The determination of spectral signatures for scenes with a high degree of spatial complexity is considered as one of the most persistent problems in atmospheric radiation, especially at the surface, where satellite observations can only be used indirectly to infer energy budget terms. In the shortwave (solar) spectral range, it is especially challenging to derive consistent albedo, absorption, and transmittance from spaceborne, aircraft, and ground-based observations for inhomogeneous cloud conditions and is closely related to the long-debated discrepancy between observed and modeled cloud absorption.
The cloud spatial structure is revealed as a spectral signature in shortwave irradiance through the physical mechanism of molecular scattering. However, the study of specific mechanisms is rather complex since the satellite instruments cannot completely describe the spatial distribution of cloud and the variability of scattering and absorption properties. For this reason, several studies deal with the problem described above, as a challenge for estimating spectrally the cloud optical properties (such as the albedo and transmittance) as well as scattering and absorption processes taking place in the cloud system with adequate resolution. Hence, the above mechanisms can be described using three dimensional (3-D) radiative transfer models. Those models receive auxiliary information from cloud imagery and radar observations. The molecular scattering (Rayleigh) was the only one directly dependent on the wavelength of the vertical radiative flux. Moreover, it was considered as a spectral perturbation of backtracked horizontal exchange of solar radiation due to the inhomogeneous distribution of cloud. The horizontal photon transport is highly correlated to its spectral dependence.
Concerning the presence of cirrus or ice clouds, the effect of their phase function and the vertical distribution were evaluated on the scattering of far infrared radiation. Thus, the accurate reconstruction of the phase function of cirrus clouds potentially indicates the need for application of a radiative transfer model. This specific module necessarily includes scattering parameters, while the accuracy of its calculations needs to be verified against real measurements.
For several applications the preliminary detection of those portions of the scene affected by the presence of clouds (cloud detection) is mandatory. For studying properties of Earth's surface targets affected by the presence of clouds are flagged just to exclude them by further analyses. In some case clouds themselves are the object of interest. In both cases the identification of clouds (and their classification) is mostly done by using (combination of) specific spectral signatures. Generally speaking clouds are highly reflecting VIS/NIR radiation showing (due to their heigth) brigthness temperatures (in the TIR region) lower than underlying surfaces. Thin or semi-transparent clouds are still detectable for their higher reflectance over the sea which represents a quite dark bacground in the VIS/NIR/SWIR region. Over land (much more reflecting) such a test is not more efficient and more sophisticated tests (e.g. Brigthness Temperature Difference in the split window bands around 11 and 12 microns) are required. In presence of very cold, high reflective backgrounds (e.g. snow, glaciers, etc.) both tests on the VIS reflectance and on TIR brigthness temperature could fail. More specific tests exploiting the reflectance drop of snow in the SWIR (where clouds are still saving their higher reflectance) helps to discriminate the presence of clouds from clear sky conditions even over a snow background. In the microwaves clouds are quite transparent except when coupled with coarse particles related to rain, snow, hailstones (precipitating clouds). In that case Mie scattering dominates strongly reducing the amount of radiance collected at the sensor (lower brigthness temperature in the microwave spectral range).