Dynamics of shallow lake cover types in relation to Paraná River flood pulses: assessment with multitemporal Landsat data

Abstract

The presence of shallow lakes (SL) is typical of large floodplains with wetland landscapes. By taking into account the dynamic behavior of SL, multitemporal remote sensing data can be used to delineate and monitor water extent and SL cover types. The aim of this study was to analyze the spatial and temporal patterns of SL coverages related to flood pulses in the Lower Paraná River floodplain. SL were delineated with a frequency analysis over a 1987–2015 time series of the Normalized Difference Vegetation Index derived from Landsat 5-TM and 8-OLI data. A total of 303 SL were identified (22% of the study area). Most of the SL were small, only eight were larger than 1.5 km 2 (66% of the total SL area). Five dominant SL cover types were identified by using multitemporal Landsat data. The dynamics of these cover types were analyzed along seven flood pulses and we described six typical temporal patterns that repeated and characterized most of the SL areas. Vegetation quickly recovered after floods, except for extreme floods related to the occurrence of El Niño-Southern Oscillation events. The historical Landsat data turn out to be valuable tools for the reconstructions of historical processes in floodplains of large rivers.