2026-10-13T00:12:59Z https://www.scidb.cn/oai

10.57760/sciencedb.j00001.01706 2026-09-20T16:23:58Z

2026-09-20 A Surface Evapotranspiration Dataset for Southwest China Based on Multi-Model Fusion doi:10.57760/sciencedb.j00001.01706 en This dataset is developed based on multi-model simulations from CMIP6 and constructed using the Bayesian Model Averaging (BMA) approach to generate a monthly gridded evapotranspiration (ET) dataset over Southwest China. The workflow is as follows. First, the performance of 28 CMIP6 models in simulating ET was systematically evaluated, and the five best-performing models were selected. Then, the high-quality SiTHv2 dataset was used as a reference, and the Quantile Delta Mapping (QDM) method was applied to bias-correct both historical simulations and future projections of the selected models, thereby reducing systematic errors and improving simulation reliability. Based on these corrected outputs, the BMA method was employed to integrate multi-model results. The final product is a spatially consistent ET dataset with a resolution of 0.1°, covering both historical and future periods. Specifically, it includes CMIP6 historical all-forcing simulations for 1982–2020 and three Shared Socioeconomic Pathway (SSP) scenarios for 2026–2100 (SSP1-2.6, SSP2-4.5, and SSP5-8.5). All data processing procedures were implemented in R. The dataset has a total size of 299 MB and consists of two types of GeoTIFF files: (1) BMA_hist_YYYY_YYYY.tif, representing monthly ET during the historical period (1982–2020), approximately 44 MB; and (2) BMA_fut_sspXXX_YYYY_YYYY.tif, representing monthly ET under future scenario XXX (2026–2100), approximately 255 MB. The spatial coverage ranges from 78.4°E to 110.2°E and 21.1°N to 36.5°N, encompassing Sichuan Province, Chongqing Municipality, Yunnan Province, Guizhou Province, and the Xizang Autonomous Region. Evapotranspiration; Southwest China; Bayesian Model Averaging; CMIP6 Zhao Qingmei Wang Genxu Sun Juying Lin Shan Hu Zhaoyong PUBLIC https://creativecommons.org/licenses/by/4.0/ dataset http://www.doi.org/10.11922/11-6035.nesdc.2025.0229.zh Science Data Bank