Item talk:Q145630
Post-wildfire landscape change and erosional processes from repeat terrestrial lidar in a steep headwater catchment, Chiricahua Mountains, Arizona, USA
Flooding and erosion after wildfires present increasing hazard as climate warms, semi-arid lands become drier, population increases, and the urban interface encroaches farther into wildlands. We quantify post-wildfire erosion in a steep, initially unchannelized, 7.5 ha headwater catchment following the 2011 Horseshoe 2 Fire in the Chiricahua Mountains of southeastern Arizona. Using time-lapse cameras, rain gauges, and repeat surveys by terrestrial laser scanner, we quantify the response of a burned landscape to subsequent precipitation events. Repeat surveys provide detailed pre-and post-rainfall measurements of landscape form associated with a range of weather events. The first post-fire precipitation led to sediment delivery equivalent to 0.017 m of erosion from hillslopes and 0.12 m of erosion from colluvial hollows. Volumetrically, 69% of sediment yield was generated from hillslope erosion and 31% was generated from gully channel establishment in colluvial hollows. Processes on hillslopes included erosion by extensive shallow overland flow, formation of rills and gullies, and generation of sediment-laden flows and possibly debris flows. Subsequent smaller rain events caused ongoing hillslope erosion and local deposition and erosion in gullies. Winter freeze-thaw led to soil expansion, likely related to frost-heaving, causing a net centimeter-scale elevation increase across soil-mantled slopes. By characterizing landscape form, the properties of near-surface materials, and measuring both precipitation and landscape change, we can improve our empirical understanding of landscape response to environmental forcing. This detailed approach to studying landscape response to wildfires may be useful in the improvement of predictive models of flood, debris flow and sedimentation hazards used in post-wildfire response assessments and land management, and may help improve process-based models of landscape evolution.
