Temporal evolution of measured and simulated infiltration following wildfire in the Colorado Front Range, USA: Shifting thresholds of runoff generation and hydrologic hazards

摘要

Destructive flash floods and debris flows are a common menace following wildfire. The restoration of protection provided by forests from post-fire floods and debris flows depends on the recovery of infiltration and attendant reduction of infiltration-excess surface runoff generation. This work examines seven years of post-fire infiltration measurements and temporal relations fit to soil-hydraulic properties from the Colorado Front Range, USA, to assess infiltration recovery with increasing time since fire. Point-scale Green-Ampt simulations of infiltration across a full spectrum of rainfall events are used to evaluate infiltration changes and shifts in surface runoff generation thresholds with post-fire temporal recovery. Measured and simulated infiltration generally recovered monotonically with increasing time since fire. This indicates a reduced vulnerability to infiltration-excess runoff generation as time elapses, with the greatest risk in the first two years after the fire. The threshold for infiltration-excess runoff advances with increasing time to rainfall events with higher intensity and greater return intervals; by the third year after wildfire, only extreme events (30-100 year recurrence) generate surface runoff and by the fifth and seventh year even extreme rainfall events typically fail to generate surface runoff. Remotely-sensed vegetation indices indicate linked, or at least contemporaneous, recovery of understory vegetation and field-saturated hydraulic conductivity at this field site, suggesting coincident recovery of multiple hillslope properties impacting surface runoff generation. This work indicates the importance of coupled assessments of hillslope property recovery and stochasticity of high-intensity rainfall for flash flood and debris flow hazards arising from infiltration-excess runoff generation; these hazards likely shift to subsurface mechanisms with increasing time since fire. Post-fire hazard assessments using static hillslope properties could fail to predict flash floods and debris flows associated with infrequent extreme rainfall events that strike during the post-fire recovery period when hillslope properties are partially recovered.