Logo image
Back
Environment-Conditioned Error Characterization of IMERG over Mountainous Terrain using Radar Networks
Journal article   Peer reviewed

Environment-Conditioned Error Characterization of IMERG over Mountainous Terrain using Radar Networks

Yagmur Derin, Pierre-Emmanuel Kirstetter, Dominique Faure and Daniel Watters
Journal of hydrometeorology
08/11/2026
DOI: 10.1175/JHM-D-26-0017.1

View Online

Abstract

Satellite-based precipitation products (SPPs) face persistent challenges over mountains due to complex orographic processes. This study uses high-resolution radar-gauge precipitation data and environmental reanalysis parameters to evaluate IMERG V07B Final Run (IM-F) performance across six mountain ranges (U.S. West Coast, Rockies, East Coast; French Alps, Central Massif, Pyrenees). We condition the analysis on environmental clusters defined by orographic upward motion, moisture flux convergence, and stability, to address two questions: (1) To what extent do environmental parameters explain the distribution and extremes of radar-observed precipitation in each mountain range? (2) How much of the variation in IM-F retrieval error across these mountain ranges is explained by environmental parameters? Results show that environmental conditions strongly control precipitation distributions – for instance, convective windward regimes yield the highest median and 99th-percentile rain rates, whereas stable, stratiform conditions produce much lower rainfall totals. Conditional validation reveals that a substantial portion of IM-F error is attributable to these environmental factors: precipitation detection skill and bias vary systematically with cluster conditions. Cluster 1 (unstable upslope flow) produces frequent heavy precipitation that IM-F underestimates when cloud ice is minimal, whereas Cluster 4 (stable, downslope conditions) is associated with light stratiform drizzle that IM-F often misses. These findings expose how orographic lift, moisture convergence, and stability interact to shape both precipitation extremes and satellite retrieval biases. This is the first study to systematically dissect IM-F error by environmental regimes across multiple global mountain ranges, providing novel insight into physical drivers of SPP errors. Building on these insights, we propose adaptive algorithm enhancements – for example, incorporating orographic uplift and moisture convergence indicators – to improve mountainous precipitation estimation. Our results underscore a new process-informed pathway for refining satellite QPE in complex terrain.

Details

Metrics

1 Record Views
Logo image