Assessment of Future Possible Meteorological Drought for Isparta Province

Abstract

There is an increase in drought events with climate change. Drought is a disaster that occurs over long periods of time for large areas compared to other natural disasters. For this reason, it is important to forecast the drought and take precautions. In this study, a longshort term memory network (LSTM) method is proposed to predict drought in the long term with high accuracy. Firstly, monthly precipitation values that are likely to occur in the next 10 years were estimated using the LSTM method with the monthly total precipitation measured between 1929-2021 belonging to the Isparta meteorology station. When the descriptive statistical properties of both precipitation series, which are predicted for both the current and the next 10 years, were examined, the similarities were seen. Then, standardized precipitation index (SPI) values were calculated for the years 1929-2021 and the next 10 years. At the same time, the trends of the SPI values obtained for the precipitation data between 1929 and 2021 and the SPI values calculated for the next 10 years were examined and the similarities of the trend analysis of all time intervals and the next 10 years’ data with the RCP4.5 precipitation projections were discussed.

Keywords

• meteorological drought
• spi
• lstm
• isparta