Selection of Parameters Values to Model Post-Fire Runoff and Sediment Transport at the Watershed Scale in Southwestern Forests

Abstract

Erosion and runoff have been observed to increase following fire. Land managers and Burned Area Emergency Rehabilitation (BAER) teams must be able to estimate these post-fire changes. Studies of post-fire erosion on burned watersheds show that the concentrations of sediment eroded from burned rangeland and forested hillslopes in the southwestern United States can be extremely high. Since wildfire primarily impacts soils and vegetation cover on hillslopes, it is appropriate to assume that changes in hillslope conditions will result in changes in runoff peak, volume and sediment yield. The AGWA (Automated Geospatial Watershed Assessment www.tucson.ars.ag.gov/agwa) hydrologic modeling tool employs both an empirical model (SWAT) and a more process-based model (KINEROS2). In order to study how these models should be modified to provide land managers with a means to assess the impact of fire, the models were applied on two burned watersheds. Analysis of data from the Marshall Gulch watershed near Tucson, Arizona, indicates that changes in runoff volume are small compared to changes in peak runoff. The application of the KINEROS2 model to burned conditions at the Starmer Canyon near Los Alamos, New Mexico shows a pattern of change over time that is consistent with watershed recovery. Calibrated hillslope roughness values are consistent with independent estimates for roughness under bare conditions following the fire to roughness consistent with forested conditions three years later. The modeling also indicated that increasing hillslope roughness over time accounts for much of the change in runoff response. Some of the physical changes following fire that have been identified to contribute to changes in hydrologic response include (DeBano et al. 1998):

Incorporating spatially heterogeneous infiltration capacity into hydrologic models with applications for simulating post‐wildfire debris flow initiation

“…For comparison purposes, we also employed an infiltration model based on the SCS‐CN method for a subset of simulations at our study site. The CN approach is computationally efficient, does not require infiltration data, and has been applied in other post‐wildfire settings to estimate run‐off (Canfield, Goodrich, & Burns, (); Chen, Berli, & Chief, (); Earles, Wright, Brown, & Langan, (); Foltz, Robichaud, & Rhee, ()). The CN method can be used to relate excess rainfall at time t , denoted P e ( t ), to P ( t ), the rainfall depth at time t , the potential maximum retention ( S ), and an initial abstraction ( I a =0.2 S ) through the equation (Gregoretti et al ) where S =1000/ C N −10 and t i denotes the time at which the initial abstraction has been exceeded.…”

“…In lieu of data constraining, the correlation length of K s at our study area, we performed simulations where was computed for an uncorrelated saturated hydraulic conductivity field (denoted ) and a correlated field with λ =6m (denoted ), which are within the range of correlation lengths associated with soil water repellency in other burn areas (F. Pierson et al, (); Woods et al, ()). We also perform simulations using an infiltration model based on the SCS‐CN (Cronshey, ) in order to compare a model based on with another simplified infiltration model that has been applied to burn areas (e.g., Canfield et al, (); Earles et al, ()). The timing of ground vibrations recorded by the geophones (which serve as a proxy for the actual hydrograph) provide tight constraints on flow timing, which we use to evaluate the different run‐off models.…”

Meta‐analysis of field‐saturated hydraulic conductivity recovery following wildland fire: Applications for hydrologic model parameterization and resilience assessment

“…If K fs measurements or estimates were made in the first 2 months following the wildfire, the arithmetic mean of the K fs values within the first 2 months was used to approximate . This averaging method was used to estimate for the Àgueda Basin (Shakesby et al, ), Starmer Canyon (Canfield et al, ), and Bitterroot Valley (Robichaud et al, ) sites (Table ). If no K fs measurements or estimates were made in the first 2 months after the wildfire, then a linear regression through the time origin of the first 2 years of data was used to estimate .…”

Effects of wildfires on peak discharges in watersheds

“…Post-fire measurements show that changes in peak discharges are usually larger than changes in runoff volumes (Hessling 1999, McLin et al 2001, Moody & Martin 2001, Canfield et al 2005. For this reason, the unit-area peak discharge is considered to be the most sensitive parameter for the description of the modified watershed response after a wildfire (e.g., Rowe et al 1954).…”