SARAL/AltiKa Waveform Analysis to Monitor Inland Water Levels: A Case Study of Maithon Reservoir, Jharkhand, India

Abstract

In the present study, behavior of the SARAL/AltiKa (Satellite with ARgos and ALtiKa) waveforms over Maithon Reservoir (∼65 km2 of surface area), Jharkhand, India, has been studied. The estimated water level has been compared with the in situ measurements at hydro-gauging station at the dam site. The problem of minimization of errors in the water level retrieval from AltiKa measurements has been resolved by improvement of the retracking method. A real retracking gate detection algorithm based on statistical analysis harnessing various physical parameters of the waveform has been developed, which has been applied to SARAL/AltiKa waveforms over the Maithon reservoir. Comparing the in-situ measurements with altimetry data (from cycle 1, 19 March 2013 to cycle 12, 8 April 2014) showed that it is crucial to improve the retracking method. Results showed accuracy of water level monitoring increased by nearly 76% by the newly developed waveform retracking algorithm over non-retracked water level. We also compared this new method with the existing ice-1 algorithm and found that with the new method there is improvement of ∼27% over ice-1 retracked water level. The correlation coefficient values and root mean square values without retracking, with ice-1 algorithm and with newly developed retracking algorithm were 0.87, 0.91, and 0.95, and 8.12 cm, 2.08 cm, and 1.42 cm, respectively. This shows the proposed retracker performed better than ice-1. The retracking procedure helped in outliers' identification and substitution and with waveform fitting and waveform parameter extraction. This algorithm should have good performance capability for retrieving water level over inland water bodies like Maithon reservoir.