Estimating discharge of the Ganga River from satellite altimeter data

We use the water level data from multiple satellite altimeter missions to estimate discharge at different reaches of varying channel width (130 m to 2 km) of the Ganga River in India. We have established five (Kachla bridge, Kanpur, Shahzadpur, Prayagraj, and Mirzapur) virtual stations in the middle and two (Azmabad and Farakka) in the lower reaches of the Ganga River. For these stations, we acquired the water level from different satellite altimeter mission ERS-2 (1995-2007), ENVISAT (2002-2010), and Jason-2 (2008-2017) from publicly available databases. We applied datum and offset corrections on the altimeter data to make them comparable with the water level measured at the nearest gauge station. At each location, water level from the altimeter and gauge station show a good agreement with root mean square (RMS) error in a range between (22 -71 cm). We plot the altimeter water level as a function of their corresponding discharge measured at the nearest gauge station to establish a stage-discharge rating curve for each location. We then use these rating curves to estimate monthly discharge of the Ganga River from the altimeter water level. Based on the overall performance analysis of the statistical parameters, i.e; Nash-Sutcliffe efficiency (NSE); 0.86-0.98, RMS-observations Standard deviation Ratio (RSR); 0.15-0.38, Percent Bias (PBIAS); 13-27, and the coefficient of determination (R-2); 0.87-0.98, we show that the estimated discharge from altimeter water level accord well with the in-situ discharge measured at the gauge station. According to the Moriasi guideline, our estimate of discharge at all the virtual stations (except Kanpur) can be categorised between good to satisfactory.

Automated summary

This study estimated discharge in different reaches of the Ganga River in India using water level data from multiple satellite altimeter missions. By establishing stage-discharge rating curves, monthly discharge of the river was estimated based on the altimeter water level. The analysis showed a good agreement between estimated discharge and in-situ measurements, indicating the effectiveness of the method.