Epsilon Archive for Student Projects

Abstract

In today´s forestry, assessments of needs for thinning are based on basal area measurements. This is an indirect measurement. The approach and the developed assessment tables are based on an assumed production forest, which implies that management proposals in stands that deviate from the basic model may be misleading. Since Walter Bitterlich in 1948 launched the idea of the relascope, several modifications have been developed for different purposes. In common is that all are based on a certain angle measurement principle. Theoretically it is possible to apply the same principle for tree crowns as well. By simply counting trees based on their height and crown length average stand-wise crown ratios can be estimated for a rapid assessment of thinning needs in non-thinned stands. The objective of this study was to find an alternative to the current methods for determining thinning needs, where direct studies the green crown rather than of the basal area would be the core idea. The studies mainly have involved assessments of the accuracy of the proposed method for estimating crown ratio.
The study began with the development of an instrument followed by a phase of measurement error assessment; the, purpose was to identify and quantify errors caused by missing or wrongly selecting trees. A second field study was carried out with the aim of determining the sampling error when using the instrument in stand surveys. For that purpose a total number of 120 sample plots was inventoried in four stands by two different plot layouts.
The results show that the crown relascope works in both theory and practice. The measurement error study presents a relatively low measurement error of 10, 6 % for height and 8, 8% for crown length, which is well comparable to the measurement error for the conventional relascope of 5-15 %. The sampling error study shows that crown length and tree height can be estimated with a precision of about 4-8 % and 4-9 % of the mean, respectively, which in turn allows crown ratio estimation with a standard error as low as 2,3 to 3 %.