A Study On Yield Estimation And Uncertainty Of Spring Wheat Based On A Net Primary Production Model

Wheat is the most important food crops in the world,timely and accurate monitoring and forecasting on wheat yield, has important significance.Study on wheat net primary productivity and production potential,on one side can by fitting the relationship between actual yield and production potential,establish the estimation model,and the estimation precision of the model uncertainty analysis has important significance for accurate prediction of the yield of wheat;on the other hand,you can find the reason why there is a gap between potential production and the actual output of wheat,so you can improve production conditions,enhance the cultivation technology level,to maximize the advantages, explore the crop production potential,to improve the actual yield.The study on Spring Wheat in Gansu province Baiyin district as the research object,using the 2010 spring wheat growth season(4~7) key meteorological data and meteorological data of 250 m MODIS resolution to estimate the silver key area of spring wheat growth season(4~7 months) NPP,according to the relationship between spring wheat and the dry weight of NPP conversion to calculate the production potential, spring wheat yield estimation model was established by fitting the relationship between spring wheat production potential and actual output in Baiyin district,using the D-value of spring wheat production potential and actual production,the obtained of spring wheat may enhance the space map in Baiyin district,and the spring wheat area extraction uncertainty affect estimation of uncertain causes are analyzed. The main conclusions are as follows:(1) The southern area of spring wheat production potential and NPP were greater than the northern region of Baiyin district;at the same time,the net primary productivity of spring wheat(NPP) and the production potential has evident seasonal variation, consistent growth, and physiological characteristics of spring wheat.(2) Spring wheat yield estimation model has better estimation accuracy in Baiyin district,the calculated value is linear and positive correlation with actual output growth season of each month;statistical model of production potential and final yield of spring wheat in the key,that between the productivity and yield of each month is in good linear relationship. And all through the accuracy test.(3) Wuchuan County in the north of Baiyin district except for the few valley area of spring wheat may enhance the small space, central Wang Xian town of spring wheat may enhance the larger space, southeast four Silong town promotion space is small,the southern town of Shuichuan water to enhance the larger space;obtained by principal component analysis,3 principal component factors and yield of spring the wheat silver slope factors, surface rock factors and irrigation factors.(4) Spring wheat yield estimation by remote sensing in the whole process has a certain degree of uncertainty in Baiyin district,on the remote sensing images used in this study,the extraction of silver spring wheat area of uncertainty increases with the increasing of the slope,with the broken degree of surface increase with the spring wheat; pixels in the extraction process of misclassification pixels the uncertainty is greater than a missed pixel uncertainty.