Main content area

Practical Extension of a Lake States Tree Height Model

Bragg, Don C.
Northern journal of applied forestry 2008 v.25 no.4 pp. 186
forest trees, height, prediction, estimation, Pinus strobus, data analysis, basal area, measurement, accuracy, nonlinear models, allometry, equations, tree and stand measurements, elite trees
By adapting data from national and state champion lists and the predictions of an existing height model, an exponential function was developed to improve tree height estimation. As a case study, comparisons between the original and redesigned model were made with eastern white pine (Pinus strobus L.). For example, the heights predicted by the new design varied by centimeters from the original until the pines were more than 25 cm dbh, after which the differences increased notably. On a very good site (50-year base age site index [SI50] = 27.4 m) at the upper end of the range of basal area (BA; 68.9 m2/ha) for the region, the redesigned model predicted a champion-sized eastern white pine (actual measurements: 97.0 cm dbh, 50.9 m tall) to be 51.3 m tall, compared with 38.8 m using the original formulation under the same conditions. The NORTHWDS Individual Response Model (NIRM) individual tree model further highlighted the influence of these differences with long-term simulations of eastern white pine height. On a moderate site (SI50 = 18.7 m) with intermediate (BA = 15 m2/ha) stand density, NIRM results show that the original model consistently predicts heights to be 20-30% lower for mature white pine.