SAR-based detection of flooded vegetation - a review of characteristics and approaches

The ability of synthetic aperture radar (SAR) to detect flooded vegetation (FV) (the temporary or permanent occurrence of waterbodies underneath vegetated areas) offers a great benefit in the research fields of flood and wetland monitoring. The growing demand for near real-time information in flood monitoring and an increased awareness of the importance of wetland ecosystems are strong drivers for the ongoing research in these fields, where FV constitutes an essential part. This study reviewed 128 publications summarizing the knowledge about the relationships between the SAR parameters and the environmental conditions for the detection of FV. An advanced review of 83 studies was carried out to gain insights about applied classification techniques and SAR data for the extraction of FV. Although some trends emerged about which wavelengths, polarisations, or incidence angles to use, there is variation in the application of different classification techniques or using SAR-derived information depending on the data sets and the study area. Notable throughout the analysed articles is the growing demand for unsupervised and computationally efficient methods of higher accuracy for the extraction of FV. Based on the advances in SAR with regard to spatial and temporal resolution, the development of robust approaches for the extraction of FV from various and complex environments has to be further pursued.