Probabilistic Assessment of Future Drought Characteristics under the Quality Boosted Regional

Abstract

Water is one of the most essential components for sustaining human and other forms of life; thus, it must be supplied to meet life's necessities. Understanding the consequences of global warming and the trends in climate change is essential for the effective operation and strategic planning of water resource projects. Consequently, accurately and efficiently characterizing drought at the regional level constitutes a more complicated and challenging obstacle. The primary goal of this study is to predict future drought characteristics in Turkey's Konya Closed Basin (KCB) region by employing the Quality Boosted Regional Drought Index (QBRDI) framework. This study uses precipitation time series data from nine sites spanning 57 years to examine drought features in Turkey's KCB. The validity of the weighting procedure in QBRDI is compared with the findings of the simple model average (SMA) using statistical measures: root mean square error (RMSE) and relative absolute error (RAE). It is observed that the RMSE value (6.282) for the weighting procedure in QBRDI is lower than the RMSE value (15.158) computed within SMA. Similarly, the value of RAE across weighting procedure in QBRDI is 0.08525, while that for SMA is 1.409. The lower values of both RMSE and RAE indicate that the weighting procedure utilized in QBRDI is more reliable than that used in the SMA. Consequently, these measurements suggest that the weighting procedure utilized in QBRDI possesses higher reliability and validity for drought assessment than that in the SMA. For application, the long-term phenomenon of drought is determined using the steady-state probability of Markov chain. The findings from steady-state probabilities show that extreme drought tends to have a higher probability than extreme wet across most timescales. This pattern suggests that drought events are marginally more probable than extreme wet events, reinforcing the need for drought-related risk evaluation and mitigation measures. In addition, the Mann-Kendall test is employed to assess the presence of a monotonic trend, while Sen's slope estimator is used to quantify the magnitude of that trend, as it serves as a complement to the Mann-Kendall test by providing a robust estimate of the trend slope.