Vetter, M., Höfle, B., Hollaus, M., Gschöpf, C., Mandlburger, G., Pfeifer, N. and Wagner, W. 2011. Vertical vegetation structure analysis and hydraulic roughness determination using dense ALS point cloud data - a voxel based approach. ISPRS Workshop Laser Scanning 2011, 29-31 August, University of Calgary, Alberta, Canada.
Doctoral Programme on Water Resource Systems
Karlsplatz 13/222, A-1040 Vienna, Austria
Identifying vertical vegetation structure with remote sensing
Remote sensing is a fast and accurate way to collect real-time information about the earth’s surface. Many creative applications are emerging that can help scientists understand water processes.
Airborne laser scanning is an efficient and cost-effective method to gain accurate data for large areas. Laser pulses are able to penetrate through the gaps in vegetation foliage. A dense laser data set shows a three dimensional image of the vegetation such as trees, shrubs and grassland, as well as showing the land surface under the vegetation. This information on the texture of the surface translates into “surface roughness”, i.e. how much the surface deviates from being smooth. The roughness of a floodplain has a strong influence on how water will move over the land surface during a flood event. It is therefore an important parameter for hydrologists interested in understanding water movement during a flood.
Surface roughness helps hydrologists model water flows
Michael Vetter and his co-authors were interested in determining how laser scanning data could be applied to give information to hydrologists on surface roughness. They did this by dividing up the laser data set into cells (1m by 1m), voxels (cubes of 1m by 1m at an equal height 0.5m), and connections (groups of voxels that usually represent a single plant or tree). Specific criteria and corresponding roughness scores were developed (such as maximum height of lowest connection less than 0.15m) and applied to the data set. The resulting estimates of surface roughness can be easily applied to a hydraulic model to more accurately simulate flood water flow over an inundated surface. A major advantage of using laser scanned data is that information on the surface geometry (i.e. slope gradient, depressions and ridges) is also provided. And that this data corresponds to exactly the same point in time as the information on the vegetation and surface roughness. Until now, corresponding data sets for the same point in time had not been available. This data resource is a major benefit for hydraulic modellers.