Collocated cokriging of iron deposit based on a model of magnetic susceptibility: a case study in Morvarid mine, Iran

This paper describes a general framework of incorporating magnetic data as prior information in the modeling of an iron deposit based on sparse drilling boreholes. Since multivariate kriging of a sparse pattern of drilling yields lower uncertainty in mineral deposit modeling, a model of magnetic susceptibility as a second variable which has an acceptable level of correlation with the first variable, iron, is used to make a model of iron grade. Before applying an inversion technique to recover a 3D model of magnetic susceptibility, the normalized full gradient (NFG) approach is used to generate an approximate visualization of probable magnetic anomaly. Based on the NFG result, we have acquired appropriate dimensions of composites to apply the geophysical inversion method. The subsurface of the desired area of the magnetic anomaly is divided into a mesh with a large number of rectangular prisms (composites) with unknown susceptibilities. The recovered magnetic susceptibility model which has a correlation equal to 0.543 with the iron grade is used as a second variable in order to apply a multivariate set of kriging equations. The collocated cokriging method as a modified version of the cokriging system is used to interpolate the iron grade based on a model of magnetic susceptibility. This method has been applied to a real case study in Zanjan Province of Iran, showing its effectiveness when we encounter with such lack of drilled boreholes in the study area.