The June 2010 Medano fire in Great Sand Dunes National Park, Colorado burned 6000 acres of the upland portion of a large watershed and was still smoldering as winter set in. Several small rainstorms in the following months had significant erosion impacts in these upper watersheds, and it is expected that spring snowmelt runoff and larger summer thunderstorms in 2011 will have more dramatic and widespread effects that may affect the visitor?s area. The early months of 2011 offer a unique opportunity to track the debris flow and sediment-laden flood responses in a wilderness area that has not been burned in more than a century. The project will measure the coupling between runoff and erosive response prior to and during vegetation re-establishment, create a predictive hazards map for the watershed, and characterize surficial materials and relate them to types of mass movement.
The proposed work will advance understanding of post wildfire erosion processes and the role of wildfire in landscape evolution, provide decision tools such as hazard maps for the National Park Service to address safety and other concerns, and provide an educational component for the general public through materials provided in the Park.
The purpose of this project was to monitor and assess debris-flow (also known as "mudslide") hazards following the 2010 Medano Fire at Great Sand Dunes National Park and Preserve in Colorado. Using previously published hazard assessment models, we created a set of hazard maps for use by Park management (see attached figure). These maps confirmed some aspects of the published models, but also noted conditions under which the models did not perform well: namely, that the hillside and channel sediments in the Medano Fire area contain a significant component of fine windblown sand from the dunes that influences the local flooding and sediment-transport processes. Monitoring of rainfall (using continuously-recording rain gauges) and debris-flow occurence (using pressure transducers drilled into bedrock in stream channels) showed that for this geologic setting and following wildfire, the shorter the measurement of a rainfall intensity burst, the more distinction there is between bursts that produce debris flows and those that do not. Ten minute intensity measurements seem to be sufficiently short to distinguish the reactions. The lag between high intensity episodes and the occurrence of debris flows is less than 10 minutes, and is frequently less than 5 minutes (see attached figure). Monitoring also demonstrated recovery of the drainage basin following the wildfire in 2010. During the monitoring period May 25 – October 11, 2011, 47 storms were recorded, producing at least 8 debris flows from 6 different drainage basins. During the monitoring period April 26 – November 30, 2012, 31 storms were recorded, 4 of which exceeded the previously established duration-intensity threshold, but produced no debris flows. This indicates substantial recovery of the basin towards pre-fire conditions, within a two-year period following the wildfire.