Navigation mit Access Keys



Vertical Vegetation Structure Classifier (VVSC): a new tool for ArcGIS

Cothereau, P.
42 Seiten

Cothereau, P., 2016: Vertical Vegetation Structure Classifier (VVSC): a new tool for ArcGIS. Master thesis. 42 p.


Cothereau P. 2016. Vertical Vegetation Structure Classifier (VVSC). A new tool for ArcGIS. Master Thesis ETHZ, D-USYS. Supervisors: J. Bolliger, C. Ginzler WSL.


Recent technical advancements in Light Detection and Ranging (LiDAR) remote sensing offers new opportunities and challenges to scientists and researchers. The possibility to measure, quantify and visualize the ecosystem structure in 3D at large spatial scales, and at relatively low costs, has allowed for new insights in our environment’s structure. The novel tool, the Vertical Vegetation Structure Classifier (VVSC) is an ArcGIS model that aggregates LiDAR data into different classes of user-defined vegetation heights (woody biomass) and provides the user with multiple information on these classes such as their area, the point density and the spatial distribution of each height classes across the landscape. The aggregation of the height classes is performed in specific user-defined regions. This allows for a broad range of ecological applications. Among many, the main goal of VVSC is to offer the user a visualization of the connectivity of the 3D vegetation structure. The division of the vegetation height categories is up to the user, as its relevance and significance depend on the user’s interest. Thereafter, VVSC will first differentiate between vegetation and buildings by classifying the LiDAR points according to the ASPRS (American Society for Photogrammetry and Remote Sensing) using LAStools from rapidlasso. Second, VVSC extracts and sorts the vertical structure of vegetation into the chosen height classes, up to a maximum of six classes. Here again, it is up to the user to choose the ASPRS classes of points he wants to include into the analysis and classification, as, occasionally, buildings represent a potential habitat for certain species (such as bats for example) and might need to be included into the process. VVSC requires only two input data sets; the first is the LiDAR data of the area of interest in the LAS or LAZ format, and the second is a shapefile in the multipoint format containing the center points of the areas of interest. The first step in the model is the creation of buffers based on the points of interests from the input shapefile. The centers of these buffers are the coordinates of the points from the input shapefile, whereas the radiuses of the different buffers are up to the user. In a second step, the buffers will be clipped off the area of interest, in order to only keep the LiDAR data that represent the inside of the buffers. Then, VVSC will filter each of these points and select only the LiDAR points that match the conditions set by the user, and then sort them into the different classes of vegetation height. Once the LiDAR points are distributed per class, VVSC will focus on producing the various outputs. Finally, the user obtains for each buffer and each vegetation height class four files; a shapefile with the distribution of the vegetation and its area, a text file with the coordinates, the intensity and the ASPRS class of each point, and two rasters to visualize the point density and the maximum point density. By using ArcScene, these two rasters allow the user to visualize the result of the point density in 3D, i.e. the area of high density of a certain vegetation height. VVSC has a large range of possible applications in the domain of spatial ecological analysis and connectivity, forestry and agroforestry such as determination of potential habitat and potential corridors, monitoring of understory regrowth and canopy growth, connectivity analysis etc. VVSC is an easily applicable tool, also for people non-used to work with geographic information systems. Therefore it has the potential to help both, researchers and scientists, towards a more effective comprehension of the recent remote sensing tools and the new possibilities they offer. Further it will help us to realize what the current challenges are and in which direction the latest progress in remote sensing need to be made. VVSC allows to derive 3D structure of woody vegetation in user defined areas for user-defined vegetation heights.