The Influence of Wildfire on Hillslope Geometry

Wednesday, 17 December 2014
Francis K Rengers1, Assaf Inbar2, Gary J Sheridan2 and Petter Nyman3, (1)University of Colorado, Boulder, CO, United States, (2)The University of Melbourne, Melbourne, Australia, (3)University of Melbourne, Parkville, VIC, Australia
In southeastern Australia wildfire occurs regularly, resulting in increased hillslope erosion. However, post-wildfire erosion processes differ depending on hillslope aspect. Equatorial (north)-facing slopes are drier than polar (south)-facing slopes and experience overland flow erosion after wildfire. By contrast, overland flow is not an active process on polar-facing slopes, even after high-intensity wildfires. These differences in post-wildfire erosion processes are accompanied by observations that slope angle and curvature also differ by hillslope aspect. An airborne LiDAR dataset flown over our study area in the Kinglake National Park, Victoria shows that the mean slope angle of polar-facing slopes is nearly 5 degrees steeper than equatorial-facing slopes. We have sought to test the hypothesis that aspect differences in post-wildfire erosion processes are sufficient to create differences in hillslope geometry.

In order to test this hypothesis, we use a simple 1D model that simulates hillslope evolution over thousands of years. We limit our model to low-drainage area hillslopes where debris-flows are unlikely to occur. Erosion is modeled as nonlinear diffusion regardless of aspect during non-wildfire model years. Wildfire is modeled by changing the erosional processes on each slope aspect to reflect the effects of post-wildfire erosion according to a wildfire recurrence interval. For two years following a model wildfire we allow overland flow erosion to erode equatorial-facing slopes, whereas polar-facing slopes erode according to nonlinear diffusion for only one year following a wildfire. The erosion parameters on the polar-facing slopes are changed during this period to reflect higher post-wildfire erosion. In addition to erosional processes, we use an exponential soil production law to simulate new soil formation every model year.

Our preliminary results suggest that changes in erosional magnitude associated with the different wildfire erosional processes are sufficient to produce the observed differences in hillslope geometry over millennial timescales. This early result suggests that interactions between wildfire and local moisture regimes (i.e. aspect) may be a major driver of long-term hillslope evolution in tectonically quiescent, non-glaciated terrain.