ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN ANN ARBOR TECHNICAL REPORT NO. 41 INVESTIGATION OF THE SPECTRAL ENERGY DISTRIBUTION FROM CLOUDS AND BLUE SKY IN THE NEAR INFRARED By KEITH PIERCE MCMATH-~HULBER T OBSERVATORY Project M720-5 OFFICE OF NAVAL RESEARCH, U. S. NAVY DEPARTMENT CONTRACT N60NR, TASK ORDER 5 February, 1953

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TECHNICAL REPORT NO. 41 INVESTIGATION OF TEE SPECTRAL ENERGY DISTRIBUTION FROM CLOUDS AND BLUE SKY IN TEE NEAR INFRARED SUMMARY In general, the infrared spectrum of the blue sky shows a Rayleigh-type scattering in the region 0.7 to 1.0 micron, Beyond 1 micron there is a large contribution of scattered radiation approximating that of pure sunlight. The energy distribution from clouds appears to be that of scattered sunlight with a slow uniform decrease in intensity from 0.6 to 2.6 microns. In addition, the water vapor bands are greatly strengthened except for cirrus-type cloud formations. The cloud brightness is from 60 to 200 times that of the blue sky at 1 micron. OBSERVATIONS Equipment A brief investigation is reported in which observations of the spectrum of clouds illuminated by sunlight, and of blue sky, were compared with sunlight reflected from white cardboard or an aluminum sheet. The spectral region covered was from 0.6 micron to 2.2 microns.

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN The observations were made with a Perkin-Elmer infrared spectrometer, Model 12-C, with a CaF2 prism. The instrument was attached to the tail piece of the F. C. McMath 24 inch diameter Cassegrain-type reflector. The large telescope fed an F/4 beam to the first slit and filled the aperture of the prism. The spectrometer was greatly modified so as to use an uncooled Cashman-type(l) lead sulfide receiver. This modification involved adapting the exit optics of the instrument so as to image the second slit on the receiver. By introduction of a sector and motor to chop the entrance radiation at 1080 cps, we were able to use a Wilson(2) AC amplifier whose output was fed to a smoothing filter and a Brown recorder. The first slit, 1 mm x 10 mm as projected through the telescope optics, corresponds to 1/40 x 1/4 of a degree in the sky, so that it was this area of 1/160 of a square degree of the sky or of a cloud that was observed. Observations of the Blue Sky On 5 or 6 days repetitive runs were made on the light scattered by the sky. This sky light has its origin in sunlight incident upon the earth's upper atmosphere. The scattered radiation is not intense, and it was necessary to use slits of 0.5 to 0.8 mm in width and a fairly high amplifier gain, which tended to lower the resolution. With such wide slits a good many of the finer features of the spectrum were lost entirely and the stronger water vapor bands filled in by the low spectral purity. Figure 1A shows a typical trace of the sky spectrum. The vertical scale gives the intensity and the horizontal scale the wavelength in microns. The intensity is a function of the response of the lead sulfide cell and the characteristics of the spectrometer. Since the sky light is derived from sunlight, it is logical to make a comparison between sky light and sunlight. Unfortunately, the comparison cannot be made directly by pointing the telescope first to the sun and next to the sky because of the great difference in intensity. Direct sunlight is too bright for the very sensitive receiver and amplifier employed. To reduce the light entering the spectrometer, sunlight was reflected from a white card or an aluminum sheet and observed with low amplifier gain and the same slit widths as were used for the sky observations. The reflectivity of the aluminum plate and white card should be practically constant within 5 percent over the spectral region concerned. Hence, the light reflected from the card or plate should have an energy distribution closely similar to that of direct sunlight but reduced in intensity. (1) Jour. Opt. Soc. Amer., 36, 356A, 1946 (2) Astrophysical Jour., 106, 246, 1947 2 -

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN From the work of Abbot and others we know that the energy distribution of sunlight approximates that of a black-body radiator at a temperature of about 6000~C. However, radiation reaching the observer is greatly modified by absorption of water vapor and other constituents of the atmosphere. Figure 1B shows the observed spectrum of sunlight and the presence of water vapor bands at 0.92, 1.1, 1.4, and 1.8 and 2.6 microns. The ratio of blue sky to sunlight at each wavelength is given in Figure 2A. We observe that if sky light were merely re-radiated sunlight the ratio would be unity for all wavelengths. The k4 line indicates the energy distribution to be expected from a scattering medium obeying the Rayleigh law, in our case the blue sky. This law applies to those cases in which the scattering particles are very small compared with the wavelengths of the incident light(3). The observations, as indicated by the points near the upper heavy solid line which defines a mean, follow very well the Rayleigh law from 0.7 to 1 micron, while from 1 to 1.6 microns they deviate very strongly in the direction of a large solar component. In this figure all observations have been reduced so as to pass through intensity unity at a wavelength of 1 micron. One observing day was characterized by a slightly milky, whitish sky and gave quite a different curve from that of a typical blue sky. This particular day is indicated by circles in Figure 2A. The total intensity of sky light for this run was much larger than that from a corresponding clear day and has been greatly reduced in plotting it on the figure so that it passes through the point (1.0, 1.0). The most interesting result from these observations is the fact that the sky is not nearly so black in the infrared beyond one micron as would be expected on the basis of the Rayleigh theory and its fit with observations at shorter wavelengths. It is logical to conclude that sky light is the sum of two components: (1) light from molecular scattering -4 obeying the X law and (2) light scattered and diffracted from larger dust and ice particles which in a first approximation re-radiate the normal energy distribution of sunlight. The work of H. Zanstra,(4) on halos and the aureole about the sun and moon, indicates that the scattered light is probably produced by small ice crystals and would extend in diminishing intensity over the complete sky. It also seems likely that the point of separation of the two curves at one micron would probably vary with the elevation of the observer and perhaps slightly with the purity of the sky, but as yet no observations to test this point have been made. (3) A more complete discussion of scattering in the atmosphere is given by H. C. Van de Hulst in Chapter 2 of The Atmospheres of the Earth and Planets, edited by Kuiper, 1949, in The University of Chicago Press. (4) Monthly Notices, RoyalAstron. Soc., 103, 265, 1943.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN It is also possible that the energy between 11 and 1,06 could be from scattered radiation from the surroundings of the instrument. However, several large modifications of the instrumental arrangement were tried with no difference in results. Therefore, it is felt that the observed energy is a real contribution from the atmosphere. In order to test this in another way, runs were made just before and just after sunset. The results are not conclusive because the energy at sunset is so weak, but they suggest that the sunlight component beyond 1 micron tends to disappear in the twilight sky. Observations of Clouds Figure 2B illustrates the energy distribution from clouds as compared with direct sunlight observed through the intermediary of reflection from a white card or an aluminum sheet. The considerable scatter of observations is due to the difficulty of following the cloud with the telescope and to its rapidly changing form and intensities. The cloud types that were observed were: (1) cumulus clouds, "face" illuminated with sun 90-180~ away; (2) stratus, illuminated mostly by transmitted light; (3) bright cloud edges illuminated by reflected and transmitted sunlight about 30~ from sun; and (4) dark cloud centers about 30~-90~ from sun. The figure shows that the energy distributed from clouds is essentially that of scattered sunlight. The only exception to this is the great additional absorption which appears in the water vapor bands at 0.8, 1.1, 108, and 2.0 microns in cumulus-type formations. A rough estimate of the intensity of cloud edges at 1 micron in comparison with the blue sky gives a ratio of about 50 to 200 to 1. This ratio varies greatly with the nature of the cloud and the mode of illumination and its distance in degrees from the sun. The work of Abbot 5) shows that for equal areas the sun is approximately 1.5 x 105 times as bright as light reflected from clouds. Several runs were made on clouds near the sun. When sufficiently thick they appear dark in the central regions and have brillianly illuminated edges. Again the center and edge repeat the sunlight distributions but the dark center may be 10 to 50 times less bright than the edges. This investigation originated with Director R. R. McMath of this Observatory and was carried out by the author. (5) Ann Astrophys, Obs. Smithsonian Inst., 2, 136, 1922

Fig. 2a Fig. Ioa Blue Sky October 19, 1952 2. ---— B- Sky High amplifier gain. Sllt 0.700 m m - W e. Low afer gai. Slt 0.175 mm o. Qo o1.4 __,. i,.e Wave Length Fig. lb Sunlight reflected from white card - Low amplifier gain. Sits 0700 mm -4 -. - Low amplifier gain. Si.t 0.175 mm 07 i. i.s to Wave Lengt

DISTRIBUTION LIST Dr. Richard C. Lord, Chairman Dr. W. B. McLean Panel on Infrared Naval Ordnance Test Station Department of Chemistry Inyokern Massachusetts Institute of Technology China Lake, California Cambridge 39, Mass. Dr. Joseph P. Casey Jr. Dr. George A. Morton, Civilian Member Physics Branch (CRRDP) Panel on Infrared Air Force Cambridge Research Center RCA Laboratories Division 230 Albany St. David Sarnoff Research Center Cambridge, Mass. Princeton, New Jersey Mr. Frank M. Field Dr. R. Bowling Barnes, Civilian Member Hdqtrs. USAF Electronics Division ORD-EL3 Panel on Infrared Directorate of Research and Development Olympic Development Co. Room 4C-331, Pentagon Building 77 Bedford St. Washington 25, D. C. Stamford, Conn. Mr. J. R. Rone Dr. Fordyce Tuttle Armament Division (AFDRD) Eastman Kodak Company Headquarters, U. S. Air Force Camera Works Directorate of Research and Development 333 State St. Room 4C-349, Pentagon Building Rochester 4, New York Washington 25, D. C. Dr. J. A. Sanderson Mr. George E. Brown Superintendent, Optics Division Engineering Research and Development Board Naval Research Laboratory The Engineering Center Washington 25, D. C. Fort Belvoir, Vao Dr. D. Bleil Mr. Harry Dauber Physics Research Division Applied Physics -Branch Naval Ordnance Laboratory Evans Signal Laboratory White Oak Belmar, New Jersey Silver Spring, Md. Mr. Henry T. Gibbs Mr. Charles E. Keener Electrical Branch Control and Guidance Division AEEL Research and Development Division U. S. Naval Air Development Center Office, Chief of Engineers Johnsville, Penna. Room 1038, T-3 Bldg. Washington 25, D. C.

Mr. Alva L. Brothers, Jr. (WCLGB) Office of Ordnance Research and Box CM Duke Station Dr. Paul J. Ovrebo (WCLRD-3) Durham, N. C. Wright-Patterson Air Force Base Attn. Physical Science Division Wright Air Development Center Dayton, Ohio Mr. G. A. Long, Jr. Bureau of Aeronautics (EL421) Mr. Norman Stulman Navy Department Office) Chief Signal Officer Washington 25, D. C. Special Projects Branch Engineering and Technical Division Mr. John Strong Room 2B-269, Pentagon Building The Johns Hopkins University Washington 25, D. C. Baltimore 18, Mdo Col. F. F. Wilkins Mr. C. S. Woodside Office, Chief of Army Field Forces Bureau of Ships (853) Research and Development Section Navy Department Fort Monroe, Va. Washington 25, D. C. ONR/Pasadena Mr. W. Otting Attn' Arnquist (2) Physics Division Air Research and Development Command ONR/London P. 0. Box 1395 Attn: Hyde (2) Baltimore, Md. ONR/Chicago ONR/San Francisco Capt. C. A. Pappas, USAF Air Research and Development Command (RDDRE) P. O. Box 1595 Baltimore, Md. Mr. S. Nutting Bureau of Ordnance (Re4e) Navy Department Washington 25, D. C. Mr.o K. Knight Bureau of Ordnance (Re8) Navy Department Washington 25, D, C.