The application of gridded precipitation datasets as a substitute of limited ground observation over mountainous regions is challenging due to considerable biases and needs adjustments before their application in subsequent impact models. In this study, four commonly used precipitation bias correction (BC) methods were evaluated for their skills to capture various aspects of extreme precipitation over Upper Jhelum Basin (UJB) for a period of 34 years (1981–2014). The four BC methods, i.e., linear scaling (LS), local scaling intensity (LOCI), power transmission (PT), and distribution mapping (DM), were applied on ERA5 reanalysis precipitation dataset and evaluated using nine extreme precipitation indices. First, it was found that the raw/original ERA5 overestimates observed precipitation and number of wet days with little precipitation and thus inevitability needs correction in raw estimates. Second, more or less all BC methods improved the raw ERA5 estimates especially magnitude; however, clear discrepancies exist in their skills to correct wet day frequency. Overall, the DM method was found to be a good compromise to correct various aspects of extreme precipitation, followed by LOCI, PT, and LS methods. This study provides twofold potential benefits; firstly, extreme precipitation information tailored the need of relevant decision makers to devise appropriate mitigation and adaptation strategies, and secondly, provides a certain reference for evaluation, correction, and application of gridded datasets for extreme precipitation analysis in data-sparse region.
Performance evaluation of raw and bias-corrected ERA5 precipitation data with respect to extreme precipitation analysis: case study in Upper Jhelum Basin, South Asia
Ansari R.
;Grossi G.
2022-01-01
Abstract
The application of gridded precipitation datasets as a substitute of limited ground observation over mountainous regions is challenging due to considerable biases and needs adjustments before their application in subsequent impact models. In this study, four commonly used precipitation bias correction (BC) methods were evaluated for their skills to capture various aspects of extreme precipitation over Upper Jhelum Basin (UJB) for a period of 34 years (1981–2014). The four BC methods, i.e., linear scaling (LS), local scaling intensity (LOCI), power transmission (PT), and distribution mapping (DM), were applied on ERA5 reanalysis precipitation dataset and evaluated using nine extreme precipitation indices. First, it was found that the raw/original ERA5 overestimates observed precipitation and number of wet days with little precipitation and thus inevitability needs correction in raw estimates. Second, more or less all BC methods improved the raw ERA5 estimates especially magnitude; however, clear discrepancies exist in their skills to correct wet day frequency. Overall, the DM method was found to be a good compromise to correct various aspects of extreme precipitation, followed by LOCI, PT, and LS methods. This study provides twofold potential benefits; firstly, extreme precipitation information tailored the need of relevant decision makers to devise appropriate mitigation and adaptation strategies, and secondly, provides a certain reference for evaluation, correction, and application of gridded datasets for extreme precipitation analysis in data-sparse region.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
