IDENTIFYING POST-FIRE DEBRIS FLOW HAZARDS IN A PRE-FIRE CONTEXT FOR THE STATE OF COLORADO

At the directive of Senate Bill 21-240, the Colorado Water Conservation Board (CWCB) initiated a statewide post-fire susceptibility study known as Wildfire Ready Watersheds (WRW). This initiative will assess the vulnerability of Colorado’s water resources and critical infrastructure to post-wildfire impacts and advance a framework for communities to plan and implement mitigation strategies to minimize fire-exacerbated risks before wildfires occur. Post-fire hazards to be evaluated include potential hydrogeomorphological hazards like flooding and debris flows. However, no spatially continuous debris flow probability estimates currently exist for the entire state.

Over the last decade, the United States Geological Survey (USGS) developed logistic regression equations to estimate debris flow likelihood in burned catchments in the intermountain West for variable rainfall intensities (Staley et al., 2016, 2017). These models are typically employed in a post-fire context, such as when a multi-agency Burned Area Emergency Response (BAER) team is activated after fire impacts federally owned land or at the specific request of an agency. To meet WRW’s need for predictive debris flow probability spatial data, the USGS’s intermountain West M1 regression equation was applied to every Colorado watershed in the United States Environmental Protection Agency’s (USEPA) Catchments data layer. Readily available gridded data products were used for terrain, soils, and rainfall model inputs. A priori burn severity estimates were inferred from the United States Forest Service’s (USFS) FlamMap model, which were subsequently transformed to numerically-continuous representations of burn severity by calibrating the inferential burn severity classes to Colorado-specific dNBR data broken out by arid and non-arid regions. Results were compared with USGS debris flow predictions for pre-existing burn scars and disagreement between the estimates and sources of uncertainty were explored. This work provides valuable insight into the methods and tools available to the natural resource managers facing increasingly severe and erratic wildfire behavior. Results and lessons learned are readily transferable to other small and large-scale pre-fire efforts in the arid, fire-prone West.