Estimation Of Winter Wheat Area In Jining City, Shandong Province With MODIS Remote Sensing Data

Winter wheat is one of the main food crops in China. Accurately grasped of winter wheat area information and distribution is the great significance in ensuring the food security. Jining city is an agricultural city, and is a major grain producing areas in Shandong province. So it is very important to real-timely get the wheat planting area and spatial distribution for national and the government food security. The traditional method for getting the area of winter wheat is mainly through the statistical report. Because of the lack of field measurement and other reasons, it makes the statistics data often large deviation. Remote sensing has accuracy advantage in estimating crop area of information, so using remote sensing technique in estimating the area of winter wheat has the important practical significance.The main research goal is to real-timely estimate the area and the distribution of the winter wheat in Jining City, Shandong Province with MODIS remote sensing Data and provide reference for the local culture sector, to help the government departments to make a scientific and rational food policy. At the same time, explore the using feasibility of accurately measuring and calculation of winter wheat with remote sensing data, and provide information for remote sensing regions mapping in similar regions.In the research, we chose the Jining City as our study area, and MODIS data is the main data source. We introduce the integration of mixed-pixel decomposition and decision tree to extract the area of winter wheat in study area. Firstly, we get the characteristic of wheat’s NDVI through the analysis of NDVI curve changed in different step of the growing in different land types. Then a decision tree was designed to derive wheat pixel and analysis the result; Finally, wheat pixel was unmixed using linear mixture modeling to get a higher accuracy.In order to prove the feasibility of the method, the result by the integration of mixed-pixel decomposition technology and decision tree is compared to estimate winter wheat area by Landsat5TM image, resized and operated the result by pixel accumulation analysis. The results shows that the overall and pixel accuracy by the method presented in this paper is91.51%and80.14%respectively, and the kinds of surface features would deeply impact the precision. The decompose precision of single surface is better than complex. So it has verified the feasibility of the method.Land remote sensing mapping results in Jining City is344,900hectares, in line with local conditions. Finally, in order to know the winter wheat area information in recent years in Jining City, we apply the method to the year from2005, the results show the winter wheat area is280,000hectares in the year, which is lower than2010. Overall, with MODIS data, the area of winter wheat in Jining City by integration of mixed-pixel decomposition technology and decision tree method is feasible and is more suitable for the single planting structure than region of complex land surface.The innovation point of this paper is to real-timely get the area and the distribution of the winter wheat in Jining City, Shandong Province with MODIS Remote Sensing data. And discuss the feasibility in the estimating winter wheat information with remote sensing technology, which can provide new idea for the local government to comprehensively understand of winter wheat information.