枯落物发生发展对高寒草地叶面积指数遥感反演的影响与模拟

Standing litter is the main factor that affects leaf area index (LAI) remote sensing accuracy of alpine grassland. Getting standing litter change information accurately has an important role of improving the quality of LAI remote sensing products, quantitative analysis of land ecological system of material circulation and energy flow. Traditional method is time-consuming and laborious, and limited by spatial and temporal scale. Meanwhile, multi-spectral remote sensing is limited by observation angle, spatial resolution, spectral channels and discontinuous wide bands, which lead to relative low accuracy. This study applied unmanned aerial vehicle multi-angle hyperspectral remote sensing data to reveal the response characteristics of observation angle to the spectrum of vegetation canopies under the coverage of different standing litters through the combination of theoretical analysis and field experiments. An inversion model of standing litters from vegetation canopies was constructed based on multi-angle remote sensing. On different spatial scales, spatiotemporal change characteristics of standing litters and their influence on the remote sensing inversion of LAI in alpine grassland were studied more comprehensively by ground-air observation and satellite remote sensing inversion results. The inversion accuracy of LAI remote sensing products in the vegetation area of alpine grassland is improved, providing scientific basis for research on ecological environment assessment, productivity monitoring and land-gas interaction in pasture grassland.

枯落物是影响高寒草地叶面积指数（LAI）遥感反演精度的主要因素。准确获取草地植被枯落物含量变化信息，对改进LAI遥感产品质量，定量分析陆地生态系统物质循环和能量流动具有重要作用。传统地面观测方法费时费力，且时效性差、空间代表性有限。多光谱卫星遥感受观测角度、时空分辨率、有限的光谱通道及不连续波段的制约，反演精度较低。本项目利用无人机多角度成像高光谱遥感数据，通过理论分析与野外试验相结合的方式，揭示不同枯落物覆盖条件下，观测角度对高寒草地植被冠层光谱的响应特征；构建基于多角度遥感的草地植被枯落物反演模型；综合利用地面-航空观测试验以及卫星遥感反演结果，在不同的时空尺度上，更全面地研究枯落物发生发展对高寒草地LAI遥感反演影响的规律，改进与提高LAI遥感产品在高寒草地植被区的反演精度，为牧区草地生态环境评价、生产力监测及陆气相互作用等研究提供科学依据。

叶面积指数（LAI）是表征植被冠层结构最基本的参数之一，与植被光合、呼吸和蒸腾作用等生物物理过程关系密切，是描述土壤-植被-大气间物质和能量交换的关键因子。青藏高原高寒草地生态系统不仅是牧区经济赖以发展的生产资料，也是反馈全球气候变化的感应器。枯落物是影响高寒草地叶面积指数遥感反演精度的主要因素，因此，准确获取草地植被枯落物含量变化信息，对改进LAI遥感产品质量，定量分析陆地生态系统物质循环和能量流动具有重要作用。本研究以青藏高原东部的甘南地区作为典型研究区，利用无人机多角度成像高光谱遥感数据，通过理论分析与野外试验相结合的方式，揭示不同枯落物覆盖条件下，观测角度对高寒草地植被冠层光谱的响应特征；基于MODIS数据、生态环境因素和地面高光谱遥感数据，利用RF、SVM、ANN、Xgboost和Cubist机器学习算法分别构建了叶面积指数反演模型，并对2006-2020年甘南地区草层叶面积指数的时空动态变化进行了分析。