Stream Classification Using Airborne Lidar Data

  • Ms Penelope Douglas, Forestry Tasmania, Australia
  • Mr Robert Musk, Australia
  • Many mapping agencies locate streams and watercourses using air photo interpretation. Under forest canopies these data are not accurate enough to support forest management planning and sustainable silvicultural practice. In this project we employ high resolution terrain data acquired using airborne LiDAR to create a digital elevation model and a range of model derivatives including vectorised flow accumulation and normals to these vectors at 5 metre intervals. A logistic modelling approach is used to relate the field mapped locations of stream sources and stream class change points to the DEM derivatives. The model is then employed to generate a statistically defensible map product that can be used with a known level of confidence by forest planners.