A 160-Year Analysis of Extreme Precipitation Hazard Shifts in Western Asia: Perspectives from Historical Trends and AR6 Projections

Western Asia, with its diverse and complex geographical and environmental conditions, is one of the world's most vulnerable regions to climate change making it particularly susceptible to extreme precipitation events (EPE). This study develops a probabilistic risk assessment for EPEs across the region, identifying high-risk zones for heavy rainfall and flash floods through analysis of 160 years of daily precipitation data, including ERA5 reanalysis data (1941–2020) and 7 CMIP6 models under two scenarios of SSP2-4.5 and the most pessimistic one, SSP5-8.5 (2021–2100). By considering the trends of 10 certain indices (EPIs) based on both historical precipitation data and future projections, The findings categorize areas into four distinct risk levels, ranging from no risk to high risk. The statistical significance of EPIs was assessed using the nonparametric Mann–Kendall test on the significance levels of 5%. For each grid cell, trends in all EPIs were calculated, and grids exhibiting statistically significant upward trends were classified as susceptible to extreme precipitation. This criterion - positive and significant trends across indices - was systematically applied to all grids in the study area, generating a spatially explicit risk classification. Regions with overlapping susceptibility signals from multiple indices were prioritized as high-risk zones.

The same methodological framework was applied to the future climate projections derived from the selected CMIP6 models. To combine the results across the different models and enhance the reliability of the classification, a machine learning technique was employed. Specifically, the Random Forest (RF) algorithm was used to integrate the spatial outputs of all models based on the "most agreed-upon risk level" for each grid cell. The goal was to produce a unified classification map that reflects the most probable future risk zones for extreme precipitation.

The outcome of this analysis shows that although the "high-risk" area decreases from 10% to 2% under the SSP2-4.5 scenario, this percentage increases significantly to 25% under the SSP5-8.5 scenario. In other words, 25% of the region is classified as high-risk in the pessimistic scenario and indicate a pronounced shift, with a majority of previously moderate-risk areas transitioning to high-risk categories by 2100. Overall, the "no-risk" area decreases from 55% to 46% under SSP5-8.5, indicating a general increase in the area susceptible to extreme rainfall. Notably, northwestern Iran and central Turkey consistently appear as high-risk zones across all analyses, from historical part through both future scenarios, highlighting their persistent vulnerability.

These findings give confidence to the potential impacts of flooding and infrastructure challenges in regions unfamiliar to dealing with heavy rainfall. This information is important for water management strategies that are part of preparing for climate change impacts, which clearly emphasize the rise in extreme weather patterns across the Middle East.