A GEOMORPHIC PROCESS-BASED APPROACH TO DELINEATE DEBRIS FLOW ACTIVITY

Debris flows are the most prevalent and destructive mass wasting type in the southern Appalachians yet the recent identification of a debris flow topographic signature has not been incorporated into landslide inventories. A scaling transition in log-log plots of drainage area versus local channel slope that can be created from a digital elevation model provides the downstream limit of debris flow activity (A_df). Using a lidar-derived digital elevation model, we have determined an A_df range of 6 to 35 km² in the Oconaluftee River basin in the Great Smoky Mountains National Park (GSMNP) region of the southern Appalachians. Lithologic erodibility contrasts, between Ocoee Supergroup metasedimentary rocks and Grenville gneiss, induce site specific spatial variability to A_df. This diversity yields a study area that is well suited to test an application of this parameter.

We present an inventory technique that limits a land classification algorithm to areas that are upstream from the topographic signature of debris flows in five streams in GSMNP. Two inventories were created to determine if this technique improves inventory efficacy: (1) the entire study area, and (2) areas that drain to debris flow-dominated channels. We use topographic derivatives and a vegetation index map created from an aerial photograph to classify debris flow and non-debris flow areas. Classification efficacy was improved by 12% when the inventory was limited to areas upstream of the transition to debris flow-dominated channels. This method allows a more focused approach to statistically characterize the land surface, which resulted in increased inventory accuracy across a landscape with an extensive and relatively well-documented debris flow history.