Bias correction of a gauge-based gridded product to improve extreme precipitation analysis in the Yarlung Tsangpo-Brahmaputra River basin

Abstract

Critical gaps in the amount, quality, consistency, availability, and spatial distribution of rainfall data limit extreme precipitation analysis, and the application of gridded precipitation data is challenging because of their considerable biases. This study corrected Asian Precipitation Highly Resolved Observational Data Integration Towards Evaluation of Water Resources (APHRODITE) estimates in the Yarlung Tsangpo-Brahmaputra River basin (YBRB) using two linear and two nonlinear methods, and their influence on extreme precipitation indices was assessed by cross-validation. Bias correction greatly improved the performance of extreme precipitation analysis. The ability of four methods to correct wet-day frequency and coefficient of variation were substantially different, leading to considerable differences in extreme precipitation indices. Local intensity scaling (LOCI) and quantile-quantile mapping (QM) performed better than linear scaling (LS) and power transformation (PT). This study would provide a reference for using gridded precipitation data in extreme precipitation analysis and selecting a bias-corrected method for rainfall products in data-sparse regions