RADAR APPROACH TO ASSESS SOIL CHARACTERISTICS

Abstract

In this dissertation work, the scattering behavior of five types of soil mixtures having different percentage of coarse sand, slit and clay have been analyzed at various incidence angle with Horizontal- Horizontal (HH-) and Vertical-Vertical (VV-) polarizations for X-band frequency (10 GHz). The effect of roughness on the scattering coefficient is observed for various soil mixtures. The observations are carried out by indigenously assembled X-band Scatterometer. The surface roughness parameters are obtained using a surface roughness measurement profiler. The surface roughness effect was taken for the two type of roughness i.e. along the track surface roughness and across the track surface roughness. The angular and polarization effect has also been analyzed on the Scatterometer data. The suitable angle for the observation has also been found. Genetic algorithm has been applied to the Scatterometer data for the retrieval of the soil characteristics and along and across the track surface roughness. The genetic algorithm has also been applied on two ERS- SAR images of Roorkee region to retrieve the soil composition. For the both type of data Dubois model [101 has been used. On the SAR images the genetic algorithm was applied with a second approach. This second approach uses the actual ground truth data for the training of the genetic algorithm. A good agreement between the estimated soil characteristics has been noticed. Study infers that soil texture is quite sensitive to radar scattering and it is possible to retrieve the soil characteristics with radar/Scatterometer data with good accuracy and this type of retrieved results can be helpful to predict soil strength as well as soil erosion of particular area.