Abandonment is forecast using machine learning models that draw associations between weather/soil conditions, crop condition ratings and satellite data and county-level crop abandonment rates published by the RMA. Since it is not known which specific fields will be (or were) abandoned, the impacts of crop abandonment are only quantified at the draw area level. As such, the summation of acres and production from all of the individual fields within a county will generally not equal that which would be associated with that county if it were set up as its own draw area. The latter would show smaller acreage and production than the summation from its fields due to the factoring in of crop abandonment.