sub-project vegetation assessment

Objectives

A crucial point when comparing measured wind data with model results is to reproduce the characteristics of the site in the model as accurately as possible. For a high-resolution modelling of flows at forest edges, therefore, a survey of the stand by means of a manual forest inventory is not sufficient. For this reason, in addition to a direct individual tree survey, the survey was carried out with terrestrial laser scanners. The target variables include:

• Recording the spatial distribution of vegetation by means of terrestrial 3D laser scanning and inventory surveys of individual trees using photogrammetric methods
• Derivation of tree and stand parameters (tree height, diameter at breast height, position, DTM)
• Determination of structural parameters and extraction of parameters for the determination of resistance coefficients (e.g.: Plant Surface Area Density PAD(x,y,z))
• Mean free path length between flow obstacles (trees, branches)

Terrestrial Laser Scanning

In 2008, 2010 and 2011 the study area was recorded from different laser scanning positions. In addition to the ground scans, images were taken from lattice mast 1 (on the game field, height 40 m) and the main mast (in the stand, height 42 m), which enabled a comprehensive coverage of the canopy. Using artificial targets (spheres and cylinders), the individual scans were registered in relation to each other. The total point cloud consists of over 150 million individual points and covers an area of 450 m x 250 m. The site was surveyed in cooperation with the Chair of Meterorology.

TLS campaign

Results

In addition to the automatic determination of the tree height, the BHDs as well as tree positions of the trees along the transect, the Digital Terrain Model (DTM) is also derived from the terrestrial laser scanner data. The terrain around the measuring towers is described with a grid width of 50 cm. The point cloud along the 4 towers (500 m x 120 m) was transformed into a voxel space with 1 m edge lengths. Attributes are stored for each voxel, which has discrete voxel centre coordinates:

• the number of laser points in the voxel
• the number of laser points that were hit behind the voxel and passed through the voxel
• the number of points that led to a hit in front of the voxel (probability of a hit if there were no obstacle)
• Probability of reflection

From the normalised point densities in the volume elements, friction coefficients for the flow in the stand are to be parameterised later. The detected tree positions were connected with a Delaunay triangulation to form a route network. The mean tree spacing was determined using the triangle side lengths. Since not all trees were detected at the edge due to scan shadows, this manifests itself in long triangle sides. These sides were not included in the calculation of the mean path distance using the 3Sigma rule.

Vegetation visualisation

In cooperation with the ZIH (Centre for Information Services and High Performance Computing) of the TU-Dresden, a fly-over of the study area was created.

From the derived bamboo parameters (position, breast height diameter, tree height and crown height) of the terrestrial laser scans in combination with airborne laser scans, the stands around the wild field were modelled and visualised with the flow data of the numerical simulation.

Results

Executing

Terrestrial laser scanning: Chair of Photogrammetry in cooperation with the Chair of Meteorology

Publikationen