Characterization of Natural Wetlands with Cumulative Sums of Polarimetric Sar Timeseries

Wetlands are among the most productive natural ecosystems in the world, generally being important biodiversity hotspots. However, the complex nature of these landscapes together with the fragile and dynamic relationships among the organisms inhabiting these regions, make wetland ecosystems especially vulnerable to environmental disturbance, such as climate change. Thus, developing new automated systems which allow continuous monitoring and mapping of wetland dynamics is crucial for preserving their natural health. Synthetic aperture radar (SAR) systems have proven useful in monitoring and mapping the hydrological processes of wetland ecosystems through the use of polarimetric change detection techniques. Nonetheless, most of these flood change detectors rely on static detection approaches, generally covering a limited period of time (e.g., pre and post flooding scenario comparison), thus failing in providing continuous information about the diverse hydrological mechanisms. In this context, this research presents a novel approach for monitoring the hydrological dynamics of wetlands in a continuous and near-real-time manner using dense Sentinel-1 image time-series. In this work, we have enhanced our recently developed algorithm based on cumulative sums (SAR-CUSUM), to include polarimetric information, which allows to classify the type of change due to the flood. The new processing stack exploits the polarimetric information of dual-pol Sentinel-1 dense time series for detecting floods and provide some separation between open water and flooded vegetation areas. The outcomes derived from this study emphasize the capabilities of dense SAR time-series for environmental monitoring while providing a useful tool which could be integrated into rapid response and wetland conservation management plans.