耦合辐射传输模型与植被生长模型的植被覆盖度遥感反演方法研究

Fractional vegetation cover (FVC) is an important parameter for characterizing land surface vegetation coverage, and plays a key role in studies of land surface processes simulation and ecological environment assessments. However, current FVC estimation methods are mainly employed on remotely sensed data at a single time point, which lead to the FVC estimation results greatly relying on the quality of the remote sensing data and having large uncertainties. This proposal aims to develop a highly-accurate time series FVC estimation method through dynamic Bayesian network (DBN) combining multi-source information from radiative transfer model, vegetation growth model and time series remote sensing data. The main research contents of this proposal include: (1) exploring time series FVC estimation method by combining radiative transfer model and vegetation growth model, (2) constructing vegetation growth model which quantitatively describes the temporal variations of FVC, and (3) developing time series FVC estimation method from high spatial resolution remote sensing data by involving prior knowledge of dynamic variations of FVC. Based on this research, the time series FVC estimation method combining multi-source information will be built, which can solve the problem of greater reliance on the quality of remote sensing data in FVC estimation and improve FVC estimation accuracy. This study will provide data basis and technology support for improving the quantitative application levels of remote sensing data.

植被覆盖度是刻画地表植被覆盖的重要参数之一，在地表过程模拟、生态环境评估等研究中发挥着重要作用。然而，目前植被覆盖度遥感反演方法主要依靠单一时刻的遥感数据，反演结果强烈依赖于遥感数据质量，不确定性较大。本项目拟利用动态贝叶斯网络耦合辐射传输模型、植被生长模型、时间序列遥感数据等多方位信息，发展高精度的植被覆盖度时序反演算法，主要研究内容包括：（1）探索辐射传输模型与植被生长模型耦合的植被覆盖度时序反演方法；（2）构建定量描述植被覆盖度时间动态变化规律的植被生长模型；（3）发展先验知识辅助下的高空间分辨率植被覆盖度时序反演方法。通过上述研究，建立耦合多源信息的植被覆盖度时序反演算法体系，解决植被覆盖度反演过度依赖于遥感数据质量的问题，改善植被覆盖度反演精度，为提高遥感数据定量应用水平提供数据基础和技术支持。

植被覆盖度是刻画地表植被覆盖的重要参数之一，在地表过程模拟、生态环境评估等研究中发挥着重要作用。然而，目前植被覆盖度遥感反演方法主要依靠单一时刻的遥感数据，反演结果强烈依赖于遥感数据质量，不确定性较大。本项目拟利用动态贝叶斯网络耦合辐射传输模型、植被生长模型、时间序列遥感数据等多方位信息，发展高精度的植被覆盖度时序反演算法，主要研究内容包括：（1）探索辐射传输模型与植被生长模型耦合的植被覆盖度时序反演方法；（2）构建定量描述植被覆盖度时间动态变化规律的植被生长模型；（3）发展先验知识辅助下的高空间分辨率植被覆盖度时序反演方法。通过本项目的研究，研发了高效的植被生长模型构建方法，进而建立了耦合多源信息的植被覆盖度时序反演算法体系，解决了植被覆盖度反演过度依赖于遥感数据质量的问题，验证结果表明本项目研发的算法体系能够显著改善植被覆盖度反演精度，为区域生态环境监测、水土流失评价、植被退化研究等提供数据基础和技术支持。