Altimetria por satélite radar na medição do nível d'água de rios de pequeno e médio porte: um estudo de caso no Rio São Francisco

Abstract

Although radar altimetry is used primarily to study ocean topography and monitoring of the ice caps, the amount of data available favours its use in continental hydrology studies. Satellite altimetry radar can contribute to minimize problems faced in monitoring water resources, being able to generate data similar to those of conventional hydrologic monitoring networks complementing the already existing uviometric network; however data processing techniques are necessary to get precise results. Due to this, the development of methods that minimize errors is of utmost importance. In this study a retracker was used to attain water levels from data of the Envisat, Sarala ns Sentinel-3A, besides that an error minimizing algorithm was also used in the three satellites' data. A systematic method of data extraction and correction was developed to be used along the S~ao Francisco River (RSF). The data were generated at the crossing points between the satellite tracks and the river, that originated the virtual monitoring stations (EV). The derived altimetry data were compared to the in situ data, from which the Mean Square Error (EMQ) was calculated and a linear regression (RL) applied. In the second part of this studyan analysis of the waveforms (FO) was made allowing to better understand the space-time evolution of the radar signal. After applying the error minimizing algorithm in 77% of Saral's EVs and 61% of Sentinel-3A's the results were better than the ones from applying the most commonly used retracker. With Envisat this last approach was more a_ective than the one with the use of RANSAC due to the high consistency of this satellite's data. It was possible to verify that the precision of the measurements along the RSF generated from the developed method is be_tting with the expected for this three sensors and resulted in data consistent for water level measurements, especially with the use of Envisat. In addition, the space-time evolution analysis of the FO allowed to con_rm the quality of the data for Envisat, Saral and Sentinel-3A, the _rst one being the one with the Best FO patterns.