The effect of data quality on short-term growth model projections

This study was designed to determine the effect of FIA's data quality on short-term growth model projections. The data from Georgia's 1996 statewide survey were used for the Southern variant of the Forest Vegetation Simulator to predict Georgia's first annual panel. The effect of several data error sources on growth modeling prediction errors was determined, including the effect of site index measurement errors. The study suggests that for tree attributes, such as volume by species-diameter class combinations, data quality will be the largest source of prediction error. For plot attributes, site index measurement errors will be the largest source of prediction error.