The contribution of multi-sensor optical satellite data to flood detection: the performances of RST-FLOOD in Basilicata region (southern Italy)

Over recent years, flood has appeared as one of the most frequent, pervasive and devastating natural hazards in the world. One of the major challenges during flood is to get timely and frequently updated overview of the extent of the affected area. Using satellite remote sensing methodologies, based on multi-source data, can supply such an information about areas affected by floods as well as their spatiotemporal evolution, useful to water managers in better planning relief and rescue operations and in general in providing effective risk management.

In this work, the RST-FLOOD algorithm, a differential approach aimed at detecting flooded areas by processing multi-year time series of daytime optical data, has been implemented for studying the flooding event affecting the Basilicata and Puglia regions (southern Italy) in the first week of December 2013. Spatial and temporal dynamics of the flooding in the study area were revealed, calculated and analysed by processing historical images acquired by MODIS (Moderate Resolution Imaging Spectroradiometer) and VIIRS (Visible Infrared Imaging Radiometer Suite). To this aim, visible (VIS) and near infrared (NIR) MODIS data (at 250m of spatial resolution) and VIIRS VIS, NIR and short-wave infrared - SWIR bands (at 375m of spatial resolution), were used.

On the basis of the achieved findings, RST-FLOOD showed satisfactory potential in effectively and accurately discriminating flooded areas, especially when assessed with the official flood hazard map of the area, provided by the Italian Ministry of the Environment and Land Protection and compared with the outputs of another state-of-art VIIRS-based method. Flood dynamics was recognized with good reliability and sensitivity, identifying also quite large flooded areas outside the hazard map, suggesting an updating of the one currently in force to better prevent further future damages.