A physically based algorithm for land surface emissivity retrieval from combined mid-infrared and thermal infrared data

<正> On the basis of the concept of the two-channel Temperature-Independent Spectral Indices (TISI), a physically based method is developed to extract the directional emissivities in mid-infrared and thermal infrared channels from day-night space measurements. A phenomenol-ogical model with three parameters is also proposed in this paper to describe the angular variations of the reflectivity (or emissivity). Having applied the proposed method to AVHRR data on an area covering the Iberian Peninsula (rather vegetated) and on a region centered on Tunisia (arid area), one can see from the results that the terrestrial surfaces do not behave as Lambertian reflector and angular variations of bidirectional reflectivity for bare soils appear to be azimuth-independent whereas those for vegetation present a pronounced backscattering effect. As for directional emissivities, values of vegetated areas are found to be higher and remain rather constant whatever the view angle is. On the contrary, on arid areas, values are f