Epsilon Archive for Student Projects

Abstract

Bergvik Skog is one of the leading forest companies in Sweden who use airborne laser scanning for forestry information. They utilize laser data retrieved by the Swedish National Land Survey to describe the forest holdings and facilitate planning for thinning and final felling operations. However, the possibility to identify stands with clearing needs by using laser scanning data has not been tested in a large scale. The data used in the current study are field data from The Swedish National Forest Inventory in combination with laser data from the Swedish National Land Survey acquired in 2009. The collected data is used to find one or several laser variables that best describe the relationship of the field measured samples with clearing needs. Simple and multiple linear regression and logistic regression were used to examine the best laser variables who describes clearing needs best. These regression functions were used to create raster layers which describe the clearing stands. Within these grid cells an average value was counted for each cell and it also gave the cell its information. This was then compared with the resulting model with Bergvik Skogs holding records to verify how well the new model describes a population with clearing needs.

The result shows that the data used in the study estimates the height, volume, basal area and have an explanation degree (R2 -adj) of 89 %, 86 %, and 76 % which is good and comparable with similar studies. The model for the needs of clearing shows that approximately 70 % of the cases identify the clearing stands within the test area. The model also shows where the needs of clearing are larger in the stands and can therefore be used to support the planning of clearing stands.