Nonlinear Precipitation Patterns in the Mediterranean and Middle East: Insights from ERA5 Reanalysis (1940–2025)

This study investigates the spatial, multifractal, and nonlinear characteristics of monthly precipitation totals derived from the ERA5 Reanalysis dataset, spanning January 1940 to February 2025 (1,022 months), across the Mediterranean and Middle East (30°N–45°N, 10°E–45°E). The dataset, subsampled to a 1° × 1° grid encompassing 576 grid points, captures long-term precipitation variability in a climatically diverse region shaped by topography and atmospheric dynamics. Employing k-Medoids clustering with the Haversine distance metric, we delineated five distinct precipitation zones—ranging from western Mediterranean coastal areas to eastern inland arid deserts—optimized through Silhouette Score analysis (k = 5). This clustering reflects the complex interplay of physical geographical features, such as orographic lift from the Alps and Taurus Mountains, and atmospheric controls, including jet streams and teleconnections. Multifractal Detrended Fluctuation Analysis (MF-DFA) revealed scale-dependent complexity, with multifractal spectrum widths \(\:D\alpha\:\) varying from 0.7960 in eastern arid interiors (Cluster 5) to 0.9159 in central semi-arid Mediterranean zones (Cluster 2). Mountainous regions (Cluster 3, Dα = 0.8544) exhibited pronounced multifractality, driven by terrain-induced variability and seasonal convection. Concurrently, the Brock-Dechert-Scheinkman (BDS) test confirmed pervasive nonlinearity across all clusters, yielding p-values < 0.05 and BDS statistics ranging from 33.5044 (Cluster 5) to 55.8214 (Cluster 3). These results attest to chaotic atmospheric processes, including orographic effects, convective phenomena, and teleconnections like the North Atlantic Oscillation (NAO) and Mediterranean Oscillation (MO). The analysis further elucidated spatial dependencies and upper air circulation patterns—jet streams, cyclonic/anticyclonic systems—as key modulators of precipitation regimes, validated against ERA5 reanalysis data. These findings enhance understanding of long-term precipitation variability, offering a robust framework for improving climate modeling, drought and flood forecasting, and water resource management in a region highly susceptible to climate change. By integrating advanced statistical methods with ERA5’s extensive temporal coverage, this study provides critical insights into atmospheric science, with significant implications for regional sustainability, disaster preparedness, and meteorological research.