Random Errors and Strategic Planning in Forestry

1993

24 S.

15 Lit. Ang.

Bandaufführung

2693

37969

Random errors of a data material might be described as the standard deviation for the differences between the sampled values and the true values. Based on this, random variates for site quality, basal area and dominant height have been generated for an empirical study. A calculation model, which combines simulations and linear programming, is applied to evaluate the effects on strategic decision variables in forest management planning. The empirical studies show that if the random errors for the input variables are 10%, the random errors for such decision variables as fellings and net present value vary from 2% to 6%. For site quality, the effect of random errors might, in a few cases, even be more serious than the effect of systematic errors. An evaluation of systematic effects caused by random errors shows that this is an insignificant problem with the applied calculation model. The level of the random errors, and accordingly also the effects, are probably somewhat larger in practical Norwegian forest management plans than in the empirical study of this work.