Monitoring Changes in the Prespa Lake Watershed Using Remote Sensing Data

Abstract

Surface water is one of the most vital Earth resources changing in time and space as a consequence of land use/land cover (LULC) changes, climate change, and other environmental factors. Timely monitoring of the water area and its surroundings is essential for policy and decision-making processes. Monitoring the land cover changes with conventional methods may consume significant time and resources. With the development of remote sensing, land cover monitoring has become practical and effective. The most practical way for monitoring the changes is with the classification of satellite imagery. The main aim of the study is to make a five-year land cover change in a watershed using remote sensing data. For this purpose, two satellite images, from RapidEye-3 and Sentinel-2 have been classified. As a study area, the Prespa Lake watershed has been selected. The Prespa Lake is situated in the Balkan Peninsula and is shared among Albania, North Macedonia, and Greece. For the purpose of the study, a semi-automated object-based classification model has been developed. The results from the classification showed significant accuracy of 93 – 96%. For future studies, it is recommended further investigation, such as using monthly satellite imagery and modeling precipitation and meteorological data.