Investigation of water scarcity in terms of climate change and population scenarios via machine learning in the city of Bursa

With global climate change, changes are expected in the mean and extreme conditions of climatic systems and many parameters in their subsystems, both spatially and temporally. Studies show that the climate is changing on a global scale, but the Mediterranean basin is one of the regions that will be most affected by climate change in Turkey, although it is a hot-spot in terms of climate change. From this point of view, it is expected that there will be a problem in the presence of water and water scarcity with climate change in Turkey. In this context, in order to understand the water scarcity and droughts that will be caused by climate change, the volume amounts and values of the dam that provides the urban water need of Bursa Province in the Marmara Region have been determined. The water needs of the city center are met by the Doğancı Dam, and in this respect, various projections have been produced for the future dam volume. For this, firstly, the RegCM4.4 regional climate model, which is dynamically reduced to 10 km, is used. Evaporation values of the study area were obtained according to the Penman-Monteith method based on the RegCM outputs. Then, the flow values of the streams coming to the Doğancı Dam basins were calculated for the future by using the SCS-CN precipitation-flow model. After these values were obtained, a statistical model was established between precipitation, evaporation and flow values and dam volume by using the observation data for the reference period with the Support Vector Machine (SVM) method, which is frequently used in machine learning. Then, dam volumes were calculated by SVM on the basis of RCP 4.5 and RCP 8.5 scenarios using RegCM data to obtain future values of dam volume. The results obtained were converted into the Standardized Reservoir Index (SRSI) and the future dry and humid periods of the Doğancı Dam were revealed. After determining the drought values of the Doğancı Dam basin, the population of Bursa city center was estimated until 2100 by using the Arithmetic Growth Model and the Exponential Growth Model, which are among the population projections, in order to understand the water scarcity. As a result of this study, it was determined that droughts increased and extreme conditions of droughts increased when compared with the reference period. In the RCP 4.5 near future scenario, going to the dry period has been determined. In the RCP 4.5 distant future and RCP 8.5 near future scenario, dry and humid periods followed each other. While RCP 8.5 showed a trend towards drought in the far future scenario, it then showed a trend towards a humid period. In line with the aim of determining the water scarcity in terms of the dam, the ratio of water to the population was determined. The amount of usable water and the amount of water per capita were calculated by dividing the volume obtained by the population. The reference period and RCP 4.5, RCP 8.5 near and far future scenarios were compared. While the amount of water per capita was 11479 liters in the reference period, this value is 5319 liters in the RCP 4.5 near future scenario, while it is 3830 liters in the RCP 4.5 distant future scenario. While the amount of water per capita is 5514 liters in the RCP 8.5 near future scenario, it is determined as 3770 liters in the RCP 8.5 far future scenario. The amount of water per capita has decreased compared to the reference period, and it has been determined that their frequency has decreased. Compared to the reference period, decreases were observed in the volume amounts in the dam.