Estimation Of Chinese Terrestrial Net Primary Production Using LUE Model And MODIS Data

Based on the Light Utility Efficiency and CASA model, we set up a model to estimate the terrestrial net primary production (NPP) of China, which the 1KM MODIS data were used as main remote sensing data input in it. We estimated the terrestrial net primary production of China in 2001. And the temporal & spatial change of NPP was analyzed. The main content include that as following.1) This research completed the NPP estimation of whole China using MODIS data in China. In the same time, the APAR and the ε were also estimated. Firstly, the LUE estimation of whole China was realized in this study. Their spatial resolution is 1 KM, and the temporal resolution is per 16 days. All of them cover the whole China and all temporal stage of 2001 per 16 days.2) Using MODIS data, we completed the vegetation classification of China. Them include 21 classes, being composed of 7 forest, 3 grassland, 2 shrub, 4 cropland, 3 desert, city-land and water. Based on this classification, the APAR, ε and NPP of different vegetation types were compared. The result shown that, for the most classes, the APAR, ? and NPP of different vegetation types were apparently different (P0.001), and the e had smaller change range than APAR and NPP.3) The estimation result was compared with the MOD17A3 NPP production in five aspects. The compare indicates that our NPP estimation is obviously better than the production of MOD17 A3.4) The seasonal change of APAR? ε 和 NPP for different vegetations were compared. Their different seasonal changes indicated high coherence with the climate factor of their distribution regions, especially temperature and precipitation.5) The geographic spatial pattern of APAR?ε and NPP were analyzed by setting up the profiles along longitude and latitude per 5 degrees. The results of One Way ANOVA indicated that the changes of APAR? ε and NPP in different were remarkable(P<0.01). The lowest values of them all were found in 40N. In general, allof them increased forward South and forward North.6) About 10 parameters were abstracted for influencing analysis, which include APAR?and NPP, NDVI, precipitation, temperature and so on. The analysis indicate that the precipitation, temperature and vegetation cover are main influencing factor of APAR and NPP, the precipitation, temperature are main influencing factor of APAR and NPP.Otherwise, six parameter of model outputs and medial parameters, including APAR, NPP, albedo, EET, LUE were produced as public production for users. The total data size is about 5G.In this research, LUE model and 1km MODIS data were used for the estimation of terrestrial NPP of China. Compared with other research using AVHRR and non-remote-sensing data, our estimation is close to the most of them. In the same time, the temporal and spatial pattern were analyzed and discussed.However, some points still should be noticed. (1)Vegetation types should was considered for the calculation of land surface evapotranspiration. (2)Also, the different maximum of for the kind of biomes should be more reasonable than a constant value (0.389 gC/m2).