Improved Lake Level Estimation From Radar Altimeter Using an Automatic Multiscale-Based Peak Detection Retracker

• Main Authors: Jiaming Chen, Jingjuan Liao, Chao Wang
• Format: Article
• Language: English
• Published: IEEE 2021-01-01
• Series: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
• Subjects: Lake level; multiscale-based subwaveform; retracker; satellite altimetry
• Online Access: https://ieeexplore.ieee.org/document/9247396/

Satellite radar altimetry is an important technique for monitoring the water levels of oceans and inland water bodies, especially in areas where in-situ data are sparse or nonexistent. This study presented an automatic multiscale-based peak detection retracker (AMPDR). The retracker can extract a robust threshold level for each track, then the stable lake level can be obtained from the multipeak waveforms using a shortest-path algorithm. Additionally, the retracker can be used for mountain lakes and for flat lakes, and is also suitable for many kinds of altimetry data, such as those of Cryosat-2, Sentinel-3, and Jason-2/3. To validate the lake levels derived by the AMPDR retracker, the in-situ gauge data of seven lakes in the Tibetan Plateau and two lakes in a flat area are used. Moreover, seven existing retrackers are compared to evaluate the performance of the proposed AMPDR retracker. The results suggest that AMPDR can efficiently process many complex multipeak waveforms, and the AMPDR has the lowest mean of all track standard deviations over all lakes. The root-mean-squared error (RMSE) of the lake level time series obtained using AMPDR is the lowest over several lakes: The mean RMSEs of all the lakes overpassed by Cryosat-2, Sentinel-3, and Jason-2/3 are 0.149, 0.139, and 0.181 m, respectively. The AMPDR retracker is easy to implement, computationally efficient, and can give a height estimate for even the most contaminated waveforms.