Water Body Detection in Sentinel-2 Satellite Images to Integrate GIS Data for Rome Smart City

The water bodies (rivers, lakes, seas and oceans) often constitute founding elements of the cities landscape, so a monitoring action to preserve their environmental value is undoubtedly necessary. Remotely sensed images can offer quick, effective and economic support for monitoring water bodies, providing their position, form, extension, consistence and state of health. This article aims to highlight how Normalized Difference Water Index (NDWI) applied to Sentinel-2 satellite images allows to monitor water bodies and detect the variations over time connected to disturbing causes due to climate change and phenomena such as drought and floods. The city of Rome is taken into consideration particularly significant for the presence of both an important river such as the Tiber (third in Italy by length), and for a long stretch of coast on the Tyrrhenian Sea. The results demonstrate that a fast procedure is possible for water body detection in Rome to integrate Geographic Information System (GIS) data and support monitoring plan. In fact, NDWI applied in QGIS allows to obtain a map of the area which, suitably classified using maximum likelihood algorithm, produces an identification of water bodies with high accuracy (user, producer and overall accuracy values are all greater than 98%).