Padari, A., Metslaid, S., Kangur, A., Sims, A. and Kiviste, A. 2009. Modelling Stand Mean Height in Young Naturally Regenerated Stands – a Case Study in Järvselja, Estonia.   Baltic Forestry, 15 (2): 226-236

A classic approach to obtain stand height HDq in advanced stands in Estonia has been to use regression height at the quadratic mean diameter Dq. In juvenile stands, where a considerable number of seedlings have not reached breast height, the modelling of the height distribution instead of the breast height diameter dbh distribution is more practical. Also, the field assessment methodologies differ for juvenile and advanced stands. In younger stands mean height and stem number are assessed. In advanced stands (dbh larger than 6 cm) these variables are replaced with basal area, quadratic mean diameter and regression height. This study evaluates the predictive abilities of different mean height calculation functions in various tree species in naturally regenerated stands. We analyze different young stand height calculation methods in comparison with the classical stand height HDq calculation method. Two independent sets of empirical materials were employed in our study: 27 plots from Estonian Network of Forest Research Plots (ENFRP) were used in regression analysis and three forest regeneration study areas (45 plots) at Järvselja (JS plots) were used for model validation. In the current study the r-power mean height Hr, where the exponent r value 3.627 estimated on 27 ENFRP plots) was fitted to regression height HDq. We recommend using the dominant tree selection method for measuring stand height in juvenile stands. Our results indicate that the ratio of dominant trees k and stand component cluster dominant height Hk are adequate for calculating stand height in young stands.

Key words: height distribution, mean height, dominant height, Scots pine, Norway spruce, Silver birch