I have several maps (based on mostly on interpolation of geological borehole data) of an underground area that has been mined extensively. These maps represent ‘explanatory variables’ of factors which in combination may to lessor and greater extents have impacted on actual mining conditions experienced. They include, for example, depth, proximity to water source, rock type etc.

I also have a map of the same area which represents the ‘response variable’ - the actual mining conditions encountered. The response variable is a simple 4 class categorization from ‘good response’ to ‘very poor response.

Would the ‘surface response methodology’ be appropriate to determine a best-fit formula to predict ground response based on the given maps (historical data)? If not, are there alternative methods?

Thanks!