Abstract:Nitrogen in the soil is a key element in crop growth. The development of nitrogen spatial distribution mapping system not only reflects the distribution of nitrogen in the soil intuitively, but also provides technical support for variable fertilization decisions. Based on the .NET platform, written in C#, the GIS functions such as data file type conversion,interpolation analysis and raster layer rendering were realized by ArcGIS Engine, and based on soil spectral data obtained by nearinfrared spectroscopy, the system made by Python can predict the N values through a BP neural network regression model. The integration of ArcGIS Engine and Python was achieved by adopting C# command line. The system was tested with the fiftyfour soil samples which were collected by the vehiclemounted soil parameter sensor. The results showed that the system can estimate the nitrogen content corresponding to the soil sample based on the spectral data. Both inverse distance weighted and trend surface interpolation can generate the nitrogen spatial distribution map combined with the geographical coordinates of the collection points. The IDW interpolation correlation coefficient reached 0753 which satisfied the precision requirements. And through the IDW spatial distribution map, the place where nitrogen was concentrated can be identified obviously. Through the TS spatial distribution map, the whole trend of nitrogen distribution can be mastered. In summary, the spatial distribution maps made by system can provide a theoretical basis for the formulation of corresponding variable fertilization decisions.