ANALYSIS OF THE SIDE-LAP EFFECT ON FULL-WAVEFORM LIDAR DATA ACQUISITION FOR THE ESTIMATION OF FOREST STRUCTURE VARIABLES
- 1Geo-Environmental Cartography and Remote Sensing Group (CGAT), Universitat Politècnica de València, Camí de Vera s/n 46022 – Valencia, Spain
- 2Department of Cartographic Engineering, Geodesy and Photogrammetry, Universitat Politècnica de València, Camí de Vera s/n 46022 – Valencia, Spain
- 3Department of Applied Mathematics, Universitat Politècnica de València, Camí de Vera s/n 46022 – Valencia, Spain
Keywords: LiDAR full-waveform, overlap stripe differences, density changes, forest structure, canopy fuel, comparison of paired samples
Abstract. LiDAR full-waveform provides a better description of the physical and forest vertical structure properties than discrete LiDAR since it registers the full wave that interacts with the canopy. In this paper, the effect of flight line side-lap is analysed on forest structure and canopy fuel variables estimations. Differences are related to pulse density changes between flight stripe side-lap areas, varying the point density between 2.65 m−2 and 33.77 m−2 in our study area. These differences modify metrics extracted from data and therefore variable values estimated from these metrics such as forest stand variables. In order to assess this effect, 64 pairwise samples were selected in adjacent areas with similar canopy structure, but having different point densities. Two parameters were tested and evaluated to minimise this effect: voxel size and voxel value assignation testing maximum, mean, median, mode, percentiles 90 and 95.
Student’s t-test or Wilcoxon test were used for the comparison of paired samples. Moreover, the absolute value of standardised paired samples was calculated to quantify dissimilarities. It was concluded that optimizing voxel size and voxel value assignation minimised the effect of point density variations and homogenised full-waveform metrics. Height/median ratio (HTMR) and Vertical distribution ratio (VDR) had the lowest variability between different densities, and Return waveform energy (RWE) reached the best improvement with respect to initial data, being the difference between standardised paired samples 1.28 before and 0.69 after modification.