Satellite-based Vegetation Production Models of Terrestrial Ecosystem: An Overview
Wenping Yuan1, 2, Wenwen Cai1, Dan Liu1, Wenjie Dong1
1. State Key Laboratory of Earth Surface Processes and Resource Ecology,Beijing Normal University,Beijing 100875,China; 2. State Key Laboratory of Cryospheric Sciences,Cold and Arid Regions Environmental and Engineering Research Institute,The Chinese Academy of Sciences,Lanzhou,Gansu 730000,China
Vegetation,as the principal component of terrestrial ecosystem,plays an important role in sustaining global substance and energy cycle,adjusting carbon balance and alleviating the rise of atmospheric CO2 concentration and global climate change. Vegetation production of terrestrial ecosystem has been one of the major subjects for the research on global change. The satellite-based model of vegetation productivity has undergone several stages of development,including the initial simple statistical model,the later process model based on light use efficiency principle. Based on remote sensing vegetation data with spatially and temporally continuous distribution,statistical model is crucial in estimating vegetation productivity on the regional and global scale. Statistical model can be classified into two categories: one is direct establishment of the correlation between vegetation index and vegetation productivity,based on which regional estimation is possible; the other is the establishment of regression parameter vector for regional applications,which is realized through the integrated utilization of vegetation indices and other environmental factors and using regression tree,neural network and other complex statistical methods. Light use efficiency model is the major approach to estimating vegetation productivity based on remote sensing data. However,there are large differences on the calculations of the fraction of absorbed photosynthetically active radiation,environmental stress factors,and the model performance also need improve. Future studies should continue to improve model ability,develop multiple model ensemble algorithms and provide simulation uncertainties.