EP43E-04:
Observations of the Changing Spatial Patterns of Post-Wildfire Erosion and Deposition Using Terrestrial LIDAR
Abstract:
After a wildfire in forested terrain, erosion is typically much greater than background forest erosion. This is due to changes in soil erodibility and water infiltration that follow wildfire. Many prior studies have documented the changes in erosional magnitude following wildfire, but this study examines how the erosion/deposition is spatially distributed. This is only possible due to the use of terrestrial LiDAR, which allows us to track centimeter-scale changes in soil depth. With terrestrial LiDAR data we are able to show how erosion and depositional patterns change over time as the landscape recovers from the wildfire disturbance. These high-resolution data represent a critical step forward for efforts to understand the sources/sinks of erosion after wildfire.This study was conducted on a hillslope burned by the 2010 Fourmile Canyon wildfire approximately 15 km west of Boulder, CO. The first LiDAR survey was conducted within 3 weeks of the wildfire and before any significant rainstorms. Four more LiDAR surveys were conducted over a two-year period. These surveys were conducted before and after large rainstorms, so that the erosional effect of large rainstorms could be observed. The changing patterns of erosion/deposition over time were observed by differencing the topographic datasets created from each LiDAR survey.
Our results show a hysteresis in the geomorphic change. The erosional response after rainstorms was strongly influenced by the previous erosional events and pre-existing site morphology. The overall erosion pattern is patchy because much of the erosion is locally interrupted by immobile objects such as boulders, bedrock, or trees. Over the entire study period we documented that the volume of sediment eroded from hillslopes was nearly twice as high as volume eroded from convergent areas; however, when the erosion is expressed as a yield per unit area the yield from convergent areas was greater than the yield from hillslope areas.