| China’s large area of food production has the adaptability of remote sensing technology, with theimprovement of remote sensors, high-phase, high spatial resolution remote sensing data sources increased,which makes the sophisticated management of agricultural monitoring more technical support. Remotesensing monitoring of agriculture includes acreage of crops, growth of the seedlings, phonologicalinformation, yield estimation, pest and so on. In this paper, MODIS NDVI time series of images are themain data source, the growth of winter wheat in response to NDVI time series was used to extract winterwheat planting area, identifying the key phonological stage and dynamic winter wheat growth monitoringand growth of the seedlings on phonology classification. The main content and results are as follows:1. MODIS NDVI time series data affected by multiple factors such as the sensor and atmosphereduiring the satellite acquisition process, which leads to abnormal data values. The data is difficult to beused to the direct analysis of remote sensing of vegetation. In this study, Savitzky-Golay filter method wasused to reconstruct time series curve of winter wheat growth period and time series of MODIS NDVIimages. Curve reconstructed better reflected the characteristics of the two peaks and a valley in winterwheat growth, the reconstructed images had improved the outliers so that the pixel value was closer to theactual value of the ground, layed the foundation for extracting winter wheat area, phenology monitoringand growth monitoring.2. Winter wheat NDVI curve characteristic for each agricultural zoning was the basis of constructingdiscrimination rules and similarity index model for winter wheat identify in each partition agriculture.Combined these two methods of extracted winter wheat pixels, used linear spectral unmixing model to getthe winter wheat planted abundance map.Through statistical abundance of winter wheat, winter wheat areais extracted4890thousand hectares, compared to the Statistical Yearbook of the area reached92.6%accuracy, a random sampling pixel recognition accuracy of90%. Analysis results planted winter wheat inthe western hilly area including Huainan hilly area, western Henan hilly areas, mountainous areas inwestern Henan winter wheat planted area is small, large-scale cultivation is mainly concentrated in easternHenan, northern Henan Plains, Huaibei plain area, Nanyang basin region and the piedmont area, goodnatural conditions in the central region, winter wheat acreage less likely due to the adjustment of agricultural structure, production of food crops to cash crops production changes.3. Logistic model was used toextract agricultural district phenology transition time, and as a basis themaximum slope method used again to extract the distribution of the phenological transition time in Henanprovince of2011. Analysis of key phenological time distribution found in mountainous area in Huainan,Huaibei plain areas, Nanyang Basin area of winter wheat reviving appeared earlier, Yudong northernregions, the piedmont region reviving appeared late; winter wheat heading, maturity from south to northtrends over time had gradually transition period appears. By analyzing the relationship between the mainwinter wheat phenology time and location of occurrence, derived mainly winter wheat phenology time ofoccurrence was positively correlated with latitude, which increased with latitude, time of occurrencephenology went backwards, and the heat affected winter wheat growth period alternating latitude lawconsistent. The main winter wheat phenology time of occurrence was positively correlated with longitude,that is, as the longitude increases, mainly winter wheat phenology went backwards.4. Differential model in the same period correlation method was used to monitor the winter wheatgrowing every8d dynamically in Henan Province based on12images from green to heading stage, andcalculated the proportion of different grades of winter wheat growing.2011’s winter wheat growing waspoor more serious than previous years, in addition to winter wheat varieties and soil conditions, extremecold weather anomaly together with prolonged drought devastated crops. Heading winter wheat growinggradually stabilized, growing good wheat gradually increased. In addition, direct monitoring method wasused to monitor growing winter wheat in Henan province, winter wheat field sampling points of NDVIfitted a polynomial with LAI. The largest R square, the best fitting effect of cubic polynomial model wasused to inverse LAI, combining phenological recognition of the seedlings of winter wheat growing withLAI classification standards division, to better reflect the actual situation of winter wheat growing. |
