Adaptive retracking of Jason-1 altimetry data for inland waters: the example of the Gorky Reservoir

摘要

Standard altimetry data processing developed for open ocean conditions can be inapplicable for the case of inland waters, especially for narrow, elongated water-bodies and rivers, where the distance between shores is less than 10 km (while the eliminated area within the gain of the radar antenna for Jason-1,2 is about 50 km). These conditions are typical, for example, of the majority of reservoirs of the Volga River cascade (with one exception: the Rybinskoe Reservoir). Under these conditions only a few telemetric impulses fit the validity criteria, which causes a severe loss of data. Besides, errors in the water level retrieved from the altimetric measurements are enormous, as was demonstrated on the basis of comparison of in situ measurements at hydro-gauging stations for the water level of the Gorky Reservoir of the Volga River and all that is available along track 10 Hz TOPEX/Poseidon altimetry data and 20 Hz Jason altimetry data over the reservoir area. The problem of minimization of the errors can be resolved by retracking. For justification of the optimal retracking algorithm, the average impulse response of the statistically inhomogeneous surface was calculated theoretically, based on the works of Brown (1977) and Barrick and Lipa (1985) for the model of the terrain in the vicinity of the Gorky Reservoir. The model represents the main typical features of the waveform examples (e.g. high peaks or irregular complex shape), and the modelled waveforms are in good agreement with the Jason-1,2 waveforms for the same area. It was shown that for the Gorky Reservoir significant wave height (SWH) did not exceed 0.5 m (corresponding to the width of the leading edge less than 1 telemetric gate). Under these conditions the retracking algorithm based on the detection of the beginning of the leading edge of telemetric impulses is preferable for a correct assessment of variations in the water level in the Gorky Reservoir. A comparison of the data with in situ measurements at the hydro-gauging stations for the water level of the Gorky Reservoir shows that retracking dramatically increases the number of data involved in monitoring and significantly improves the accuracy of the measurement of the water level. The retracked data of water level have also been validated by comparing them with Jason-2 and Jason-1 after carrying out measurements for the Gorky Reservoir. The general principles of retracking algorithms for complex areas (land, coastal zone, inland waters, etc.), based on calculations of the waveform and taking into account statistical inhomogeneity of the reflecting surface adjusted to a certain geographic region, are discussed.