陆地生态系统uWUE模型最优k值研究

Study on Optimal k Value of Underlying Water Use Efficiency Model in Terrestrial Ecosystems

摘要: 陆地生态系统的蒸散发是陆气间水、碳和能量交换的重要纽带。底层水分利用效率uWUE模型提高了对陆地生态系统碳水耦合关系的描述，基于uWUE的蒸散发分割方法也得到广泛应用。然而，对uWUE模型中的关键参数k的不确定性知之甚少。为了排除不同植被类型引起参数k的变化，本研究讨论了单一植被类型的uWUE模型参数k的最优值问题。研究结果证明了uWUE模型参数k存在不同于模型经验值的最优值k*。单一植被类型uWUE模型的指数偏离一般经验值的原因是地理环境的差异。常绿针叶林的k*在0.2~1.1之间呈多峰分布，与uWUE模型中的一般经验值(k = 0.5)存在较大偏差。经度、纬度、标高、年平均降水、年平均温度等五个地理变量对常绿针叶林的k*均存在影响。其中，高程、经度和纬度对k*的影响权重合计达80%；而年平均降水量和年平均温度的权重占20%。本研究指出了uWUE模型中的关键参数k存在较大的不确定性，揭示了参数k是基于uWUE模型的蒸散发分割方法的不确定性的重要来源。

Abstract: Evapotranspiration in terrestrial ecosystems is an important link for water, carbon and energy exchange between land and air. The new leaf scale uWUE model can better describe the coupling relationship between carbon and water in terrestrial ecosystems, and the UWU-based evapo-transpiration segmentation method has been widely used. However, little is known about the uncertainty of the key parameter k in the uWUE model. In order to exclude the variation of parameter k caused by different plant functional types, this study discussed the optimal value of parameter k of uWUE model with a single vegetation type. The results prove that there exists an optimal value k* for the parameter k of the uWUE model, which is different from the empirical value of the model. The reason why the parameter k of uWUE model for a plant functional type deviates from the general empirical value is the differences in geographical environments. The k* of evergreen coniferous forest showed a multi-peak distribution between 0.2 and 1.1, which was significantly different from the general empirical value (k = 0.5) in uWUE model. Five geographical variables, such as longitude, latitude, elevation, annual mean precipitation and annual mean temperature, all have influence on k* of evergreen coniferous forest. Among them, the influence altitude, longitude and latitude on k* reach 80%. Annual mean precipitation and annual mean temperature account for 20% of the weight. This study points out that the key parameter k in uWUE model has great uncertainty, and reveals that parameter k is an important source of uncertainty for the evapotranspiration segmentation method based on uWUE model.