Performance of Seven Gridded Precipitation Products over Arid Central Asia and Subregions

Abstract

The evaluation of gridded precipitation products is important for the region where meteorological stations are scarce. To find out the applicable gridded precipitation products in arid Central Asia (ACA) for better follow-up research, this paper evaluated the accuracy of five long-term gridded precipitation products (GPCC, CRU, MERRA-2, ERA5-Land, and PREC/L) and two short-term products (PERSIANN-CDR and GPM IMERG) on different time scales for the whole ACA and two subregions, Central Asia (CA) and Xinjiang of China (XJ). Seven evaluation indices were used to evaluate the consistency between the seven gridded precipitation products and observations at 328 meteorological stations for 40 years from 1980 to 2019. The main findings were as follows: (1) Each product can correctly reflect the trend of decreasing annual precipitation in CA and increasing annual precipitation in XJ, except for ERA5-Land. (2) GPCC captured extreme events by 75.9% for heavy rainfall and 67.9% for drought events, and GPM IMERG outperformed PERSIANN-CDR with a capture probability of 61% for heavy rainfall and 50% for drought events. (3) Annually, except for GPCC and CRU without significant deviations (BIAS < 2%), ERA5-Land, GPM IMERG, and PERSIANN-CDR generally overestimated precipitation (20% < BIAS < 60%). MERRA-2 and PREC/L underestimated precipitation, with approximately −5% for PREC/L and −20% for MERRA-2. (4) Seasonally, GPCC outperformed the other four long-term products in all seasons with the lowest BIAS (<0.93%), and GPM IMERG (BIAS < 30.88%) outperformed PERSIANN-CDR. (5) Monthly, the areas with large deviations (BIAS > 60%) for the seven products were near the Tianshan Mountains; comparatively, they performed better in CA than in XJ, with BIAS approximately 20% for CA and 40% for XJ. Despite regional differences, GPCC performed the best across the five long-term products overall, followed by CRU, MERRA-2, PREC/L, and ERA5-Land. For the two short-term products, GPM IMERG outperformed PERSIANN-CDR.

Assessing High-Resolution Precipitation Extremes in Central Asia: Evaluation and Future Projections

“…The dynamic shifts in precipitation patterns, both spatially and temporally, exert detrimental effects on agricultural systems, water resources, hydroelectric power generation, and the environment at local and global scales 51 . Comparison of spatiotemporal simulation performance between CMIP6 models and MME with ERA5 and CPC reveals significant dry bias in most CMIP6 models across Central Asia (CA) when contrasted with ERA5, consistent with findings by 52 , who noted ERA5's tendency to overestimate precipitation across CA by 20 to 60%. Conversely, 53 found ERA5 overestimating low precipitation events while underestimating high-intensity precipitation, and 13 concluded ERA5's relatively poor performance in reproducing interannual variability over Asia.…”

“…Topography plays a crucial role in shaping precipitation distribution, as seen in the spatial differences in precipitation levels 56,57 . Despite being widely used in arid CA, ERA5's accuracy falters in mountainous terrain 52,58 61,62 . While, higher resolution models show advantages in simulating extreme precipitation, especially in regions with complex topography 63 , where extreme precipitation is closely linked to convection and cloud microphysics 64 , complicating attribution of simulation differences among models across different global regions.…”

Evaluation of CMIP6 Models and Multi-Model Ensemble for Extreme Precipitation over Arid Central Asia

“…Due to the sparse site and poor temporal persistence of observation data in ACA, the application of grid data makes up for these shortcomings. Previous studies have shown that GPCC is the most accurate in the global gridded precipitation dataset for average and extreme precipitation in Central Asia, outperforming other precipitation products on both the daily and annual scales [41][42][43][44]. As a result, it is widely used for precipitation and extreme precipitation studies in Central Asia [39,40,45].…”

Drought variability, changes and hot spots across the African continent during the historical period (1928–2017)

“…GPCC spans the period 1891–2020 and is reconstructed from gauge observations from numerous stations. This inclusion ensures the evaluation of precipitation trends and variability with enhanced reliability (Dilinuer et al, 2021; Schneider et al, 2017; Song et al, 2022). Dai and Zhao (2016) also highlighted limitations in the CRU data coverage since the 1990s, emphasizing the significance of including alternative datasets like GPCC to address potential data gaps and enhance the robustness and comprehensiveness of our historical drought assessment.…”