Pooling Data Improves Multimodel IDF Estimates over Median-Based IDF Estimates: Analysis over the Susquehanna and Florida

Abstract Traditional multimodel methods for estimating future changes in precipitation intensity, duration, and frequency (IDF) curves rely on mean or median of models’ IDF estimates. Such estimates are impaired by large estimation uncertainty, shadowing their efficacy planning efforts. Here, assuming that each climate model is one representation the underlying data generating process, i.e., Earth system, we propose a novel extension current through pooling data: (i) evaluate performance models simulating spatial temporal variability observed annual maximum (AMP), (ii) bias-correct pool historical AMP reasonably performing models, (iii) compute nonstationary framework from pooled data. Pooling enhances fitting extreme value distribution to assumes represent samples “true” distribution. Through Monte Carlo simulations with synthetic data, show return periods derived have smaller biases lesser uncertainty than those ensembles individual We apply this method NA-CORDEX estimate 24-h intensity–frequency (PIF) over Susquehanna watershed Florida peninsula. Our approach identifies significant at more stations compared median-based PIF The analysis suggests almost all least two-thirds peninsula will observe increases 2–100-yr periods.