Study On N Fertilization Recommendation In Winter Wheat Based On Multi-sources Remote Sensing Data

Winter wheat is an important crop in China, scientific field management for wheat is the necessary conditions to ensure that its high yield and quality. Nitrogen is the nutrient elements that played a key role in the growth process of the crop, the amount of investment will promote photosynthesis of crops, will also affect the economic growth and development, ensure crop yield and quality. In order to ensure productivity, farmers always supply more than less nitrogen. Excessive N in field not only reduces N use efficiencies(NUE) so greatly and results in environment pollution, but also reduces farmers’ profit. Thus, predicting crop N status accurately and applying appropriate rate N to crop are the focus of lots of studies in agriculture.In recent years, remote sensing technology is rapid developing, because of its advantage of high frequency access, suitable for large-scale and non-destructive crops, it has become an effective, rapid and non-destructive new technical means and methods of monitoring winter wheat nitrogen nutrition situation. It is as an important part of precision agriculture, has become one of the research focus of modern agriculture. Normalized Difference Vegetation Index(NDVI) is one of the important crop canopy structure parameters, it is has an important impact on groups of light distribution, energy use and crop yields.This study attempted to reference nitrogen fertilizer optimization algorithm(NFOA) and develop the model of estimating winter wheat yield potential based on the data came from winter wheat canopy hyperspectral data which actually measured and actual yields, combining meteorological data. Then NDVI that came from remote sensing images is used to validate and analysis the model. The results show that INSEY calculated with dividing NDVI by the number of days from planting to sensing where GDD>0 is exponential relationship with actual yield(R2 =0.63, n=232), the correlation coefficients between the two is r=0.759>r(0.01,232)=0.153 and has reached a significant level. Using the fertilizer recommendation model, combining high Geo Eye-1 remote sensing data, a prescription map of variable-rate fertilization could be generated recommended fertilization for experimental plots on the national experiment station for precision agriculture in 2013, fertilizer recommendation within reasonable limits. The map can provide a meaningful reference for precision variable-rate fertilization in the field.About regional scale fertilizer recommendation, firstly, MODIS NDVI time series data from 2010 to 2014 about Shaanxi Province is reconstructed by SG filtering method, the reconstruction results meet the classification requirements, according to the typical surface features characteristic curve trends about Shaanxi Province, the differences about characteristic curve between winter wheat and other crops is determined, to extract the acreage of winter wheat in Shaanxi Province. Secondly, LAI time series is reconstructed by using S-G the envelope filter method based on time-series MODIS LAI product MOD15A2 data, and monitoring results of reviving stage winter wheat is obtained by Logistic curve fitting for the filtered data. Taking the re-greening start time of winter wheat as the foundation, the jointing start time was monitored by using MODIS LST data product MOD11A2 as well as the accumulated temperature method. Shaanxi Province has big north-south span and phenological differences, most of the winter wheat began to enter reviving stage about in late February to early March, most of the wheat area has begun to enter the jointing stage time is about at the end of March to the end of April. Re-greening start time and jointing start time of winter wheat monitored by remote sensing is broadly consistent change in trend with the observed, most of the difference in control within 5 days. It was demonstrated that the method of monitoring reviving stage and jointing stage by remote sensing is introduced into the winter wheat growing areas of Shaanxi Province is equally applicable, the method is feasible. Finally, according to winter wheat phenological phasematerial monitoring by remote sensing, study area is divided to different area, fertilizer recommendation model is applied to each regional division to obtain fertilizer recommendation prescription map, the amount of fertilizer recommended within a reasonable range. In general, it is feasible to fertilizer recommendation regional scale by using remote sensing data.