Abstract: Sampling data of oxidation belt is wide and discrete，and the traditional interpolation methods are not fitting well with the discrete data.This resourch，based on the regionalized variable and variation function of Kriging along with Inverse Distance Weighted (IDW)，put forward a new algorithm nesting the two methods.Based on investigation in a large molybdenum deposit at domestic，the conclusion is that this algorithm is simple and can greatly improve the fitting accuracy of the digital terrain model (DTM) of oxidation belt.

Key words: Oxidation belt, Kriging, IDW, DTM fitting