It is widely known that water resources are facing several issues, as their limited availability has been affected by climate change, human over-exploitation and environmental pollution. The reduction of the availability of water resources along with the deterioration of water quality could have devastating consequences for human beings and natural environment. Therefore, ensuring the availability, quality and quantity of water resources is a key issue which has attracted the attention of researchers, governments and the public. The construction of dams constitutes a water storage solution in areas suffering from severe water shortage problems. Dams are artificial barriers, constructed across a river in order to hold back the rain water which falls during the wettest months, while a lake or reservoir is being formed behind them. In the current study, we map and monitor the existing water resources within the reservoir of a newly con-structed dam during its gradual filling. In that context, we acquired Sentinel-1 radar data and Sentinel-2 multispectral data which are provided free of charge to all users through the Copernicus programme. Our data sets are covering a period of one and a half year, from the completion of the dam construction and the beginning of the reservoir’s filling until today, wherein the water level is at the highest point. Sentinel-1 data were processed using random forest classifier and dividing the area into water and land sub-parts, while Modified Normalized Difference Water Index (MNDWI) was calculated for Sentinel-2 data. The multi-dated water extents were integrated in an ArcGIS environment in order to evaluate them and investigate if a combined processing of radar and multispectral data of Copernicus programme could be the right choice towards a more accurate and detailed monitoring of water resources in a dam reservoir. Validation of the results has been performed using in situ measurements and observations.
|