Remote sensing of canadian grasslands: using vegetation biochemistry to monitor grassland health

植被生化和生物物理特性的评估有助于更好地了解草地状况，进一步提高生态系统建模的性能。从遥感数据中估计了一些关键的植被生物物理参数，并将其作为不同时空分辨率的生态系统过程模型的输入。然而，控制生态系统生物地球化学过程的植被生化变量并不总是以空间明确的格式提供，并且由于缺乏在广泛区域内具有高光谱和空间分辨率的适当数据，因此必须在模型中保持恒定。近年来，高光谱遥感技术的发展为植被生物化学特性的估算开辟了新的途径。因此，我建议探索和发展前沿的遥感方法，在多个空间尺度上估计草地的空间显性生化特性（如叶绿素浓度和类胡萝卜素浓度）。具体而言，研究将：(1)建立基于指数的方法，在冠层水平上准确估算草地生化特性；2)研究了物理模拟方法在冠层水平上准确估算草地生物化学的能力；3)利用星载高光谱数据将植被冠层尺度扩大到景观水平，评估基于指数和基于物理模型的生物化学估算对景观水平混杂因子（即暴露土壤和非光合植被）的依赖性。这项工作将解决遥感生物化学和草地生态系统方面的重大知识缺口。预计这项工作的成果将成为碳封存研究、草地生态系统可持续发展、加拿大混合和高大草地生态系统气候变化和变率影响研究的重要投入，并为利益相关者提供决策工具。

Assessment of vegetation biochemical and biophysical properties can lead to a better understanding of grassland conditions and further improve performance of ecosystem modeling. Some key vegetation biophysical parameters have been estimated from remote sensing data and used as inputs to ecosystem process models at various spatial and temporal resolutions. However, vegetation biochemical variables that control ecosystem biogeochemical processes, are not always available in a spatially-explicit format and have to be held constant in the models due to a lack of appropriate data with high spectral and spatial resolutions across a broad area. Recent developments in the field of hyperspectral remote sensing have opened new ways for estimating vegetation's biochemical properties. Consequently, I propose to explore and develop cutting-edge remote sensing methods to estimate spatially-explicit grassland biochemical properties (e.g. chlorophyll concentrations and carotenoid concentrations) at multiple spatial scales. Specifically, the research will 1) develop index-based approach to accurately estimate grassland biochemical properties at the canopy level; 2) investigate the capability of physical modelling approach to accurately estimate grassland biochemicals at the canopy level; 3) scale up from the canopy level to the landscape level using space-borne hyperspectral data and assess the dependence of the index-based and the physical modeling based biochemical estimations on confounding factors (i.e. exposed soil and non-photosynthetic vegetation) at the landscape level. The proposed work will address the significant knowledge gap concerning remote sensing biochemistry and grassland ecosystem. It is anticipated that the output from this work will be an essential input for carbon sequestration studies, grassland ecosystem sustainable development, and climate change and variability impact studies in the Canadian mixed and tall grassland ecosystems, as well as a decision making tool for stakeholders.