Agriculture applications

The use of SAR imagery for agricultural applications has been studied extensively in particular because it is possible to acquire data at various times during the growing season. Single channel SAR data have been found to provide less information than multi-spectral optical data. This can be often overcome with a temporal series of SAR observations. As the crops grow and mature the backscatter characteristics change and these variations which depend on the crop can be used as the basis of crop discrimination.

The backscatter from agricultural targets is composed of surface scattering from the soil, volume scattering from the plants, and a soil-vegetation interaction term. The relative contribution of each component is a function of system and target parameters. In general, at C-band, the return is composed of a combination of these components with the soil surface term dominating early in the growing season and the vegetation volume term dominating during the peak growth period. At the end of the growing season, a mixture of the component returns is generally present with surface and soil-vegetation interaction terms being the dominant ones. This makes information extraction difficult as the soil or the crop may be the target of interest and other components produce "noise" which adds significant error to the estimation process. Once again multi-temporal observations help, but difficulties persist.

Research to date has demonstrated that the additional information in polarimetric data (both magnitude and phase) can help to increase the information content for agricultural applications thereby decreasing the need for multi-temporal imagery.

Two examples of such work are given on the following pages:

• Soil conservation: Soil tillage and crop residue
• Crop productivity / Within field variation