Study on Modeling and Spatial Pattern of Net Primary Production in China's Terrestrial Ecosystem

Terrestrial carbon cycle has been the focus in global change research. Simulating net primary production of terrestrial ecosystems has been very important in carbon cycle research for evaluating primary production at global or regional scale;simulating the growth of crop;estimating the spatial pattern of carbon flux and its interaction with climate;and predicting the variation of ecological environment.This study aims to simulate China's terrestrial NPP in 2003 using a remote sensing-based region-scale parametric model(C-Fix model) to explore the tempo-spatial patterns of China's terrestrial net primary production;and to reveal the mechanisms of their responses to various environmental factors in the light of resources balance.C-Fix model is advantageously simple and easy to be established.For these purposes,a national input database has been set up on a 1day-1km tempo-spatial resolution.Data used in this study includes: Leaf Area Index(LAI),daily temperature grid data,daily precipitation grid data and daily solar radiation grid data.The results show that,in 2003,China's total and average terrestrial NPP were 4.37 Pg C and 640.32 g C/(m~2·a),respectively.The spatial pattern of NPP is: NPP gradually decreased from southeast of China to northwest of China;the value of NPP are greatest in tropical rainforest and monsoon forest of south of Hainan Island,southwest and south of Yunnan province and southeast of Qing-Tibet Plateau,it is between(1 800) and(2 500) g C/(m~2·a),the maximum value is(2 743) g C/(m~2·a).The NPP value is below 100 g C/(m~2·a) in most parts of Qing-Tibet Plateau and Xinjiang province.The vegetation is sparse in Taklimakan Desert and its NPP is no more than 5 g C/(m~2·a).The total NPP value derived from this study is consistent with(2.24～6.13 Pg C) or close to results of other relevant researches,convinced this study's reliability.Moreover,the spatial pattern of NPP simulated by C-Fix model is in accordance with the results of other similar models.This indicates that C-Fix model is appropriate to be applied in China.Future work involves the consideration of the variations in the biophysical and biogeochemical attributes of different land cover types(i.e.forest,grassland,cropland and so on) to reduce the uncertainty of the model.