POST-FIRE DEBRIS FLOW MODEL TESTING IN THE GRIZZLY CREEK FIRE, GLENWOOD CANYON, COLORADO, USA

The hazards associated with post-fire debris flows are increasing in the western U.S. as wildfire activity increases. The U.S. Geological Survey has developed a suite of tools used to assess post-fire debris flow hazards; however, we rarely have sufficient data available to test these models. In this study, we present a rare case where we did have detailed data for model testing, and it was possible to examine the debris flow prediction/response within the Grizzly Creek burn perimeter (13,000 ha) in Glenwood Canyon, Colorado, USA. This fire triggered a large number of debris flows, and the study area had a high density of new and existing rain gauges (11) and before/after lidar. These factors gave us a unique opportunity to examine the success of two operational debris-flow hazard models that predict debris flow likelihood and volumes. Rainfall data were used to determine the applicability of USGS-produced rainfall-intensity thresholds. Our observations showed strong agreement between the rain rates that triggered debris flows and the USGS rainfall thresholds in the first-year post-fire. No debris flows were observed in the second year after the fire. In addition, a map of elevation change derived from lidar was used to calculate debris flow volumes, and these were compared with the operational USGS debris flow volume model. The model-predicted debris flow volumes were 2-3 times larger than observed volumes in the first-year post-fire in most cases, suggesting opportunities for refining our volume modeling methods.