Plant hydraulics substantially affects terrestrial water and carbon cycles by modulating water transport and carbon assimilation. Despite improved drought simulations in certain ecosystems through their integration into land surface models (LSMs), the broader application of plant hydraulics in diverse ecosystems and hydroclimates is still underexplored. In this study, we implemented the recently developed Noah-Multiparameterization Land Surface Model (Noah-MP LSM) equipped with a plant hydraulics scheme (Noah-MP-PHS) across 40 FLUXNET sites globally. Employing the Shuffled Complex Evolution-University of Arizona (SCE-UA) auto-calibration algorithm, we optimized key plant hydraulics parameters for these sites spanning eight vegetation types in both arid and humid climates. Noah-MP-PHS significantly improves the simulation of evapotranspiration (ET) and gross primary production (GPP) by better representing atmospheric and soil water stress compared to traditional soil hydraulic schemes (SHSs, such as Noah and CLM). The augmented Noah-MP-PHS models reduce surface flux overestimation and underestimation, exhibiting an average increase of 0.14 and 0.15 in Kling-Gupta Efficiency (KGE) compared to Noah and CLM, respectively. The explicit consideration of plant capacitance in PHS reveals substantial deep-layer and nocturnal root water uptake especially under dry conditions. We employed eXplainable Machine learning (XML) to quantify the model's relative sensitivity to newly introduced leaf-, stem- and root-related parameters in PHS. The sensitivity analysis reveals a rise in root parameter importance and a decline in leaf and stem parameters as conditions shift from humid to arid. These findings indicate that as aridity states vary, the most influential parameters affecting surface fluxes variation may change in parameter calibration for PHS applications. Our findings underscore the importance of incorporating plant hydraulics into LSMs to enhance simulations of terrestrial water and carbon dynamics. These findings are crucial for understanding ecosystem responses to global climate changes and guide the broader application of PHS at larger scales.

中文翻译：

---

通过植物水力学集成在陆地生态系统中提炼水和碳通量建模


植物水力学通过调节水的运输和碳同化，对陆地的水和碳循环产生重大影响。尽管通过将植物水力学整合到陆地表面模型 （LSM） 中，某些生态系统的干旱模拟得到了改进，但植物水力学在各种生态系统和水文气候中的更广泛应用仍未得到充分探索。在这项研究中，我们在全球 40 个 FLUXNET 站点实施了最近开发的 Noah-Multiparameterization Land Surface Model （Noah-MP LSM），该模型配备了工厂水力学方案 （Noah-MP-PHS）。采用 Shuffled Complex Evolution-University of Arizona （SCE-UA） 自动校准算法，我们优化了这些地点的关键植物水力学参数，涵盖干旱和潮湿气候下的八种植被类型。与传统的土壤水力方案 （SHS，如 Noah 和 CLM） 相比，Noah-MP-PHS 通过更好地表示大气和土壤水分胁迫，显著改进了蒸散 （ET） 和总初级生产力 （GPP） 的模拟。增强的 Noah-MP-PHS 模型减少了表面通量的高估和低估，与 Noah 和 CLM 相比，Kling-Gupta 效率 （KGE） 分别平均提高了 0.14 和 0.15。在 PHS 中明确考虑植物电容表明，尤其是在干燥条件下，深层和夜间根系水分吸收量很大。我们采用 eXplainable Machine Learning （XML） 来量化模型对 PHS 中新引入的叶、茎和根相关参数的相对敏感性。敏感性分析显示，随着条件从潮湿转变为干旱，根参数重要性上升，叶和茎参数下降。 这些发现表明，随着干旱状态的变化，影响表面通量变化的最有影响力的参数可能会在 PHS 应用的参数校准中发生变化。我们的研究结果强调了将植物水力学纳入 LSM 以增强陆地水和碳动力学模拟的重要性。这些发现对于理解生态系统对全球气候变化的响应至关重要，并指导 PHS 在更大规模上的更广泛应用。