A suitable model proposal via remote sensing and geographic information systems in the Bartin Stream Basin for climate change scenarios

In this study, modeling of the situation that will emerge in the Bartın Stream Basin for the year 2050 was conducted according to RCP climate change scenarios by evaluating the effects of bioclimatic data and environmental variables on land-cover and land-use changes. Modeling was carried out through maximum entropy in accordance with the purposes of; (i) study serving the protection and management objectives of the changes in landscapes partaking in the European Landscape Convention and (ii) developing certain recommendations at a sectoral scale regarding the study area. The entire study was carried out using Landsat satellite images. Thornthwaite climate classification was employed while determining the satellite image selection dates. After the satellite images were cut according to the field boundaries, the maps were obtained by calculating the normalized vegetation difference index (NDVI) for each of the years 1984-2019. The study area was divided into grids and the Shannon-Wiener diversity index was computed for each pixel. It can be said that the fragmentation and artificialization increase in areas where the Shannon-Wiener diversity index value increases, whereas in areas where the Shannon-Wiener diversity index value decreases, the natural areas are not deteriorated or slightly deteriorated. Within the scope of the thesis, it was revealed that using the NDVI data between 1984 and 2019, the prediction of 2050 can statistically be explained. The controlled and uncontrolled classification was conducted by using the NDVI map created from the satellite image of 2019 and the estimated NDVI map for the year 2050. Afterward, the maximum entropy method was employed for the modeling. Both environmental variables and bioclimate variables show high correlation within themselves and this causes a multicollinearity problem. Therefore, representative variables of 42 different variables were determined with statistical analyses. In the study area, two different changes are observed between 1984 and 2019. The first of these is the changes where the fragmentation in the area increases, artificiality increases, and natural areas are lost. The latter is the changes where natural areas are protected as much as possible or changed little, and there is no or little fragmentation. Two different models were obtained with this approach. By determining the bioclimatic variables affecting these models, land-cover and land-use situation for the year 2050 was revealed by mapping according to RCP climate change scenarios based on RCP 2.6, 4.5, 6.0, and 8.5. It was demonstrated by the statistical models created within the scope of the thesis that the changes in LULC will negatively (increase in both the artificialization and fragmentation) increase in the Bartın Stream Basin. In this context, the relationship with the previously made spatial plans in the study area was examined and a mutual evaluation was made with the outcomes of the thesis. After all these, certain recommendations were developed at a sectoral scale. As a result, it was revealed that; (i) the modeling of land-use and land-cover changes based on climate change scenarios should be taken into account when making various planning and management decisions, (ii) the effects of climate change, one of the biggest threats of our age, can be predicted by statistical modeling, (iii) and thereby it is probable to protect lands by taking various precautions.