The relationship between soil nutrient properties and remote sensing indices in the Phaeozem region of Northeast China

This study analyzed the relationships between soil nutrient characters and remote sensing image of the farmland in the Phaeozem region of Northeast China by GIS and canonical correlation analysis. The results showed that the two sets of variables, i.e. remote sensing indices and soil nutrient indices, had significantly correlative relationship (P<;;0.05) and the first canonical variable of remote sensing indices (W1) was significantly correlated with the first canonical variable of soil nutrient properties (V1) (r=0.72). In other words, it was a feasible and efficient way to estimate soil fertilizer level by analyzing remote sensing indices. SDI ( SWIR Difference Index ) dominated W1 and V1 was directly influenced by SOC, nitrogen, potassium and coarse sand. At the same time, it indicated that SDI had significant relationship with SOC and total N, and SDI was the best index among all the remote sensing indices for evaluating soil nutrient status.