Canadian Forest Service Publications
Assessing the Effects of Sample Size on Parametrizing a Taper Curve Equation and the Resultant Stem-Volume Estimates. 2019. Saarinen, N., Kankare, V., Pyörälä, J., Yrttimaa, T., Liang, X., Wulder, M.A., Holopainen, M., Hyyppä, J., Vastaranta, M. Forests, 10, 848.
Year: 2019
Issued by: Pacific Forestry Centre
Catalog ID: 40014
Language: English
Availability: PDF (download)
Available from the Journal's Web site. †
DOI: 10.3390/f10100848
† This site may require a fee
Abstract
Large and comprehensive datasets, traditionally based on destructive stem analysis or other labor-intensive approaches, are commonly considered as a necessity in developing stem-volume equations. The aim here was to investigate how a decreasing number of sample trees affects parametrizing an existing taper curve equation and resultant stem-volume estimates. Furthermore, the potential of terrestrial laser scanning (TLS) in producing taper curves was examined. A TLS-based taper curve was derived for 246 Scots pines (Pinus sylvestris L.) from southern Finland to parametrize an existing taper curve equation. To assess sensitivity of the parametrization regarding sample size, the number of Scots pines included in the parametrization varied between full census and 1 Scots pine at a time. Root mean square error of stem-volume estimates remained ≤20.9% and the mean absolute difference was relatively constant (≤9.0%) between stem-volume estimates when the sample size included ≥46 Scots pines. Thus, it can be concluded that, with a rather small sample size, a taper curve equation can be re-parametrized for local conditions using point clouds from TLS to produce consistent stem-volume estimates.
Plain Language Summary
Development of stem volume equations has traditionally been based on destructive stem analysis or other labor-intensive approaches to produce required information about stem taper and volume. Large and comprehensive datasets are commonly considered as a necessity in developing these equations for achieving consistent stem volume estimates. However, variability in sample sizes for parametrizing taper curve equations has varied and there is little research to determine how much the sample size effects on the reliability of stem volume estimates. Therefore, the aim here was to investigate and assess the effect of the number of sample trees on parametrizing an existing taper curve equation and resultant estimates of stem volume. Additionally, the stem volume estimates based on Terrestrial Laser Scanning (TLS) point clouds and the re-parametrized taper curve equation were more accurate than the estimates based on the existing taper curve equation indicating appropriateness of the approach. Thus, it can be concluded that with a rather small sample size a taper curve equation can be re-parametrized for local conditions using point clouds from TLS to produce reliable stem volume estimates.
