Understanding global precipitation patterns is crucial for addressing climate change impacts and ensuring water and food security. This study analyzes spatiotemporal variations, stationarity, heteroscedasticity, and distribution changes in global precipitation from 1891 to 2019 using multiple statistical approaches. Spatiotemporal analysis of global precipitation variability revealed divergent trends, with 32% and 21% of global regions exhibiting statistically significant increases (p < 0.05) and decreases in precipitation respectively (Based on modified Mann–Kendall test). Precipitation trends predominantly increased, particularly pronounced at 30°N latitude. Temporal regime shifts, identified through Pettitt tests and CUSUM analysis, clustered predominantly between 1955 and 1987, during which 38% of global regions experienced significant precipitation changes. Stationarity assessments further indicated non-stationary behavior in 78% of studied regions based of ADF and KPSS tests, suggesting widespread temporal dependency in precipitation patterns. These findings collectively highlight both the spatial heterogeneity and temporal non-stationarity characterizing contemporary hydroclimatic changes. Global precipitation analysis revealed pronounced spatial heterogeneity, with only 7% of regions showing regular precipitation concentration compared to 60% with moderate concentration and 33% exhibiting irregular regimes based on PCI index. Temporal analysis identified a significant shift (p < 0.05) toward increasingly irregular patterns in recent decades, suggesting intensifying hydroclimatic variability under climate change. Examination of statistical distributions across temporal breakpoints showed significant changes in 59% of cases, particularly concentrated in northern Asia (increasing), northern/northeastern North America (mostly increasing), South America (), the Middle East (Mostly decreasing), and Africa (Mostly decreasing). These findings collectively demonstrate both increasing spatiotemporal precipitation variability and fundamental changes in statistical properties, with notable regional disparities in distributional changes. The observed heteroscedasticity patterns further underscore the non-stationary nature of contemporary precipitation regimes.

HIGHLIGHTS. The purpose of the present study is to investigate the conditional density analysis of bivariate copulas to estimate the groundwater level deficien