耦合植被指数改进水分利用效率气孔导度遥感模型

Improvement of water use efficiency stomatal conductance remote sensing model by coupling vegetation index

水分利用效率(WUE)是深入理解生态系统碳-水循环耦合关系的重要指标,改进融合环境因子与遥感参数的WUE直接估算方法,对于评价陆地生态系统碳-水循环耦合与气候的交互具有重要的科学意义。研究针对中分辨率成像光谱仪(MODIS)遥感产品间接估算WUE精度低以及气孔导度模型(也称分析模型)适用性有限的问题,改进温带落叶阔叶林生态系统WUE的直接估算。在分析现有估算方法不确定性来源的基础上,分析了生态系统WUE及其主要调控因子的季节变化特征,进一步分析探讨了WUE对遥感监测植被覆盖以及冠层结构因子的线性和非线性响应,提出耦合遥感植被覆盖和冠层结构因子的WUE分析模型改进方案。结果表明,改进的WUEan×ScaledNDVI模型拓展了分析模型在8 d尺度的适用性,也对单因子线性模型实现了提升,在植被密度较高的森林站点,改进模型比原模型精度提升了36.98%,比MODIS间接估算WUE精度提高了31.00%。改进的分析模型弥补了8 d尺度WUE估算能力和适用性上的不足,也提升了基于植被覆盖和冠层结构因子的单因子模型的性能。

Water use efficiency(WUE) is an important index for deeply understanding the coupling relationship between ecosystem carbon and water cycles. Improving the direct estimation method of WUE based on environmental factors and remote sensing parameters has important scientific significance for evaluating the interaction between ecosystem carbon-water cycle coupling and climate.Aiming at the low accuracy of indirect estimation of WUE by MODIS and the limited applicability of stomatal conductance model(also known as analytical model),the direct estimation of WUE in temperate deciduous broad leaved forests was improved. Based on the analysis of the uncertainty sources of existing estimation methods,the seasonal variation of ecosystem WUE and its main controlling factors were analyzed. Then,the linear and nonlinear response of WUE to remotely sensed vegetation cover and canopy structure factors are further discussed. Finally,improved approaches of WUE analytical model coupling remote sensing based vegetation cover and canopy structure factors were proposed. The results show that the improved WUEan×ScaledNDVI model expands the applicability of the analytical model on the 8-day scale and improves the linear model. In the forest sites with high vegetation density,the accuracy of the improved model is 36.98% higher than that of the original model and 31.00% higher than that of the MODIS based WUE. The improved analytical model not only makes up for the shortcomings of WUE estimation at 8-day scale,but also improves the performance of the linear model based on vegetation cover and canopy structure factors.