Liu et al. (2025) Evaluating ISIMIP3b bias-corrected data for precipitation extremes in China during 1981–2100

Identification

• Journal: Advances in Climate Change Research
• Year: 2025
• Date: 2025-08-27
• Authors: B. Liu, Qiuhong Tang, Ximeng Xu, Siao Sun, Deliang Chen, Jinkai Luan, Huarui Ren
• DOI: 10.1016/j.accre.2025.08.005

Research Groups

• State Key Laboratory of Space Information System and Integrated Application, Beijing Institute of Satellite Information Engineering, Beijing 100086, China
• Key Laboratory of Water Cycle and Related Land Surface Processes, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China
• University of Chinese Academy of Sciences, Beijing 100049, China
• Key Laboratory of Regional Sustainable Development Modeling, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China
• Department of Earth System Science, Tsinghua University, Beijing 100084, China
• Key Laboratory of Hydrometeorological Disaster Mechanism and Warning of Ministry of Water Resources, Nanjing University of Information Science and Technology, Nanjing 210044, China
• School of Hydrology and Water Resources, Nanjing University of Information Science and Technology, Nanjing 210044, China

Short Summary

This study evaluates the accuracy of ISIMIP3b bias-corrected data in representing historical precipitation extremes in China and projects future changes, confirming its reliability for climate change impact assessments and forecasting increased heavy precipitation and fewer dry days under higher emission scenarios.

Objective

• To evaluate the accuracy of ISIMIP3b bias-corrected data in representing historical precipitation extremes in China (1981–2010) compared to observational records.
• To project future changes in extreme precipitation events across China under various Shared Socioeconomic Pathways (SSPs) and Representative Concentration Pathways (RCPs) from 2015 to 2100.
• To qualitatively compare ISIMIP3b projections with those from a broader range of Global Climate Models (GCMs) to assess the reliability and limitations of using a reduced GCM ensemble.

Study Configuration

• Spatial Scale: China, divided into eight sub-regions (Northeast, East arid, Western arid (semi-arid), North China, Southwest, Qinghai‒Tibet Plateau, Central China, and South China zones). Data resolution is 0.5° × 0.5°.
• Temporal Scale: Historical period: 1981–2010. Future projection periods: 2015–2100, with specific focus on near-future (2031–2060) and far-future (2071–2100). Daily data.

Methodology and Data

• Models used: Five GCMs from ISIMIP Phase 3b (ISIMIP3b): GFDL-ESM4, IPSL-CM6A-LR, MPI-ESM1-2-HR, MRI-ESM2-0, and UKESM1-0-LL. A multi-model ensemble average (MME) was calculated.
• Data sources:
  • ISIMIP3b bias-corrected daily precipitation data (1850–2100) under SSP1-RCP2.6 (SSP126), SSP3-RCP7.0 (SSP370), and SSP5-RCP8.5 (SSP585) scenarios.
  • Observational reference dataset: CHM_PRE daily gridded precipitation dataset (1961–2022, 0.5° × 0.5° spatial resolution), derived from 2839 station records.
  • Bias adjustment and statistical downscaling method: ISIMIP3BASD v2.5, using W5E5 v2.0 as the observational reference.
  • Extreme precipitation indices: Consecutive Dry Days (CDD), Maximum 5-day Precipitation (RX5D), Simple Daily Intensity Index (SDII), and Days of Heavy Precipitation (R10mm).
  • Evaluation metrics: Pearson correlation coefficient (r) and relative error (RE).

Main Results

• The ISIMIP3b multi-model ensemble mean (MME) accurately reproduces historical (1981–2010) average annual and extreme precipitation patterns across China. Spatial correlation coefficients (r) for average annual precipitation, CDD, R10mm, and RX5D range from 0.93 to 0.96 (p < 0.001), with relative errors (RE) between -1.3% and 6.5%. For SDII, r is 0.84–0.88 and RE is 2.1%–6.3%. The MME improves overall performance, achieving r values of 0.87–0.96 and RE of -0.2% to 4.9%.
• Future projections (2015–2100) indicate a pronounced increase in heavy precipitation intensity and frequency (RX5D, SDII, R10mm) and a decrease in consecutive dry days (CDD). These changes are more significant under higher emission scenarios (SSP585) and in the far-future period (2071–2100).
• By the end of the 21st century under the SSP585 scenario, relative to 1981–2010: CDD is projected to decline by 7.4 days (11.4%); days of heavy precipitation (R10mm) to increase by 2.5 days (15.2%); maximum 5-day precipitation (RX5D) to increase by 16.3 mm (21.7%); and simple daily intensity index (SDII) to increase by 0.7 mm/day (11.6%).
• Spatially, CDD generally declines across China, but increases are projected in Central China, South China, and the Southwest region, particularly under SSP370 and SSP585 in the far-future. Conversely, R10mm, RX5D, and SDII show widespread increases, with the Qinghai‒Tibet Plateau, Western arid (semi-arid) zone, and East arid zone exhibiting notable increases for R10mm and RX5D, and North China and the Northeast zone for SDII.
• A comparative analysis suggests that the ISIMIP3b data, despite using a smaller GCM ensemble, effectively captures key aspects of projected changes in extreme precipitation, though some regional discrepancies with broader CMIP6 projections were noted.

Contributions

• Validates the reliability and suitability of ISIMIP3b bias-corrected data, derived from a reduced GCM ensemble, for assessing historical and projecting future extreme precipitation patterns in China.
• Provides detailed spatio-temporal projections of key extreme precipitation indices (CDD, R10mm, RX5D, SDII) across China under different SSP scenarios, offering refined regional insights.
• Justifies the use of ISIMIP bias-corrected data for climate change impact assessments, particularly in regions with complex climate dynamics like China, despite its smaller GCM ensemble.
• Contributes to informing climate modeling efforts, enhancing hydrological impact assessments, and supporting strategic planning for flood risk management, water resources, and agriculture.

Funding

• National Natural Science Foundation of China (U2243226, 42222101)
• Tsinghua University (100008001)

Citation

@article{Liu2025Evaluating,
  author = {Liu, B. and Tang, Qiuhong and Xu, Ximeng and Sun, Siao and Chen, Deliang and Luan, Jinkai and Ren, Huarui},
  title = {Evaluating ISIMIP3b bias-corrected data for precipitation extremes in China during 1981–2100},
  journal = {Advances in Climate Change Research},
  year = {2025},
  doi = {10.1016/j.accre.2025.08.005},
  url = {https://doi.org/10.1016/j.accre.2025.08.005}
}