Spatial and temporal pattern of rainfall extremes in Iran

Extreme rainfalls are important hydrometeorological variables for water resources management, flood mitigation, and soil conservation in Iran. This study examined annual and monthly maximum 24-hour rainfall from 135 stations across Iran, focusing on trends, frequency distributions and stochastic characteristics to investigate spatial and temporal patterns. Although a few stations exhibit statistically significant trends, most western and northern regions show increasing trends, while decreasing trends dominate the central and eastern semi-arid areas. Frequency analysis identified the Generalized Logistic distribution as the most prevalent distribution for extreme rainfall in Iran, with no clear spatial pattern in distribution type. In addition, spatioal analysis of L-moment statistics revealed high L-coefficients of variation in arid and semi-arid regions, while skewness and kurtosis did not show distinct spatial patterns. Lag-1 and Lag-12 autocorrelation coefficients of monthly extreme rainfall were also examined, revealing weak temporal memory and seasonal autocorrelation for most stations. Seasonal autocorrelation was more pronounced in the humid and semi-humid western and northern regions compared to the arid and semi-arid regions of Iran. These results highlight significant spatial heterogeneity in extreme rainfall patterns and underscore the challenges of predicting extreme rainfall events due to their low temporal predictability and high uncertainty. The results emphasize the need for robust hazard and risk management strategies to address rainstorm- and flood-related risks across Iran.