Study on the spatial variability and the sampling scheme of soil nutrients in the field based on GPS and GIS

With the introduction of GPS and GIS, the spatial variability of soil properties was measured and analyzed using traditional statistics method and geo-statistics. 63 sampling points were measured on a 50m grid in the soil surface (0~20cm) of a 13.3 hectare winter-wheat field and were positioned by the GPS receiver. The results show that all soil nutrients are normal distribution and represent moderate spatial variability. The organic matter, total-N, NH 4-N, and available K have strong spatial correlation, while the available P exhibits moderate spatial variability, and the range of soil properties is 265.8-349.7m. The standard errors of the 63 sampling points were respectively computed in kriging interpolation method. The results indicate that the errors increased with the decrease of the sampling points. The 50m X 100m and 100m X 100m sampling schemes were recommended for their higher precision and economical efficiency. Thus, they are preferable in the large-scale sampling.