Determining the vineyards areas in mersin province tarsus district using remote sensing and geographical information systems with sentinel-2 satellite images

With this study, it is aimed to determine the vineyard areas, which are of great importance for the regions agriculture, with remote sensing methods, quickly and reliably, and to contribute to the creation of a database related to these areas. In the study, Sentinel-2 satellite images of 2019 and 2020 provided free of charge by the European Space Agency were used. Grapes, plums, nectarines, citrus, fallow (wheat), olives, forests and non-agricultural fields, which are grown in 6 neighborhoods in Mersin province Tarsus district, where viticulture is intense, were examined in 8 classes. Test data were created as a result of Farmer Registration System and field studies. Object Based Controlled Classification was made with 2 satellite images of 2019 and 8 satellite images of 2020 from Sentinel-2 satellite images. In the thematic maps obtained, the classification accuracy was determined as 75% and 71% in July and August 2019, respectively. In the spring and autumn images of 2020, the accuracy was low, and 71% accuracy in the summer months. The κ value obtained as a result of the study was determined as 0.67 and 0.64 in July and August 2019, respectively. It was found to be low in the spring and autumn images of 2020, and 0.64 in the summer months. The highest user accuracy of the vineyard areas in Tarsus District was obtained from the classification of the July 2019 image with 91%. Vineyard areas cover the largest area with 1,966 hectares, followed by olive fields with 1069 hectares. Keywords: Vineyards, Sentinel-2, Tarsus, Remote Sensing, Geographic Information Systems