Public examination of a doctoral dissertation in the field of forestry
Doctoral candidate: M.Sc. Jussi Peuhkurinen
Date and venue: 26.8.2011, at 12 noon, BOR100, Joensuu campus
ALS-based inventory methods are replacing traditional field inventories in the production of stand-level data for operative and management purposes. Despite the advances made in the species-specific inventories of mean stand variables, ALS has not succeeded in providing accurate growing stock descriptions for operative wood procurement planning, for example, containing information on tree quality, tree size distribution and the distribution of the logs in timber assortment classes. The aim of this thesis is to evaluate and develop ALS-based methods of predicting tree size distributions and timber assortment recoveries.
The experimental work was carried out in two inventory areas both located in Eastern Finland: Matalansalo, representing a typical managed boreal forest, and Koli, located in the Koli National Park. The remote sensing material for Matalansalo consisted of low and high pulse density ALS data and digital aerial images, and the material for the Koli area of high density ALS data only.
The investigated estimation methods were individual tree delineation (ITD) and areabased statistical approach (ABSA), which were also compared within the same test areas. The performance of ITD in estimating tree size distributions and theoretical timber assortment classes was found to be better than that of the compared methods (ABSA and field assessments) in cases where individual trees could be discerned from the ALS data. In the aggregate, the different ALS methods were comparable when estimating volume and basal area, but ITD tended to produce a bias in saw log volumes and tree size distributions because of the errors in tree delineation. It was stated that the errors in both of the methods, ITD and ABSA, were in correlation with the tree size distribution and the spatial distribution of tree locations.
The estimation of theoretical and actual saw log recoveries was investigated using two area-based methods. The results of the linear regression indicated that it is possible to obtain accurate saw log recoveries using an area-based ALS method. The second method employed k-nearest neighbour imputation and harvester-collected stem data bank. The method produced species-specific saw log recoveries although the estimation accuracies were not as good as expected. The method could be improved by using a more representative stem data bank and additional search variables.
The harvester data from final cuttings was found to be suitable material for validating the diameter distributions and theoretical saw log recoveries estimated from ALS data, although there were challenges considering the delineation of the stand, tree positioning accuracy and different bucking preferences. The use of stem data bank as an auxiliary data source was more challenging because the stem data bank did not include reliable information on stand delineation and bucking parameters.
The doctoral dissertation of Jussi Peuhkurinen entitled “Estimating tree size distributions and timber assortment recoveries for wood procurement planning using airborne laser scanning” will be examined at the Faculty of Science and Forestry. The opponent in the public examination is Dr. Johannes Breidenbach, The Norwegian Forest and Landscape Institute, Ås, Norway and the custos is Professor Matti Maltamo of the University of Eastern Finland.
Photo available for download at http://www.uef.fi/vaitoskuvat
Contact: Jussi Peuhkurinen, tel. +358 50 587 5585, firstname.lastname@example.org
Publishing year: 2011Back to this years article listing