Deployment of a high quality DEM for landform classification
Landform maps have long been recognized as an important data source for modeling erosion, characterizing watersheds, and mapping land components for the purpose of land information management. Landform maps are also of interest to climatologists for developing climate models, especially in topoclimatology. Landform classification which delineates topographic classes was usually conducted by interpretation of aerial photographs and field survey in the past. Automatic landform classification on the bases of digital terrain models (DEMs) and geographical information systems (GIS) has attracted great attention. To some extent, the successful implementation of automatic landform classification depends on the quality of DEM available. Nowadays, increasingly available LiDAR (light detection and ranging) data offer capability of producing high accuracy and high resolution DEMs in a fast and cost-effective way. This leads a great potential to use LiDAR-derived high quality DEMs to efficiently map and classify landforms in more details.
This paper presents ways for detailed landform classification by using LiDAR-generated high quality DEM in a large Victoria Volcanic Plain (VVP) area, which is located in the south western Victoria, Australia. Selected terrain attributes such as elevation, slope, relief, curvature, break of slopes and topographic wetness index (compound topographic index) are extracted from DEM. These attributes are used as the main input for landform element classification. These classified landform elements (crests, simple slopes, depressions and flats) are combined together to produce a single landform layer in a GIS database. It can be further used to infer the relationship between landform pattern and other biophysical pattern. The study shows that landform classification results can be significantly improved by using LiDAR-derived high quality DEM.