全球地表辐射收支平衡是研究一系列全球环境和气候变化问题的基础，而表征其收支平衡的净辐射估算一直是国际上一个富有挑战性的研究课题。针对现有估算算法和净辐射产品的不足，本项目拟基于遥感及其他多源数据，探索地表气象因子及其他状况与净辐射的关联规律及净辐射自身的变化机制，比较现有算法的适用性和不足，在此基础上集成多源信息，运用线性回归、人工神经网络及机器学习等方法，研究并发展可适应于不同地表覆盖的全天候日地表净辐射估算模型，并对各参量对模型的影响进行敏感性分析。同时结合全球地表验证数据及现有净辐射产品，验证分析并评价各净辐射估算方法的精度。本项目还将探索定量遥感以及相关学科的一系列关键科学问题，比如测量数据的质量控制、误差传播、不同模型的评价体系、海量数据的建模理论、及尺度转换等等。

Global land surface radiation budget is the foundation for studies of on global environmental and climate changes. Estimation of surface net radiation that characterizes surface radiation budget has been a challenging topic in the world for a long time. Based on remote sensing and other ancillary data, this study will explore the dynamics and dependence of net radiation on a series of environmental factors, and evaluate existing estimation methods. Furthermore, this project will develop new statistical models for estimating the all-sky surface daily net surface, which are adaptive in the worldwide. Meanwhile, the sensitive analysis and validation of these new models will be carried out using the global observations and other net radiation products. At the same time, this project will explore some common scientific issues in quantitative remote sensing and relevant disciplines, such as quality control of ground measurements, error propagation, evaluative criteria of different methods, modeling with huge amount of data, and scaling.