Can CMIP6 Models Accurately Reproduce Terrestrial Evapotranspiration Across China?

Abstract

Listen

Translate article

ABSTRACTTerrestrial evapotranspiration (ET) plays a fundamental role in the climate system. The Coupled Model Intercomparison Project Phase 6 (CMIP6) provides a valuable framework for assessing global climate model performance, but gaps remain in evaluating its ET estimates, particularly in China. To fill this gap, we employed the Global Land Evaporation Amsterdam Model (GLEAM) and the water balance ET method to validate the CMIP6 ET outputs from 1980 to 2014 at both national and river basin scales. Key findings include: (1) GLEAM ET performs comparably to the water balance method, making it reliable for validating CMIP6 ET outputs. From 1980 to 2014, the annual mean ET in GLEAM for China ranges from 355 to 411 mm/year. In contrast, most CMIP6 models overestimate ET, with the multi‐model ensemble (MME) mean ranging from 524 to 542 mm/year, showing considerable variation among models. Spatially, the MME overestimates ET across over 90% of China. Bayesian model averaging (BMA) results align closely with reference data, with overestimation concentrated in southwest China. (2) At the national scale, CMIP6 trends range from −0.36 to 0.58 mm/year2, which contrasts sharply with the GLEAM trend of 1.27 mm/year2. At the basin scale, most models overestimate annual ET compared to GLEAM, with discrepancies particularly evident in the major river basins. The smallest difference in ET trend simulation occurs in the Northwest River basin, where model distributions are more concentrated, while the largest discrepancies appear in the Pearl River basin, where model performance is more scattered. Furthermore, signal‐to‐noise ratio (SNR) analysis reveals high ensemble consistency in regions such as the Haihe, Yellow, Yangtze, Pearl and Songliao River basins, indicating more reliable model performance in these areas. This study contributes to enhancing the reliability and accuracy of climate projections, which is essential for informed decision‐making and policy formulation in atmospheric science.