Forest stand height determination from low point density airborne laser scanning data in Roznava Forest enterprise zone (Slovakia)

Abstract

Abstract: The presented paper discusses the potential of low point density airborne laser scanning (ALS) data for use in forestry management. Scanning was carried out in the Rožnava Forest enterprise zone, Slovakia, with a mean laser point density of 1 point per 3 m2. Data were processed in SCOP++ using the hierarchic robust filtering technique. Two DTMs were created from airborne laser scanning (ALS) and contour data and one DSM was created using ALS data. For forest stand height, two normalised DSMs (nDSMs) were created by subtraction of the DSM and DTM. The forest stand heights derived from these nDSMs and the application of maximum and mean zonal functions were compared with those contained in the current Forest Management Plan (FMP). The forest stand heights derived from these data and the application of maxima and mean zonal functions were compared with those contained in the current Forest management plan. The use of the mean function and the contour-derived DTM resulted in forest stand height being underestimated by approximately 3% for stands of densities 0.9 and 1.0, and overestimated by 6% for a stand density of 0.8. Overestimation was significantly greater for lower forest stand densities: 81% for a stand density of 0.0 and 37% for a density of 0.4, with other discrepancies ranging between 15 and 30%. Although low point density ALS should be used carefully in the determination of other forest stand parameters, this low-cost method makes it useful as a control tool for felling, measurement of disaster areas and the detection of gross errors in the FMP data. Through determination of forest stand height, tree felling in three forest stands was identified. Because of big differences between the determined forest stand height and the heights obtained from the FMP, tree felling was verified on orthoimages.