Exploring Post-Wildfire Hydrologic Response in Central Colorado Using Field Observations and the Landlab Modeling Framework.

Friday, 19 December 2014
Jordan Marie Adams, Tulane University, New Orleans, LA, United States, Francis K Rengers, University of Colorado, Boulder, CO, United States, Nicole M Gasparini, Tulane University of Louisiana, New Orleans, LA, United States, Gregory E Tucker, Univ Colorado, Boulder, CO, United States, Sai Siddhartha Nudurupati, University of Washington, Seattle, WA, United States, Daniel E. J. Hobley, Univ of Colorado, Boulder, CO, United States, Erkan Istanbulluoglu, Univ of Washington, Seattle, WA, United States and Eric W.H. Hutton, Community Surface Dynamics Modeling System, Boulder, CO, United States
In May 1996, the Buffalo Creek fire burned nearly 5,000 hectares of National Forest and private land 60 miles southwest of Denver, Colorado. Several weeks later, convective storms initiated severe flooding in the Buffalo and Spring Creek watersheds, leading to significant overland and fluvial sediment transport. To assess landscape response to fire, the U.S. Geological Survey monitored hydrologic and sedimentologic conditions between 1997 and 2000. In addition to precipitation, discharge and grain size distribution measurements over the study period, ground surveys and photogrammetry were done at representative cross-sections near the downstream channel outlet to estimate volumes of transported material after different precipitation events. These data can be used to validate numerical models that simulate landscape hydrologic and erosional responses post-fire. Modeling is a critical tool in understanding landscape response to fire across these short time scales, as post-fire erosion events can disrupt steady-state landscapes, subsequently affecting short and long-term landscape evolution. Anthropogenic climate change can exacerbate these morphologic changes over time, as precipitation and fire regimes will experience changes in event recurrence. Landlab, an open-source, componentized model written in Python, can be used to explore landscape evolution across both short and long time scales. The observed hydrologic conditions from Spring Creek are modeled across a 1-meter digital elevation model (DEM) of the watershed to simulate observed flow and sediment transport events. Once the flow and sediment transport data are validated using Landlab components, the model results can be extrapolated to understand landscape response to precipitation and discharge events that occurred immediately post-fire that were not measured in the field. Particularly, sediment transport by overland flow through sheet-flow and rilling are more significant in burned than unburned landscapes and the Landlab post-fire model can be used to quantify these differences. These results shed light on which processes are drivers of post-fire erosional responses in central Colorado, and how those responses can drive morphological changes over much longer time scales, particularly when climate change is considered.