P-LSHv2：基于遥感反演及土壤湿度胁迫的全球逐日地表实际蒸散发数据集

准确量化土壤水分可利用性对蒸散发的影响，是提升蒸散发反演精度的关键。然而，目前大多数全球卫星蒸散发数据集未显式纳入土壤水分约束机制，导致在水资源受限区域存在显著不确定性。本研究提出一种基于分位数方法的增强型土壤水分约束方案，可有效捕捉土壤水分对植被蒸腾与土壤蒸发的影响。该方案仅依赖土壤湿度数据，避免了异质性土壤水力参数带来的不确定性。我们将此方案整合到基于过程的陆面蒸散发/热通量算法中，开发出改进版本P-LSHv2。基于全球106个通量塔观测数据，我们校准了不同生物群落与气候带的参数，并量化了多气候与土地覆盖类型下的水分约束效应。P-LSHv2在蒸散发估算上取得显著提升：均方根误差降低至0.67毫米/天，相关系数提高至0.81，尤其在干旱区表现优于前代P-LSHv1。对比分析表明，P-LSHv2在捕捉土壤水分异常对蒸散发影响方面优于Penman-Monteith-Leuning模型与全球陆地蒸发阿姆斯特丹模型，提升了全球蒸散发估算精度。基于P-LSHv2算法，我们生成了1982–2023年全球逐日蒸散发长时序数据集，为陆地水热循环与气候变化研究提供了重要数据资源。

Accurately quantifying the impact of soil water availability on evapotranspiration (ET) is critical for improving the accuracy of ET retrieval. However, most current global satellite-based evapotranspiration datasets do not explicitly incorporate soil moisture constraint mechanisms, leading to significant uncertainties in water-scarce regions. This study proposes an enhanced soil moisture constraint scheme based on the quantile method, which effectively captures the impacts of soil moisture on vegetation transpiration and soil evaporation. This scheme solely relies on soil moisture data, avoiding uncertainties introduced by heterogeneous soil hydraulic parameters. We integrated this scheme into a process-based land surface evapotranspiration/heat flux algorithm, and developed an improved version named P-LSHv2. Based on observational data from 106 global flux towers, we calibrated parameters across different biomes and climate zones, and quantified the moisture constraint effects under multiple climate and land cover types. P-LSHv2 achieves remarkable improvements in evapotranspiration estimation: the root mean square error (RMSE) is reduced to 0.67 mm/day, and the correlation coefficient is increased to 0.81, especially outperforming the previous version P-LSHv1 in arid regions. Comparative analysis shows that P-LSHv2 outperforms the Penman-Monteith-Leuning model and the Global Land Evaporation Amsterdam Model (GLEAM) in capturing the impacts of anomalous soil moisture on evapotranspiration, improving the accuracy of global evapotranspiration estimates. Based on the P-LSHv2 algorithm, we generated a long-term global daily evapotranspiration dataset spanning 1982–2023, providing an important data resource for terrestrial water and heat cycle and climate change research.