大尺度物种丰富度与植被遥感参数时间序列特征之间的关系研究——以新疆地区为例

Biodiversity is human's essential living condition. Exploring large-scale biodiversity patterns and their drivers is the key point of current ecology research, and it also has practical significances. Most of the previous studies were focused on the impact of climate, topography,soil, productivity and other factors, while the time-series plant growth characters were rarely considered for the limited data. Current time-series remote sensing data can easily detect plant growth period and the variation in one calendar year, which provides effective data source for analyzing the impact of plant growth dynamics on diversity. Based on the Xinjiang Uygur Autonomous Region, we will analyze the plant diversity formation mechanism; besides climate, topography, geography and soil characters, the vegetation temporal indicators, for example, vegetation growth period, the annual maximum/minimum vegetation parameters, and cumulative vegetation parameters derived from the remote sensing parameters products(including vegetation index (VI), Leaf Area Index (LAI), Fraction of Photosynthetically Active Radiation (FPAR), etc. ). Vegetation temporal indicators will be introduced as factors influencing plant diversity. Regression, redundancy analysis, and neural network models will be conducted to build plant diversity driving model. In this study, we will mainly solve three important questions in model building, including the selection of optimal characters in time series, selection of proper model and consistent dataset generation in temporal and spatial scale. This research is expected to improve the explanation power of large-scale biodiversity models by fully exploiting remote sensing big data.

生物多样性是人类赖以生存的条件，探索大尺度生物多样性格局及其驱动因素是当前的研究热点并具有现实意义。目前的研究多集中于气候、地形、生产力等因素对其影响，受限于数据条件，植被生长时序特征与生物多样性的关系少有研究。时间序列遥感参数可以详细刻画植被的生长周期及季相变化等，为研究植被生长年内变化对生物多样性的影响提供了有效数据源。本研究以大尺度物种丰富度为切入点，拟选取新疆地区为试验区，在获取常规环境因子的同时，引入时间序列的植被参数产品（植被指数、LAI、FPAR等），提取年尺度上的植被生长特征，包括生长期，年内变异度等，结合地面植物多样性数据，用回归、冗余分析和神经网络模型等分析方法，探讨物种丰富度与植被遥感参数时间序列特征之间的关系。本研究拟重点解决植被时序特征参数厘定、模型优选及遥感和地面参数时空不一致问题。研究可充分挖掘遥感大数据信息，提升遥感数据在生物多样性格局驱动模型中的解释能力。

生物多样性是人类赖以生存的条件，探索大尺度生物多样性格局及其驱动因素是当前的研究热点并具有现实意义。本项目针对当前大尺度植被生长物候特征与生物多样性的关系研究缺乏的问题，选取新疆地区为试验区，构建了新疆地区物种丰富度、植被遥感参数、地形、气象等空间数据集；利用矩阵分析、空间相关等分析了物种丰富度空间格局及其尺度效应；利用探索性回归、随机森林等方法，筛选影响大尺度物种丰富度格局的主要因素，对不同植被参数（植被指数、LAI、FPAR等）在不同生境类型中对生物多样性的解释能力进行分析；同时，引入时间序列的植被参数，提取年尺度上的植被生长特征并进行优选，包括生长季均值、成熟季均值、枯落季均值、年间变化幅度均值、年间方差均值、年间累积均值等时序特征，结合空间化的物种丰富度数据，探讨了物种丰富度与植被遥感参数及其时间序列特征之间的关系。研究成果充分挖掘了遥感大数据信息，提升了遥感数据在生物多样性格局驱动模型中的解释能力。项目共发表论文12篇，联合培养博士1名，硕士1名，依托本基金申报中英合作交流项目1项，拓展了项目成果的应用和影响力。顺利完成了项目既定目标。

内容获取失败，请点击重试