Incorporate additional predictors into meta-forecast eval model by location into model-based evaluation

Likely additional location specific confounders that are not currently included in the planned analyses. 
Are there other predictors we might be able to include, either on the wastewater side of demographic side? 