基于植物光学模型的光合色素多指数协同反演方法研究

项目名称： 基于植物光学模型的光合色素多指数协同反演方法研究

项目编号： No.41201358

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 地理学

项目作者： 施润和

作者单位： 华东师范大学

项目金额： 25万元

中文摘要： 植物叶片内的光合色素与其光合作用密切相关，其在宏观尺度上对陆地生态系统与大气间的物质和能量交换有重要影响。针对当前光合色素的单一光谱指数存在抗干扰能力弱、尺度效应显著、森林地区多尺度研究不充分、类胡萝卜素物质研究缺乏等问题，本研究选取浙江天童亚热带常绿阔叶林为研究区，选取栲树等当地优势乔木为研究对象，研究一种以光谱指数法为基础，将模型模拟、敏感性分析和最优化技术相结合的光合色素多指数协同反演新方法，并将其用于叶片、冠层、地块三个尺度下的叶绿素和类胡萝卜素定量反演研究。该方法利用阔叶辐射传输模型PROSPECT5，耦合适当的冠层和土壤模型，模拟出不同尺度下与实测光谱具有良好一致性的模拟光谱；通过引入一种全局敏感性分析方法，筛选出若干具有代表性的光谱指数；通过最优化方法实现多指数间的协同，从而在一定程度上弥补单一光合色素光谱指数稳定性差的不足，探索该方法对森林植被类胡萝卜素多尺度估算的可行性。

中文关键词： 叶绿素；类胡萝卜素；光谱指数；敏感性分析；人工神经网络

英文摘要： Photosynthetic pigments existing in plant leaves are closely related to their photosynthesis processes. They have a major impact on the exchange of materials and energy between the ecosystem and the atmosphere at a large scale. This proposal aims to the present problems of using a single spectral index (SI) for photosynthetic pigments estimation, including their weak stability, distinct upscaling effects, inadequate researches at forest regions at multiple scales, and lack of carotenoid retrieval studies. The proposed research area is at Tiantong, a subtropical evergreen broadleaf forest. Castanopsis fargesii is one of the dominant trees at the area and it is also one of our objective species. Retrieval of chlorophyll and carotenoid at leaf, canopy and plot scales using a new collaboration strategy combining model simulation, sensitivity analysis and optimalizaiton techniques is the goal of this proposal. The proposed strategy is as follows: (1) To create a simulation spectra set at differect scales which show good consistency with measured ones by means of PROSPECT5, a broadleaf radiative transfer model and coupling with suitable canopy models and soil models. (2) To select several representative SIs using a global sensitivity analysis method. (3) To optimize the selected indices so as to accomplish the collabo

英文关键词： chlorophyll；carotenoid；spectral index；sensitivity analysis；artificial neural network