Multi-sensor Space-Borne Earth Observation Data for Characterizing Fluvial-Geomorphic Provinces: A Case Study from Kosi River, India

Abstract

The dynamism of geomorphic provinces in fluvial systems present considerable ambiguities in mapping by remote sensing. This necessitates use of multiple satellite data to characterize such depositional provinces. We use, an integrated dataset to characterize the geomorphic provinces (e.g. active flood plain, older food plain, fan etc.) of the Kosi River (Bihar), India. This is done using contrast in spectral signatures derived from multispectral bands (of IRS-P6 LISS III), radiant temperature (from ETM+) and radar-roughness (from radar brightness image RISAT-1). ASTER DEM has been used in deriving topographic profiles. The optical imagery, enables regional characterization through direct tonal changes (e.g. active flood plain is brighter than older flood plain). The radiant temperatures show variations across provinces. Geomorphic transitions are represented by topographic breaks. Radar backscatter imagery, show differences in radar-return from different sub-provinces. Observations made using specific sensor characterize each provinces and is supplementary/complimentary to the parameter(s) from other sensors.