Evaluation of 24‐Hour Probable Maximum Precipitation Methods for the Soummam Watershed, Algeria

ABSTRACT Extreme rainstorms are a major concern for flood risk management, particularly in regions with high precipitation variability. This study evaluates 24‐h Probable Maximum Precipitation (24 h PMP) maps for the Soummam watershed, which covers an area of 9200 km 2 in northeastern Algeria. A daily precipitation database was created at a 5 × 5 km resolution, based on data from 56 rain gauge stations within the watershed. These precipitation data, along with dew point temperature and wind speed data from the ERA5‐Land database, were then used to calculate 24 h PMP at the watershed scale. Two PMP estimation methods were applied: the Modified Hershfield (HM) method and the Local Storm Maximisation (LSM) method. The results illustrated in maps showed that, with the HM method, the highest values (430–571 mm) were observed in the northwest and a small area in the east, while the lowest values (137–210 mm) were located in the central and southern regions. In contrast, the LSM method yielded the highest values (630–871 mm) in the northern, eastern, and central regions. Moderate values appeared across most of the watershed, while the lowest values (130–230 mm) were confined to the southern and western areas. The LSM method, which systematically provides higher 24 h PMP estimates than the HM method, shows a ratio between the two approaches ranging from 0.6 to 3.28. Percentage differences vary from 0.11% to over 100%, with particularly marked discrepancies in the southwest and certain parts of the southeast. This notable difference between HM and LSM estimates primarily arises from the reliance of HM on statistical analysis of historical extremes, while LSM incorporates physical maximisation factors and the sensitivity to local climatic and topographic conditions. The highest 24 h PMP values are observed in the northern, southern, and eastern regions of the watershed.