Comparison and correction of satellite measurements using in-situ observations of lake surface heights: A case study in lake Baikal

Abstract

Over the last few decades, satellite radar altimetry has developed enabling a wide range of applications, including the monitoring of lakes. As a method to measure lake water level time series, satellite radar altimetry contributes to gaps in data of in-situ networks in time and space, which is freely available, continuous, global, regular, and precise. However, depending on the location, shape, size, and surrounding environment of the lakes, the accuracy of water heights inferred from altimeters varies significantly. The accuracy of water heights also strongly depends on the quality of space-born instruments, with modern models having evolved from radar altimeters launched in the 1990 s. To determine the level of accuracy of radar altimeters, it is necessary to use ground observations of lake heights. However, the measurements performed onboard by altimeters and on the ground by water level gauges are not directly comparable because the physical properties and scaling of the measurements are completely different. In this study, we present an approach which permits realistic inter-comparison of water levels, in relation to a large lake with a large number of hydrological stations (at least 10) and the presence of at least 10 satellite measurements per month (on the example of lake Baikal), in order to provide diagnostics on the quality of the radar altimeter and statistics of the errors.