Abstract:
Based on the soil nutrient sample data collected in Zhaoyang District of Zhaotong Municipality, the content of exchangeable magnesium (Mg) of the whole district was simulated by means of the Inverse Distance Weighted (IDW), Radial Basis Function Interpolation, the Ordinary Kring and the Regression Modeling methods. It was indicated by the study that the forecasting accuracy of different simulation methods varied because of the distribution complexity of the intermediate elements. The Inverse Distance Weighted method emphasized the impact of the weight of spatial distance, it could not represent the distribution rule of Mg element very well. Forecast of the Radial Basis Function Interpolation method was not ideal either due to the great uncertainty of the data collected from the sampling points in the study area. The accuracy of the Kriging method was the highest with the least error. The Regression Modeling method created the relatively greater error in the forecast, however it could best reflect the regional characteristic and the vertical change feature of the spatial distribution of exchangeable Mg.