Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Royaume-Uni → 1987 → Angular models for interpreting satellite-based measurements of shortwave and longwave radiances over deserts

University of London (1987)

Angular models for interpreting satellite-based measurements of shortwave and longwave radiances over deserts

Brooks, D.R

Titre : Angular models for interpreting satellite-based measurements of shortwave and longwave radiances over deserts

Auteur : Brooks, D.R

Université de soutenance : University of London

Grade : Doctor of Philosophy (PhD) 1987

A major problem for interpreting satellite-based measurements of the Earth’s radiation is inferring the radiant energy reflected or emitted from a point on the Earth’s surface into an entire hemisphere from a single measurement of radiance along the line of sight from the point on the Earth to a detector on the satellite. Especially over desert surfaces, dependence of both shortwave and longwave radiance upon solar zenith angle, viewing zenith angle, and azimuth is prominent in both the reflected and emitted portions of the earth’s electromagnetic spectrum. Data from several satellites with differing orbital properties have been used to examine these directional effects, including the GOES, METEOSAT, Nimbus 7, and ERBS. The desert areas examined include the Sahara, Sahel, Saudi Arabian Peninsula, Kalahari, and western Australia. For analysis of the clear surface-atmosphere system, a variety of techniques are employed to reduce cloud contamination in satellite data. One- and two-dimensional thresholding and spatial coherence are discussed and these methods are applied to produce data sets that are considered to represent the cloud-free desert-atmosphere system. For the GOES, METEOSAT, and Nimbus-7 satellites, the data are binned in 1o x 1o regions. For such areas, the clear desert-atmosphere surface has a typical daytime variability in the shortwave of less than 10 percent under constant viewing conditions,and a longwave variability on the order of 4 percent. For the ERBS, data are taken along single overflights across the selected desert regions, with variability in the shortwave and longwave similar to that observed for the other satellites. For longwave measurements, the two most prominent phenomena are limb darkening and the diurnal cycle. For limb darkening, a parameterization in terms of atmospheric path length is developed that allows longwave radiant existence to be calculated from a measurement of radiance at a single value of viewing zenith angle. Some of the available measurements exhibit azimuthal effects that are thought to be due to shadowing from topographic features, rather than being an intrinsic property of desert surfaces. The diurnal cycle is parameterized in such a way as to separate the magnitude, lag time, and decay of the diurnal signal. This signal lags behind local solar noon by an hour or two and decays for several hours after sunset before approaching an essentially equilibrium condition just before the next sunrise. For shortwave measurements over deserts, the most prominent features are the large variability in surface reflectance (or albedo) for different desert surfaces or within the same desert region on a spatial scale of several hundred kilimeters, and the strong anisotropy of the reflected signal. This pronounced surface variability favours examining the bidirectional function on a spatial scale small enough to retain a high degree of spatial homogeneity. Parameterizations are developed to allow inferrence of shortwave radiant existance from a measurement of radiance taken at specified conditions of solar zenith angle, viewing zenith angle, and relative azimuth. Analytical formulation of the bidirectional reflectance function makes possible the use of restricted measurements over small regions, although data from more than one satellite is typically required to obtain good estimates of the parameters involved. Inclusion of data from more than one satellite requires intercalibrations that are accomplished by matching data to the same time and angular viewing conditions whenever possible. The desert-atmosphere system is shown to be predominantly backscattering. However, the wide range of parameters obtained demonstrates that surface variability is a significant problem. The results indicate that data obtained from a single satellite may contain geographical biases within the observed range of angles, and should be treated with caution. In summary, data from severa


Version intégrale (10,4 Mb)

Page publiée le 12 février 2015, mise à jour le 25 novembre 2019