigan's High Altitude Engineering Laboratory for similar measurements on balloon flight tests at an altitude of approximately 115,000 feet. One of the most important factors in the interpretation of the data measured by a radiometer is a knowledge of its operational characteristics, a thorough understanding of which is best obtained by frequent, careful calibrations. Ideally these calibrations would be made under the same environmental conditions experienced by the satellite-borne instrument. Also it is desirable to make calibrations both before and after flight data are obtained, so that possible changes in instrument characteristics can be detected~ The requirement of calibration under proper environmental conditions was certainly met in the preparation of the radiometers for the TIROS flights,3 however, post-flight calibrations are impossible for the satellite work since the radiometer is not recovered. The radiometer on the balloon flight tests can provide some knowledge of the integrity of the instrument characteristics after it has been used for some time under near-space conditions~ The TIROS 5-channel radiometer No. 103A was flown on balloons four times between May 1961 and September 1962o In addition, the radiometer was operated under simulated balloon-flight environmental conditions four times during this periodo The total time of operation under near-space conditions during this period approximates 75 hourso Calibrations were made in May 1961, in August 1961, and in September 1962 at the NASA Goddard Space Flight Center; additional calibrations were made at The University of Michigano The purpose of this report is to discuss a portion of these calibration data to see what they reveal about the long-term integrity of the characteristics of the visible channels of this radiometero Future reports will consider the visible channel calibrations at Michigan, the thermal channel calibrations at NASA and Michigan, and will attempt to establish the absolute accuracy of all of these calibrationso 1

2o DESCRIPTION OF THE RADIOMETER The principle of the TIROS 5-channel radiometer is illustrated schematically in Figo 1, showing the basic components of a single channel of the radiometero The field of view of each channel is about 5~ wide at the half power-point (the field of view down to 5% of maximum response is 8 to 9~). The chopper disk-prism arrangement causes the radiometer to look alternately in opposite directionso During flight and during calibration, a reference target is viewed in one direction, and the earth or calibration target in the opposite direction. The reference target is outer space during flight and is a liquid nitrogen-cooled black body during calibration. As the half mirrorhalf black chopper disk rotates, the radiation reaches the detector alternately from opposite directions. The resulting chopped signal is amplified and rectified to produce a DoC. signal, which in the ideal instrument is proportional to the difference in energy flux received from the two directionso The two viewing directions are called "wall" and "floor" side of the radiometer according to their location in the TIROS satellite, i.e., one view is through the wall side, the other through the floor side of the satellite. The nominal characteristics of the two visible channels, channels 3 and 5, with whose calibration we are concerned in this report, are given in Table I. TABLE I NOMINAL CHARACTERISTICS OF TIROS RADIOMETER CH L AD HANNEL AND CHANNEL 5 Channel Number 5 Purpose of "Reflected" Solar Comparison Measurement Radiation with TV-Photos Wave -length (Nominal).2-6 microns ~55-.75 microns A 203 AX2a03 Optics (lenses) Bo SO2 BoF2 SiO2 Interference Filters --- and Chance Glass 2

3. METHOD OF CALIBRATION The method of calibration used in the NASA calibrations of the visible channels of the TIROS 5-channel radiometer is referred to in the literature as the "Near-Extended Source Method."4 A diffuse* source, larger than the radiometer aperture, which will completely fill the radiometer field of view, is placed within a short distance of the aperture so that the transmissivity of the medium between the source and the atmosphere is essentially unity. The equations which apply can be written as follows. Assuming that: a. The source is an ideal diffuse source of known spectral radiance Nc. b. The transmissivity of the medium between the source and radiometer is unity. The characteristics of the calibration source are given by NC = / NKcd we "o W = WxcdX o Ltts-cm-2-steradian-1 watts-cm-2 (la) (lb) and the relations Nc = Xc -Wc \Jhe~~~T (2a) (2b) Where: Nc WXc Nc is the spectral radiance, watts-cm-2-steradian-l-micron-1 is the spectral radiant emittance, watts-cm-2-micron-1 is the radiance, watts-cm-2-steradian-1 Wc is the radiant emittance, watts-cm-2 sense of a Lambert radiator! *Diffuse in the 3

The irradiance of the radiometer aperture is He = n NI0 dX watts-cm-2 (3) where 2 is the solid angle viewed by the radiometer. The total radiant power at the radiometer aperture is HcAr = ArV. NX dcX watts (4) where Ar is the effective area of the radiometer aperture. If the response of the radiometer is constant over its entire field of view, the voltage output of the radiometer is given by Vc = R~Ar' ~ /X NXc.dX volts (5) O where R is the responsivity constant of the detector in voltsowatt-1, and B is a quantity called the effective spectral response of the instrument, with maximum value < 1, which accounts for the spectral response of the detector, the transmission of optical components such as filters and lenses, and the reflectivity of mirrors. We can write the last equation as Vc = RoAr~.Nc (volts) (6) where 00 N' = f NXc/k~d watts-cm-"2 (7) ~~~o ~Steradian"' is said to be the "effective" radiance of the source, i.e., the.radiance of the source modified by the spectral response characteristics of the instrument In the practical case in which the response of the radiometer is not uniform over the entire field of view, we can write the voltage output of the radiometer as Vc = ArJNC R(cu).d) (8).a 4

where R(c) is the responsivity of the radiometer as a function of the solid angle w. Even in this case we can write Vc = R^'Ar~oQNc (9) where R' is the responsivity of the detector averaged over the entire field of view, i.e., R(w)dwd R' = J___ (10) Thus the equation which describes the voltage output of the radiometer is (9) above, or, for this "diffuse' calibrating source, Vc = R'-Ar W (11) Where WI is the "effective" radiant emittance of the diffuse source, i.e., the spectral radiant emittance of the source modified by the spectral response characteristics of the instrument, ioeo, 00 W J W= AdX^ (12) The calibration procedure for the visible channels of the radiometer unit No 103A consisted of the following steps: 00 co a. Vary WC Wcd?\ of the diffuse source of known spectral distributiono bo Record corresponding values of Vc and Wc. c. For a given value of Wc, calculate a corresponding value of WI. This can be done using Eq. (12) above if WXc and ~H are known for the diffuse calibrating source and the radiometer respectively. d. Plot curves of Vc vso Wc, or Vc vs. WI (the latter is used in this report)o These curves are the "calibration curves" of the radiometer for a diffuse target of the given spectral radiant emittanceo Since the calibration curves of the radiometer are not completely independent of the temperature of the

radiometer itself, there is a set of such calibration curves covering the range of temperatures over which the instrument is expected to operate, see Figo 2o The calibration at room temperature was made with a "standard" lamp as a source and is the absolute calibration. The possible changes of the calibration with radiometer temperature was investigated with another lamp, and only relative values were obtained~ Ideally, the instrument calibrated as indicated in this section would be used only with diffusely radiating unknown targets of the same spectral radiant emittance Wic as the calibrating source, in which case the use of the calibration curves is quite straightforward. The interpretation of the data in the case in which the target is not diffuse or does not have the spectral distribution of the calibrating source is more difficult and will not be discussed further hereo The usefulness of the instrument depends on the calibration curves remaining constant over long periods. This stability is the subject of the reported investigation. 6

4. THE EFFECTIVE SPECTRAL RESPONSE The effective spectral response of a channel of the radiometer is given by: = Rm(Rbm-R f R.(Rx-RxI).fX.^A (13) where Rp is the spectral reflectivity of the radiometer prism RA is the spectral reflectivity of the aluminized half of the chopper disk b R is the spectral reflectivity of the black half of the chopper disk f% is the spectral transmissivity of the filter-lens portion of the given channel Ax is the spectral absorptivity of the thermister detector. In the actual computation of By, the expression y = Bl-R)f ~x - Rx(l-Rx).fX (14) M was used, thus assuming that RN = 1 and A = 1. The data used to compute By for the visible channels of the radiometer are given in Section 7. 7

5. THE RADIANT EMITTANCE OF THE SOURCE Figure 3 is a schematic diagram of the calibration arrangement. The diffuse calibration source was produced by reflection of light from white Kodak paper. The source of light was a type G.E. 30A/T24/3 lamp No. 6271, calibrated by the National Bureau of Standards and at the Naval Research Laboratories. The apparatus consisted of a calibration rack containing fixed mounts for the standard lamp and the radiometer. Two white Kodak paper targets (of fixed radii of curvature 42.0 cm and 90.0 cm) could be placed at various distances from the lamp in order to vary the intensity of light reflected to the radiometer. The target of 42 cm radius of curvature was used at distances of 45.7, 48.2, 50.7, 55.2 and 55.7 cm from the lamp; the 90.0 cm radius of curvature target was used at 55.7, 57.7, 62.7, 72.7, 82.7, 92.7 and 102.7 cm from the lamp. During the calibration measurements the standard lamp was operated at 15.30 amperes and 94.60 volts (RMS). The voltage was measured during the calibration with an NLS digital voltmeter with A.C. converter. Voltage fluctuations of approximately 0.5% were observed. The radiometer output voltage was measured with a Hewlett-Packard 410-B voltmeter. Calibration data supplied by NASA personnel for the standard lamp consisted of: a. The Bureau of Standards data; i.e., the luminous intensity of the lamp is 3455 candles under the above operating conditions. b. Acurve of relative spectral radiance 4X (in arbitrary units) obtained by Naval Research Laboratory personnel (see Fig. 4) for the same operating voltage. From this data the spectral radiant intensity of the source, can be calculated as follows. The luminous intensity of the lamp is given by 00 I = 685 k Xkxd = 3455 candles (15) where I is the luminous intensity in candles (lumens-steradians-1) X is the wavelength, in microns 8

685 is the luminous efficiency of monochromatic flux of wavelength 0.555 microns (lumens-watt-1) *x is the relative spectral radiance, in arbitrary units kk is the relative luminosity of a standard observer for good lighting conditions (see Fig. 5), i.e., the luminous efficiency at wavelength X is 685 k. lumens/watt k is a constant Thus if we solve for k from Eq. (15), we find k = 5455 685 ~ XkXdX% 0 (16) and the spectral radiant intensity of the NBS standard lamp is J% = kxK watts-steradian' -micron 1 (17) The value of k obtained for the standard lamp No. 6271 obtained in this fashion is 39.084. Since J% max for the spectral radiant intensity of the NBS standard lamp occurs at a wavelength Amax of 0.91 micron, the spectral distribution of the lamp is approximately that of 31700K black body (see Fig. 4 for comparison). The Kodak white paper is curved and the lamp located so that the radiation from the lamp is very nearly normally incident upon the paper. Under these conditions, the spectral radiant emittance of the calibration source (the Kodak paper) is Wk c = r2 Xc - R2 watts-meter -micron (18) Where r\ is the diffuse reflectivity of the white Kodak paper (Fig. 6) and R is the distance (in meters) from the standard lamp to the white paper. The radiant emittance of the calibrating source is therefore 00 Wc = J XrX dX W Jc R2 00 = k *_rz.) R2 oO watts-meter-2 watts-meter-2 (19) (20) 9

6. THE "EFFECTIVE" RADIANT EMITTANCE OF THE SOURCE The "Effective" radiant emittance of the calibration source was given by Eqo (12) above, ioe,, W - J WXcdX t watts -meter 2 where Wc and By are given by Eq. (18) and (14) above, respectively. The integration was performed numerically, by summing the contributions to the integral from each small increment of wavelength A1i, thus W = Z (WC) ix~iA c i R2 i 1iri iAi watts-meter-2 (21) 10

7. RESULTS The data used to specify the calibration characteristics of the TIROS 5 -channel radiometer No. 103A are given in Table II for channel 3, and in Table III for channel 5. *r and rX are plotted in Figs. 4 and 6, as noted above. By and the product k*XrX$k (the "effective" spectral radiant emittance of the Kodak white paper source) are given in Figs. 7 and 8 respectively for both channels. The summation indicated in Eq. (21) was carried out, and the "effective" radiant emittance of the calibration source was found to be (Channel 3) W = 74 watts per meter2 (22) (Channel 5) W = 4.6 watts per meter2 (23) c R The calibration data, i.e., values of Vc versus Wc for 14 values of R taken with the radiometer at room temperature are given in Table IV for channel 3, and in Table V for channel 5. Complete sets of data were taken in May and August of 1961, and spot checks were made in September 1962. Radiometer temperatures during the three calibration runs are shown in Table VI. Figure 9 shows the temperature dependence of the calibration data (relative calibration curves) for channel 3, wall side; for convenience, a portion of the data is re-plotted in Fig. 10. Similarly, the temperature dependence of the calibration curves for channel 5, wall side, is shown in Fig. 11, with a portion of the data re-plotted in Fig. 12. The data of May 1961 have been corrected to the radiometer temperature of the August 1961 and September 1962 data. For channel 3 the data were corrected to the radiometer temperature of 30~C. For channel 5 the data were corrected to the radiometer temperature of 31~C. The corrected calibration results are shown in Tables VII and VIII respectively, and are plotted in Figs. 13 and 14 respectively. The May 1961 and August 1961 calibrations agree within the accuracy of the measurements. The September 1962 data show somewhat more scatter but are also essentially in agreement. 11

TABLE II CHANNEL 3 CALIBRATION CHARACTERISTIC DATA Wavelength Microns.25.26.28.31.41.51.61.71.81.91 1.01 1.21 1.41 1.61 1.81 1.91 2.01 2.21 2.41 2.61 2.81 3.02 3.50 4.10 4.22 4.34 4.46 4.58 f % Filter-lens Transmis s ivity.5155.6704.7592.7799.7887.7980.7999.8021.8026.8083.8085.8133.8135.8183.8239.8231.8233.8225.8188.8329.8238.8240.8420.8259.8261.8214.7984.7789 1-RB Chopper Reflectivity.240.371.642.788.788.852.836.795.730.750.782.812.830.850.860.870.875.890.905.86.86.87.88.89.90.90.89.89 RP Prism Reflectivity.619.619.636.756.870.870.800.778.750.768.800.834.851.868.877.886.875.890.905.860.86.87.88.89.90.90.89.89 Effective Spectral Response.077.154.310.465.541.592.535.496.439.466.506.551.575.604.621.634.630.652.671.616.609.624.652.654.669.665.632.617 r% k*% k*%A/ArA White Paper Lamp Spectral White Paper Effective Reflectivity Radiant Intensity Spectral Radiant Emittance.60.60.63.66 -.78 5.081 2.144.88 24.232 12.601.88 77.386 36.449.88 113.344 49.531.897 127.414 50.201.90 132.104 55.352.89 126.241 56.808.88 106.699 51.749.83 85.203 40.642.70 66.052 27.94.75 49.637 23.131.71 50.028 22.513.66 41.82 17.397.60 32.831 12.837.49 30.876 10.158.46 23.45 6.636.39 9.38 2.232.32 ---........._........._ _ _...... _...

TABLE III CHANNEL 5 CALIBRATION CHARACTERISTIC DATA Wavelength Microns.500.525.550.575.6oo.625.650.675.700.725.750.775.800 F-1 \^ f% Filter-lens Transmissivity.005.o60.241.459.532.350.342.358.250.245.137.010.000.003.016.008.013.037.062.062.o6o.066.074.079.oo6.000 1-RB Chopper Reflectivity.854.856.854.850.842.833.828.815.805.799.783.767.752.819.824.832.852.862.876.840.850.860.868.870.865.870 RP Prism Reflectivity.860.845.830.814.798.782.767.748.732.887.892.895 i^ Effective Spectral Response.0037.043.171.318.223.228.217.218.147.174.096.0069.0022.012.0059.0093.0220.041.044.043.049.056.o6o.oo45 rA White Paper Reflectivity.88.88.88.88.88.88.88.88.88.887.892.895 k*t Lamp Spectral Radiant Intensity 21.494 29.455 41.794 55.726 71.647 85.977 98.714 107.471 113.441 117.82 122.198 125.781 121.402 117.82 109.859 87.967 68.065 50.153 43.386 33.037 31.445 24.678 17.514 9.951 2.786 White Paper Effective Spectral Radiant Emittance.0633 1.099 6.156 15.321 13.789 16.969 18.512 20.261 14.481 17.816 10.199.741.238 1.167.539.691 1.003 1.477 1.236.843.741.630.447.0195 1.10 1.13 1.20 1.40 1.60 1.80 2.00 2.20 2.40 2.60 2.70 2.80 2.90.89.89.88.84.69.75.84.85.860.868.875.865.865.89.89.88.84.69.75.67.60.50.46.43.39.36

TABLE IV CALIBRATION RESULTS FOR CHANNEL 3 (Uncorrected for Temperature Dependence) Floor Side Wall Side Wc May, Aug. Sept, May Aug. Sept. (watt-m-2) 1961 1961 1962 1961 1961 1962 Volt s Volt s 319o6 6.02 5.86 5.94 5.58 5.08 516 287.3 5 46 5.46 -- 4 4.52 259.7 4.98 490 — 4.54 4.24 4.24 235.8 454 4.44 -- 4.20 3.96 215.2 4.18 4. 16 4.00 3.78 3.68 3.64 219.0 4.28 4.14 -- 3582 3 66 3.44 200.5 3592 3582 -- 56 5328 5.18 169.8 3.38 326 -- 318 2.90 145.6 2.88 2o84 2o64. 274 2.46 2.52 126o 3 2 54 2 46 2.42 2.12 11o.6 2.28 2.o18 -- 212 1o88 1.78 97.6 1i86 1.82 -- 174 1 62 77.7 1.62 1.58 1.44 1046 o1.5 1,52 63. 3 1o32 1.25 -- 1.20 1.10

TABLE V CALIBRATION RESULTS FOR CHANNEL 5 (Uncorrected for Temperature Dependence) Floor Side Wall Side Wc May Aug. Sept. May Aug. Sept. (watt-m-2) 1961 1961 1962 1961 1961 1962 Volts Volts 22.53 2.90 260 2.90 2.94 2,64 2.78 20.0 2,64 2.36 -- 2.64 2.38 -- 181 2.54 2.14 -- 2.42 2.14 2.36 16,4 2.16 1.90 -- 2.20 1.92 -- 15.0 2.00 1.72 1l96 2.00 1.72 1.94-1.98 15.5 1.98 1.76 -- 2.10 178 — 14.0 1.84.60o -- 1.94 1o72 1.74 11.8 1.56 1.38 -- 1.64 1.44 10.1 16 1.20 1.20 1.40 1220 1.26 8.8 1.18 1.06 -- 1.22 1.08 -- 7.7 1.04.90 1.10.92.98 6.8.86.76 --.90.76 -- 5.4.70.62.68.74.64.64 4.4.58.52 --.60 ~52 TABLE VI RADIOMETER TEMPERATURES DURING CALIBRATION RUNS Channel 5 Channel 5 Wall Side Floor Side Wall Side Floor Side Date Begin, End, Begin, End, Begin, End, Begin, End, ~C C 0~CC C ~ C ~ C 0C C~ May, 1961 25.8 27.5 28.5 29.0 28.4 27.6 29.5 29.2 Aug., 1961 29.8 50.5 31.1 51.0 30.7 51.1 31.0 31.1 Sept., 1962 29.1 29.9 50.6 50.6 50.0 50.2 506 50o.6 15

TABLE VII CALIBRATION RESULTS FOR CHANNEL 3 AT 30~C (Corrected for the Radiometer Temperature) Wt (watt-m-2) 319.6 287.3 259.7 235.8 215.2 219.0 200.5 169,8 145.6 126.3 110o6 97.6 77.7 63.3 May, 1961 5.30 4.79 4.31 3~99 3~59 3.63 3~38 3,02 2,60 2,30 2.01 1 65 1 39 1,14 1.14 Wall Side Aug., 1961 Volts 5,08 4,52 4.24 3.96 3.68 3.66 3028 2.90 2.46 2,12 1,88 1.62 1033 1.10 Sept., 1962 5.16 4.24 3.64 3.44 3.18 2,32 1.78 1,32 16

TABLE VIII CALIBRATION RESULTS FOR CHANNEL 5 AT 30 C (Corrected for the Radiometer Temperature) Wc (watt-m-2) 22o3 20.0 18,1 16,4 15.0 15.5 140o 11,8 10.1 8.8 7~7 6,8 5.4 4.4 May, 1961 2,72 2,44 2,24 2,04 1.85 1,80 1052 1030 1,13 1,02 o.83.68.55 Wall Side Aug., 1961 Volt s 2.64 2038 2,14 1.92 1.72 1,78 1.72 1,44 1,20 1,08,92,76.64.52 2.78 2.36 1.96 1074 1,26.98.64 Sept,, 1962 17

8. DISCUSSION AND CONCLUSION The accuracy of this evaluation of the long-term integrity of the channel 3 and channel 5 calibrations of the TIROS 5-channel radiometer depends on the ability to repeat and measure precisely the calibration source characteristics, the radiometer operating characteristics and the radiometer output voltage The factors which affect the repeatability of the calibration source characteristics are the stability of the light output of the lamp, the stability of the reflectivity characteristics of the Kodak white paper, the repeatability of the lamp to white paper distances and the lamp operating voltage The lamp is operated at about 85% of its normal operating voltage and thus should have a lifetime (500 hours?) much longer than the time it was actually usedo Although it is assumed that the lamp characteristics at the given voltage did not change during the 16 month interval over which the calibrations were carried out, it is felt that this assumption has not been checked well enough and that further use of this calibration apparatus would warrant additional calibrations of the characteristics of the lamp itselfo The reflectivity of the Kodak white paper can vary somewhat depending on the nature of the accumulation of surface dusto Similar reflecting surfaces of magnesium oxide show a 2% decrease in reflectivity after a short time of use, followed by a rather long period of stable reflectivity characteristicso We assume *the same behavior for the Kodak white paper target and thus assume a maximum uncertainty of 2% in reflectivity characteristicso The lamp voltage and lamp to white paper distance are repeatable within ~ 05% and so contribute only a small uncertainty to the resultso The radiometer operating characteristics should repeat if all radiometer power supply voltages are kept at their proper valueso Precautions are taken to monitor these voltages at intervals before and during the calibration runs and critical voltages are regulatedo The radiometer output voltage is read with a maximum error of ~ 5% of full scale Table IX summarizes the estimates of the uncertainty which these various factors contribute to the calibrations. If these possible sources of error combine randomly, the net uncertainty would be ~ 4%. 18

TABLE IX UNCERTAINTY OF THE CALIBRATION DATA Possible Source of Error Uncertainty Lamp Characteristics- - Lamp Voltage ~ 0.5% White Paper Reflectivity ~ 2o0% Lamp to White Paper Distance ~ O05% Radiometer Operating Voltages -- Radiometer Output Voltage '~ 3% The data in Figso 13 and 14 are seen to have scattered less than this amount and so it must be concluded that, under laboratory conditions the calibrations of channel 3 and channel 5 of the TIROS 5-channel radiometer maintained their integrity over this 16 month interval within the limits of the uncertainty in the measurements (~ 4~%)o 19

/ / / J Thermistor Bolometer -i / Prism / / Lens / / Filter / / a / / / flack / / Paint Rotating Chopper Disc Fig. 1. Diagram indicating principle of operation of a channel of the TIROS 5-channel radiometer. 20

W' C vc Radiometer Fig 2. Set of calibration curves s radiometero temperature s - White Kodak Paper ( -- standard Lamp Radiometer Fig 3 Schemati diagram f calibration arrangement

4.0 +(X) -3180~ K Black Body 2.0 PO- O / 0.2 1.0 2.0 3.0 Wavelength, microns Fig. 4. Relative spectral radiance of G.E. 30A/T24/3 standard lamp No. 6271.

1.0 - 0.8 0 e~, 0.6 -J 0.2 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Wavelength (zI) Fig. 5. Relative luminosity curve for a standard observer

100 90 80 > 70 60 4 -50 t <, 40 - \ Q) 30 20 10 0.2.4.6.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 Wavelength, Microns Fig. 6. Diffuse reflectivity of Kodak white paper.

1.0 r CD a) o. () a, La) 0.8h Channel 3 0.6 0.4 0.2 I I I I I I I 1.0 2.0 3.0 4.0 5.0 6.0 7.0 X, Wavelength, microns 0. or 0.1 COL a) -X,.,0.4 a Ln Lul. Channel 5 1.0 2.0 3.0 4.0 5.0 6.0 7.0 X, Wavelength, microns Fig. 7. Effective radiometer. spectral response of channels 5 and 5 of TIROS 5-channel 25

60 50 40 h< -es 30 20 10 Channel 3 1.0 2.0 3.0 4.0 5.0 6.0.7.0 X, Wavelength, microns 30 ^20 10 Channel 5 1.0 2.0 3.0 4.0 5.0 6.0 7.0 X, Wavelength, microns Fig. 8. White paper "effective" spectral radiant emittance, TIROS 5-channel radiometer channels 3 and 5. k*%jkrx, for 26

Channel 3 (. 2-5. 5 u) Wall Side VQV VRAoD 12 0 15 1 0 10. 0 _^ — 9&0 8.0 6.0 ' 10 5.0 4.0 0 3.0 20 0 -1~S -Q _I~ r~a I Aa ~~k- ~~-~E 1.0 C0, 5 n ~ ~ _osd~~ -pl~ S_ — w~ ~~~~ -10 0 10 20 30 40 Radiometer Housing Temperature, ~C Fig. 9. Temperature dependence of calibration data for channel 3, bulb voltage vs. radiometer temperature. 27

15 10 - 5 Channel 3 (. 2-5.5 j) Wall Side Temperature May,'61 Data 45. 4 ~C I I,.._ I I I I 2 4 6 6 10 12 Radiometer Voltage Fig. 10. Temperature dependence of calibration data for channel 3, bulb voltage vs. radiometer voltage. 28

20 Channel 5 (. 55 —. 75 ) VaAIWall ] Side O,RAD 12.0V OC I 4~1.0 7, 0 1 6. 0,,,-5.0 == 1 ~4.0 100 l 3.0. 20 _ 01.0.0 20 30 40 ~Fig, 11.~ @Radiometer Housing Temperature, C voltge v. Teraturae deeneeof calibration dat c, b 29

20 Channel 5 (.55. 75 p) Wall Side May,'61 Data Housing Temperature 31. 70C 5.5~C I / / /27. 9 ~C 15 - CD C" -, 0 -. m:::: 10 2 4 6 8 10 12 Radiometer Voltage Fig. 12. Temperature dependence of calibration voltage vs. radiometer voltage. data for channel 5, bulb

800 Channel 3 (.20 5. 5x) Wall Side 700 g700 o May 1961 o August 1961 A September 1962 600 1 500 on ^ " 400 300 _ 200 1 00 0 2 3 4 5 6 7 Radiometer Voltage Fig. 13. Channel 3 calibration data, normalized to 30~C. Fig. 15. Channel 5 calibration data, normalized to 30~C. 31

80 Channel 5 (. 55 -. 75 u) Wall Side o May 1961 o August 1961 70 A September 1962 60 E COn / 50 -,/, LA r 40._4 Cu 30 20 10 -, I I I.. 1. I 1 1 0 2 4 6 8 10 12 Radiometer Voltage Fig. 14. Channel 5 calibration data, normalized to 31~C. 32

9o ACKNOWLEDGMENTS We are deeply indebted to the Aeronomy and. Meteorological Satellite Branch of NASA for financial support, for the use of their equipment in performing these calibrations, and for the kind assistance of their personnel in helping to set up the equipment and provide supplieso 33

10o REFERENCES lo Astheimer, R Wo, R DeWaard and Eo Ao Jackson, "Infrared Radiometric instruments on TIROS IT," Jo Opto Soco Amer, 51, pp 1386-1393, 1961 2o Bandeen, Wo Ro, Ro A. Hanel, Jo Licht, R. Ao Stampfl and Wo Go Stroud, "Infrared and. Reflected Solar Radiation Measurements From the TIROS TI Meteorological Satellite," J. Geophys Res., 65, pp 3169-3185, 1961. 3. National Aeronautics and Space Administration and United States Weather Bureau, "TIROS II Data Users' Manualo" NASA Goddard Space Flight Center, Greenbelt, Mdo, 15 August 1961, (Available upon request from the Aeronomy and Meteorology Division, NASA Goddard Space Fli ght Center)o 4o Holter, M. Ro, S. Nudelman, G, Ho Suits, Wo Lo Wolff, Go Jo Zissis, "Fundamentals of Infrared Techno logy,' p 54o The McMillan Coo, 1962o 34

UNIVERSITY OF MICHIGAN 3 9015 02223 2402