© 2002 by Institute of Chartered Foresters
| ||||||||||||||||||||||||||||||||||||||||||||||||||
Height estimation in complete stem analysis using annual radial growth measurements
1 School of Environmental Science and Management, Southern Cross University, PO Box 157, Lismore, NSW, Australia
A tree annual radial growth (TARG) method for estimating the position where annual height growth ceased within a log section was developed using annual ring widths from stem analysis data. The performance of this method on height estimation in complete stem analysis was assessed against four other documented methods at 1.5 m and 3 m sampling intervals. Whilst Carmean's (1972) and Lenhart's (1972) methods gave progressively less precise but acceptable estimates at 1.5 m sampling intervals, they significantly overestimated the true tree height at 3 m sampling intervals (P < 0.05). The methods developed by Newberry (1978) and Fabbio et al. (1994) significantly overestimated the true tree heights at 1.5 m and 3 m sampling intervals. However, the new TARG method did not show any significant difference and was the most accurate at both sampling intervals. The height estimates and age data from Carmean's, Lenhart's, Newberry's and the TARG algorithms were used to fit a non-linear, height–age model using Richards' (1959) equation. The asymptotic 95 per cent confidence intervals of the parameter estimates for Newberry's method overlapped by at least 39.7 per cent with those of the other methods. Using the same parameter estimates, the Lenhart, Carmean and TARG algorithms overlapped by more than 81.0 per cent with each other at both sampling intervals. The height–age curves based on parameter estimates from the TARG method gave more accurate height estimates than those from the other methods at both sampling intervals.
Received 15 January 2001.