Investigation of spatial interpolation of available soil nitrogen in paddy field based on rice canopy spectral information

The objectives of the present study are to examine the relationship between soil nutrients, and rice canopy spectral indices, and to interpolate the soil nutrients by the means of Cokriging method which uses soil sample data as the variate and some rice canopy spectral indices as covariate. The area selected for this study was a 3.34 hectare paddy field on Taozhuang Farm in Jiashan county. The soil nutrients analyzed consisted of total N, total P, available K, available N, and organic matter(OM). The results showed that some rice canopy spectral indices calculated according to TM, SPOT bands at the main growing stages, especially spectral indices such as TM4/TM3, B3/B2, were well correlated with some soil nutrients such as available N, OM. The spatial pattern of soil available N in the paddy field was predicted using Cokriging which used soil sample data as the variate and ratio spectral indices TM4/TM3, B3/B2 at the tillering stage as covariate. The Cokriging estimation variances were less than those for ordinary Kriging, furthermore when the amount of soil samples was cut down or the relativity between the variate and covariate was better.