Abstract <p>This study represents a new statistical approach to controlling rain data, using a compilation of statistical methods that have not previously been used in this context. Our application has been tested on three karstic subbasins in western Algeria at different scales, from the smallest to the largest: the Khemis, Sikkak, and Mouileh Basins, subject to a semi-arid climate. Daily precipitation data from stations covering the studied basins over thirty years was used. We also brought in high-resolution reanalysis data from ERA5-Land. This provided a far more thorough spatio-temporal representation of the rainfall patterns. We have successfully demonstrated the efficacy of the statistical methods and validated them as effective new tools for monitoring critical rainfall data. The methods applied drift, estimated variance, intensity curves of duration frequency (IDF), and analysis of their climate trends allowed us to identify the behavior of precipitation at different scales of the subbasins studied and draw a table showing how precipitation changes depending on the conditions in the subbasins. This finding also confirms that the methods used are effective for analysing rain data in any type of basin, regardless of scale.</p>