Estimation of cropland net primary production using remote sensing methods

Net primary production (NPP) is an important parameter for the global carbon cycle research, and the remote sensing methods to estimate of it are widely concerned. In this paper, we proposed a modified CASA model to estimate NPP. Firstly, we used a new method to estimate solar radiation based on the relationship between sunset angle and sunshine hour, and then the distribution of solar radiation needed for the regional NPP calculation was acquired. Secondly, we utilized the local parameters of maximum light use efficiency 0.51gC/MJ for winter wheat and 0.66gC/MJ for summer maize specifically. Thirdly, a new water limitation factor based on LSWI (The Land Surface Water Index), which can reflect the irrigation effects, was used. Lastly, a FPAR retrieval strategy based on searching table for MODIS replaced the former CASA method. Using the common meteorological data, 8-day MODIS reflection data (MOD 09) and the distribution map of winter wheat and summer maize at Huang-Huai-Hai plain, the NPP map was acquired using the above parameterized and optimized CASA model. The yearly NPP of winter wheat is 6.028 · 10
¹³
gC, and that of summer maize is 4.132·10
¹³
gC at this region. The trend at time and space is clear. Double apex distribution of NPP at temporal scale is shown, the spatial distribution of winter wheat NPP is obviously related with latitude (r=-0.640). The value of NPP is related with climate. Winter wheat NPP is related with temperature (r=0.637) and rainfall (r=0.446), but summer maize NPP is most affected by solar radiation (r=0.662) and sunshine (r=0.614). In the different subregions of the research area, the highest winter wheat NPP (1098.89 gC/m
²
a) and summer maize NPP(805.20gC/m
²
a) both appear in eastern Shandong district, but the lowest average NPP at subregion level is in the coastal area for winter wheat (216.62gC/m
²
a) and in northern Jiangsu plain for summer maize (482.10gC/m
²
a).