Mapping Floods in Lowland Forest Using Sentinel-1 and Sentinel-2 Data and an Object-Based Approach

The impact of floods on forests is immediate, so it necessary to quickly define the boundaries flooded areas. Determining extent flooding in situ has shortcomings due possible limited spatial and temporal resolutions data cost collection. Therefore, this research focused flood mapping using geospatial remote sensing. area located central part Republic Croatia, an environmentally diverse lowland Sava River its tributaries. Flood was performed by merging Sentinel-1 (S1) Sentinel-2 (S2) mission applying object-based image analysis (OBIA). For purpose, synthetic aperture radar (SAR) (GRD processing level) were primarily used during period possibility all-day imaging all weather conditions detection under density canopy. pre-flood S2 imagery, a summer acquisition, as source additional spectral data. Geographical information system (GIS) layers—a multisource forest inventory, habitat map, hazard map—were sources assessing accuracy interpreting obtained results. signature, geometric textural features, vegetation indices applied OBIA process. result work developed methodological framework with high speed production. overall classification 94.94%. Based conducted research, usefulness C band S1 leaf-off season determined. paper presents previous describes SAR parameters characteristics floodplain significant classification.