Using high-resolution DEMs for debris flow detection based on topographic signatures: A case study in the Quebrada del Toro, NW Argentina

Mountainous high-relief terrains in climatically sensitive regions are often subjected to natural extreme events such as debris flows and landsliding. With people and infrastructure at risk, it is important to identify, measure, and comprehend the driving forces and mechanisms of slope movements in these environments at regional scale. Geomorphologic analyses and hazard assessments in these regions are, however, often limited by the availability of good-quality high-resolution digital elevation models (DEMs). Publically available data often have lower spatial resolution and are distorted in high-relief areas. In contrast, airplane-based lidar (light detection and ranging) data provide highly accurate information on 3D structure, yet, acquisition is costly and limits the size of the respective study area. Finding adequate, economical alternatives for creating high-resolution DEMs is therefore essential to study Earth-surface processes at regional scale, which may enable the detection of spatial variations, clusters and trends.

In areas with sparse vegetation, stereogrammetry has proven to be a viable tool for creating high-resolution DEMs. Here, we use SPOT-7 tri-stereo satellite imagery to create DEMs at 3 m spatial resolution for the Quebrada del Toro (QdT) in the Eastern Cordillera of NW Argentine Andes, an area with extreme gradients in topography, rainfall and erosion. Over 5000 GPS points collected during fieldwork ensure the spatial coherence of our DEMs.

Field observations in this high-elevation area show that the hillslopes of the deeply incised QdT gorge are characterized by debris flow deposits of various extent. Debris flows have a specific slope-drainage area relationship that curves in log-log space. Using high-resolution topographic data, we are able to provide further evidence for this phenomenon and characterize the distinct topographic signature of debris flows. We specifically focus on the transition zone between debris-flow and fluvial processes, which is variable in the different catchments. The transition is characterized by a pronounced kink revealed in slope-drainage plots, as well as an increase of slope scatter in the drainage area logbins. We propose that the presence and location of this kink reflects the nature of the dominating transport processes in the corresponding catchments. In light of these observations we discriminate between debris-flow and fluvially dominated catchments in the QdT and identify regions that primarily exhibit slope movement. Our new results reveal a cluster of fluvial catchments to the SE of our study area – an area that receives significantly more moisture than upstream regions. In contrast, debris flows are prominent in areas of sparse vegetation, where occasional extreme rainfall events are efficient in transporting large amounts of talus downhill. These observations are key to a better understanding of the relationships between the impact of extreme rainfalls at high elevation and the formation of large volumes of sediment in the arid highlands of the Andes.