Evaluating FVS Predictions in Red Pine Stands at Cloquet, MN

Abstract

Data from a remeasured Pinus resinosa (red pine) spacing study were used to compare Forest

Vegetation Simulator (FVS) predictions of diameter at breast height (DBH) and volume growth at both

the stand and the individual tree level. Four scenarios were run, differing by whether FVS was used to

predict DBH growth, height and/or volume. The results from the scenarios differed depending on

whether the evaluation was at the stand or tree level. Using FVS heights instead of modelled heights

produced a negative bias in volume prediction at the tree and stand level. Using FVS predicted DBHs

produced a very large error in volume and DBH at the tree level, but its effect at the stand level varied

by spacing; FVS usually over-predicted DBH growth. This led to over-prediction of volumes at

narrower spacings. This was particularly true for trees of smaller DBH. FVS did not show bias in the

middle spacings. The bias across all spacings and reps was 1.3% of initial DBH, and 8.6% of mean

DBH growth. Across all spacings, FVS underestimated the spread of DBH growth.