To what extent horizontal resolution improves the simulation of precipitation in CMIP6 HighResMIP models over Southwest China?

Abstract

Southwest China (SWC) is located in the eastern part of Tibetan Plateau (TP) with large elevation differences and complex topography, which has always been a challenge to the simulation of precipitation in climate modeling community. In this study, the differences in the simulation of precipitation over the SWC are evaluated using the lower and higher resolution models (LR and HR) from the High–Resolution Model Intercomparison Project (HighResMIP) protocol in Coupled Model Intercomparison Project Phase 6 (CMIP6). Our results indicate that the spatial patterns of annual precipitation over the SWC for the period 1985–2014 are well reproduced in most of the HR and LR models, with an increasing tendency from the northwest to southeast. Compared with LR models, the wet biases over the eastern TP and the dry biases over the Sichuan Basin are significantly reduced in HR models. The bias for annual precipitation of the multi–model ensemble mean (MME) has been reduced from 0.97 mm/day (LR) to 0.72 mm/day (HR). In addition, the simulation of extreme precipitation is significantly improved in the finer horizontal resolution models, showing effectively reduced simulation biases in the Sichuan Basin compared with the LR models. The frequency and intensity of extremes are represented by heavy precipitation days (R10 mm) and maximum consecutive 5 days precipitation (Rx5day), which the relative changes have been decreased from 66% (LR) to 47% (HR) in R10 mm and decreased from 23% (LR) to 19% (HR) in Rx5day. We further examine the possible reasons for the difference between LR and HR models in precipitation simulation, showing that the HR models could generate “additional” cyclonic circulation and promote more upward motion with the water vapor convergence, thus correcting the dry biases of precipitation simulation over the Sichuan Basin. This indicates that atmospheric circulation and moisture conditions could be simulated more realistically in climate model with a finer resolution, further improving precipitation simulation performance.

Exploring the Influence of Improved Horizontal Resolution on Extreme Precipitation in Southern Africa Major River Basins: Insights from CMIP6 HighResMIP Simulations

“…This study utilized precipitation indices developed by the Expert Team on Climate Change Detection and Indices (ETCCDI; Zhang et al 2011) of the World Meteorological Organization to characterize daily precipitation extremes (Table 2). The indices employed in this study have been widely used in characterizing daily precipitation extremes (e.g., Bador et al 2023;Jin et al 2023;Reboita et al 2022;Faye and Akinsanola 2022;Ayugi et al 2021). In particular, Rx1day, R95p, and R99p are employed primarily to define intense daily rainfall, while CDD is utilized mainly to define severe daily dryness.…”

“…Recently, several studies (Ajibola et al 2020;Bador et al 2020;Strandberg et al 2021;Liang-Liang et al 2022;Chen et al 2022;Scoccimarro et al 2022;Liang et al 2023;Jin et al 2023;) have been dedicated to assessing the importance of enhancing spatial resolution in simulating precipitation characteristics using CMIP6 HighResMIP models. Several of these studies concluded that increasing the spatial resolution of GCMs within the CMIP6 HighResMIP experiment improved the model performance in simulating precipitation characteristics.…”

“…Moreover, simulating precipitation across southern Africa is a challenging task for climate models, primarily due to the influence of various global and regional processes on the region's climate (Reason et al 2006;Driver et al 2019;Desbiolles et al 2020;Samuel et al 2022). Previous studies (Zhang et al 2016;Iles et al 2020;Strandberg et al 2021;Jong et al 2023;Jin et al 2023) have highlighted that enhancing the spatial resolution of GCMs could be useful for improving their ability to simulate precipitation characteristics.…”

“…However, contrasting findings exist, with other studies indicating that an increase in spatial resolution results in deterioration in model performance. For instance, Chen et al (2022) and Jin et al (2023) reported that an increase in spatial resolution reduces biases in simulating precipitation extremes over the Tibetan Plateau and southwest China, respectively. In contrast, Bador et al (2023) reported that global climate models (GCMs) with enhanced spatial resolution under the CMIP6 HighResMIP project often exhibit lower skill than their low-resolution counterparts in simulating global land extreme precipitation.…”