Model And GIS-Based On Precise Cultivation Prescription Generation Technology In Wheat

Wheat (Triticum aestivum L.) is one of the most important food crop in China. Today, precision agriculture develops rapidly, the technology of precise cultivation prescription generation bases on different variety types, ecological environments and production levels, it needs to give full play to the advantage of the precision crop management model and combine with spatial information acquisition and analysis techniques. In this study, based on years of meteorological data in Jiangsu province, the temporal and spatial distribution of major climatic resources in the wheat growth period was analysed. Considering the auxiliary variable of the nutrient content of soil, terrain attribute data, remote sensing vegetation index, four interpolation methods were used to predict the spatial distribution of soil organic matter content on the county scale. A technology system on precise cultivation prescription generation on wheat was developed, which is based on soil information and meteorological data analysis of the spatial difference at different scales. The results provide technical support for precision management of wheat.Based on the meteorological data of four historical periods including1972-1979(1970s),1980-1989(1980s),1990-1999(1990s) and2000-2008(2000s) from69meteorological stations in Jiangsu province, the major climatic resources including total sunshine hours, total growing degree days above0℃, total precipitation, total precipitation days, average daily precipitation in each site during wheat growing period were counted. The daily meteorological data during four periods in wheat growing regions were processed to generate the gridded daily meteorological surface data using ANUSPLIN and ARCGIS. The spatial and temporal distribution characteristics of the climate resources for the wheat production during wheat-growing periods between four decades(1960s to2000s)were calculated and analyzed based on the grid cell. As compared with1970s, the total sunshine hours in wheat-growing periods of2000s decreased by4.36%, the total growing degree days above0℃showed an increasing trend, the rate was about6%to22%. Average total precipitation showed an increasing trend which was modest, about1.77%. Average total precipitation days showed a decreasing trend, but the magnitude was very small, about1%. The average daily precipitation tended to increase about2.88%. The average temperature in January tended to increase about0.8℃. The results will provide management decision support for wheat production in Jiangsu Province.The study aera is situated at Rugao city in Jiangsu province, the soil nutrient data of soil sampling points in2005was analysed. Combining of the auxiliary variables composed of terrain attribute factors obtained from DEM, the vegetation index extracted by remote sensing images, the nearest distance to the river network, the SOM content was interpolated by using ordinary kriging (OK), multiple linear stepwise regression (MLSR), regression kriging(RK) and cokriging(CK). Then, the results interpolated by the different methods were evaluated. The results show that, RK method predict more accurate than the other methods. The value of the mean squared deviation ratio (MSDR=0.82) is more close to1, which means the goodness of fit of the theoretical estimate of error. Error analysis results show that, compared with OK, MLSR, CK, with Rk method the value of the coefficient of determination (R) between the simulated and measured values from182validation points is0.711which is increased0.144,0.346and0.162respectively, the value of mean absolute error(MAE) reduces0.25,0.69and0.28g.kg-1respectively, the value of root mean squared errors (RMSE) reduces0.42,0.85and0.47g.kg-1respectively. Viewing the spatial distribution maps of SOM from different interpolation methods, RK method can display more details. Combination of multi-source auxiliary variables, the spatial distribution of SOM in county scale can be predicted more accurately using RK method.On the basis of a model and GIS-based decision system for wheat precision management, By combining with the knowledge model for wheat management, spatial analysis technology for different regional scale base on geostatistics and management zone technology, a technology system for precise and quantitative cultivation prescription generation in wheat base on model and GIS was developed. The difference reflected in management unit and controlling factor of difference between management unit. This work laid a technical support for the precision management of other crops, and has provided a basic platform for more large regional scale research.