For those unfamiliar with regression results:

 

https://tinyurl.com/k5vqmvo (ArcGIS.com)

 

Preliminary tests to evaluate whether our data was a good fit for the regression model were all promising, but yet the coefficients came out strange. We attribute this to be a result of us using pixel area values for our area coverage of each lithology within the watershed. Perhaps if the area were calculated differently more can be taken from the model and applied to the map. 

 

Yet, there are some valuable conclusions to be made with the data. There seems to be a positive relationship between both the Intrusive and Jurrasic and many economic minerals.  The regression model ran has an average multiple R of ~ .52, so using it we can explain around 52% of the variation in concentration and relative area of each lithologic unit. 

 

As with any regression, we plan to use this model as a tool in the interpretation. By analyzing the positive and negative relationships and their relative weighting on a element-by-element basis we have a better prediction of where to look for economic prospects,