Methodology of Adaptation of Satellite Measurements for the Assessment of Water Levels of Large Reservoirs (on the Example of Reservoirs in Russia)

Abstract

In the last 15-20 years, satellite altimetry has been actively used to determine the water levels of land water bodies. At the same time, satellite measurements contain significant errors, that can be eliminated by correcting them based on ground-based observations. As practice shows, the smallest deviations of satellite data from the results of ground observations are characteristic of the largest lakes, and the largest — for reservoirs. The article, using the example of five large Russian river-type reservoirs, examines the reasons for the reduced accuracy of satellite data and outlines a methodological approach for correcting real-time and monthly averaged water level values using the results of ground-based observations. Previously, the series of satellite values are ‘smoothed out’, and then corrected by removing ‘outliers’ and eliminating systematic errors. Further, the corrected (‘filtered’) satellite series are checked for representativeness, i. e. the degree of compliance with long-term series of ground-based observations. The verification criteria are acceptable differences in the values of standard deviations and coefficients of linear correlation of long-term series of corrected satellite and ground data. After correcting the satellite data on the reservoirs under consideration, performed according to the proposed methodology, and obtaining ‘filtered’ series, subsequent verification by quality criteria showed their representativeness and suitability for practical use.