Exploring the influence of improved horizontal resolution on extreme precipitation in Southern Africa major river basins: insights from CMIP6 HighResMIP simulations

This study investigates the impact of enhanced horizontal resolution on simulating mean and extreme precipitation in the major river basins of southern Africa. Seven global climate models (GCMs) from the Coupled Model Intercomparison Project Phase 6 (CMIP6) High-Resolution Model Intercomparison Project (HighResMIP) are used, which are available at both high-resolution (HR) and low-resolution (LR). The models are assessed using three observational datasets from 1983–2014 during December-January–February. The performance of the models in simulating nine extreme precipitation indices, as defined by the Expert Team on Climate Change Detection and Indices (ETCCDI), is quantitatively assessed using various of statistical metrics. Results show that the distributions of daily precipitation from the HR models are nearly identical to those of their LR counterparts. However, model biases are not consistent across the three observations. Most HR and LR models reasonably simulate mean precipitation, maximum consecutive dry and wet days (CDD and CWD), number of rainy days (RR1) and heavy precipitation events (R10mm and R20mm), albeit with some biases. Enhanced horizontal resolution improves the simulation of mean precipitation, CDD, CWD, RR1 and R10mm, as indicated by high spatial correlation coefficients (SCCs), low root mean square errors (RMSEs), and reduced biases in most HR models. The majority of the HighResMIP models (i.e., both LR and HR models) overestimates extreme wet days precipitation (R95p and R99p), maximum one-day precipitation (Rx1day), and simple daily intensity (SDII), with a pronounced wet bias in HR models for R95p and R99p. Most LR models outperforms HR models in simulating R95p, R99p, and SDII. By means of a Comprehensive Ranking Metrics, EC-EARTH_HR is identified as the best performing model for simulating all nine extreme precipitation indices across the basins, except for the Zambezi, where EC-EARTH_LR performs best. Our findings indicate that increased resolution can either improve or worsen performance depending on the model and basin. Therefore, no clear evidence exist that enhanced horizontal resolution under HighResMIP enhances the simulation of extreme precipitation in southern Africa.

SAMUEL Sydney;  TSIDU Gizaw Mengistu;  DOSIO Alessandro;  MPHALE Kgakgamatso; 

2024-10-03

SPRINGER

JRC137093

0930-7575 (online),   

https://link.springer.com/article/10.1007/s00382-024-07325-9,    https://publications.jrc.ec.europa.eu/repository/handle/JRC137093,   

10.1007/s00382-024-07325-9 (online),