Relative error transmission and detection in Strategic Forest Management Model

Abstract

The relative error (term error in the thesis always stands for relative error) transmission from Forest Resource Inventory (FRI) data to Strategic Forest Management Model (SFMM) outputs based on different FRI data to SFMM outputs bases on FRI survey factors such as age, stocking, height, and their combination, and species was studied. A basic input file from the Fort William Forest Management Unit was used to produce different experimental data sets which were entered into SFMMTOOL kit to generate SFMM input files. Each experimental data set was produced through modifying the basic data to make a given error rate inherent within. Through running SFMM input files of the experimental data sets, various SFMM outputs inherent error were produced, and were compared using statistical analysis technology and other analysis. It was concluded that FRI data errors such as the errors of species, age, stock, and combined errors of them could be transformed into SFMM outputs at different rates depending on the different survey factors. The results from the study indicated that species errors caused large and various SFMM output errors, depending on the original forest conditions. Age errors could cause small SFMM output errors except for the case with the age error of more than 15%. Stock errors can be transmitted into SFMM outputs at the same rate as the stock error value. Combination error can be transmitted to SFMM outputs at the same rate as the combination errors, but with a sharp increase of the rate when the combination error surpassed 20%. Age had an additive effect and interacting effect on the SFMM output errors when the combination error was equal to or greater than 20%. Based on the study, some suggestions to deal with the problems associated with FRI and SFMM application were made.