A spectrometer/radiometer now undergoing development is designed to be used aboard a spacecraft to measure the heights of cloud tops on Earth. The spectrometer/radiometer performs functions of three instruments - two spectrometers and a radiometer - that share a common field of view. Each of these instrument techniques implements a technique that has been used before, by itself, to measure the heights of cloud tops. By combining the three techniques in a single instrument package, the design of the spectrometer/radiometer makes it possible to determine cloud-top heights more accurately than can be done by use of one of the techniques alone. Moreover, the three techniques are complementary, so that at least one of the techniques can yield a useful measurement under conditions in which the other two techniques are deficient.
The three techniques are the following:
The spectral radiance of a cloud top viewed from above is measured at a wavelength of 11 µm, and the temperature of the cloud top is inferred from the measured radiance. Then the height of the cloud top is inferred from the climatological temperature-versus-altitude profile of the atmosphere, under the assumption that the cloud is in thermal equilibrium with the atmosphere. This technique fails in the presence of an isothermal atmosphere or of a convective cloud, which is not in thermal equilibrium with the atmosphere.
Molecular-Oxygen "A"-Band Absorption
Back-scattered sunlight is spectrally analyzed to determine the amount of absorption of light in oxygen molecules in the wavelength range from 750 to 780 nm. The depth of the atmospheric column above a cloud (and thus the height of the cloud top) is inferred from the differential absorptions in this wavelength range. This technique is vulnerable to errors in that the accuracy of the inferred cloud-top height depends on accurate correction of nonoxygen absorption at the cloud boundary layer.
Fraunhofer-Line Filling-in Effect (Ring Effect)
The Fraunhofer Ca, H, and K lines (which are absorption spectral lines in the solar spectrum) in the wavelength range of 390 to 400 nm are partially filled in by scattering in the terrestrial atmosphere due to the frequency shift in rotational Raman scattering. The measured spectrum of back-scattered sunlight in this wavelength range is compared with the corresponding extraterrestrially measured spectrum of light coming directly from the Sun to determine the amount of filling in of the Fraunhofer lines. Then the depth of the atmospheric scattering column above the cloud top (and thus the height of the cloud top) is inferred from the amount of filling in. This technique is not vulnerable to the boundary-layer inaccuracy of the molecular-oxygen "A"-Band technique, but its accuracy depends on high spectral resolution.
In the spectrometer/radiometer (see figure), wavelength dispersion is achieved by use of a diffraction grating. The spatial period and orientation of the grating are chosen to diffract the three wavelength bands of interest in different orders that emerge in different directions. Filters provide additional spectral selectivity for sorting the orders.
The 11-µm radiation is diffracted in the zeroth order of the grating, goes through a band-pass filter, and im-pinges on an HgCdTe photodetector cooled by liquid nitrogen. The 750-to-780-nm radiation is diffracted in the first order of the grating, goes through a long-wavelength-pass filter with a cutoff wavelength of about 550 nm, and impinges on part of a 1,024-pixel linear array of silicon photodiodes on a focal plane. The 390-to-400-nm radiation is diffracted in the second order, goes through a short-wavelength-pass filter with a cutoff wavelength of about 500 nm, and impinges on another part of the linear array of photodiodes. The spectral resolution is 0.4 nm in the first order and 0.2 nm in the second order.
This work was done by Hongwoo Park of Goddard Space Flight Center, Peter F. Soulen of the University of Maryland, and Coorg R. Prasad of Science& Engineering Services, Inc. GSC-14022