A novel method of satellite based river discharge estimation using river hydraulic geometry through genetic algorithm technique

Abstract

Many countries experiences floods due to heavy flows carried by trans-boundary Rivers. Hydrological modelling in un-gauged catchments and trans-boundary Rivers has become a challenging task for hydrological modellers. This research aims at estimating river discharges computed solely using satellite images, with no ground-based or a prior information at all. It is also aimed at testing the applicability of approach in very well defined channel to braided channels and in very wide range of discharge conditions. The approach works owing to principle of hydraulic geometry of natural rivers. Monitoring these discharges in upstream areas will help in forecasting flows in the downstream areas. In this paper, an attempt is made to estimate the river discharge using river flow width information derived from temporal satellite remote sensing data alone. The approach is based on Many Stations Hydraulic Geometry (AMHG) principle through Genetic Algorithm Technique. The technique is demonstrated in four major rivers in India for various discharge values using temporal remote sensing data of Indian Remote Sensing Satellites (IRS) and LANDSAT. Computed discharges are validated with observed discharges in all the identified river reaches using statistical method and found to have very good match. This novel approach is useful globally for quantifying river discharges with continuous imaging of the rivers and has strong advantage in flood early warning and water resources assessment.