Extreme precipitation return levels for multiple durations on a global scale

Quantifying the magnitude and frequency of extreme precipitation events is key in translating climate observations to planning engineering design. Past efforts have mostly focused on estimation daily extremes using gauge observations. Recent development high-resolution global products, now allow extremes. This research aims quantitatively characterize spatiotemporal behavior extremes, by calculating return levels for multiple durations domain Multi-Source Weighted-Ensemble Precipitation (MSWEP) dataset. Both classical novel value distributions are used provide insight into spatial patterns Our results show that traditional Generalized Extreme Value (GEV) distribution Peak-Over-Threshold (POT) methods, which only use largest estimate not spatially coherent. The recently developed Metastatistical (MEV) distribution, includes all events, leads smoother local For 5 10 days, however, there less per year fit (37 22 average, respectively), leading larger inter-annual variability possible overestimation While GEV POT methods predict a consistent shift from heavy thin tails with increasing duration, MEV method predicts relatively constant heaviness tail any opening up an important question what ‘correct’ different durations. generated corresponding parameters provided as Global EXtremes (GPEX) These data can be useful studying underlying physical processes causing variations distributions.