Topographic, climatic, and age controls on the reworking of volcanic debris avalanche deposits
Volcanic debris avalanches have deposited as much as 1000 km 3 of largely unconsolidated material on landscapes and remodeled existing drainage networks. The landscape disturbances created by these events pose severe, cascading downstream sedimentation hazards that can require long-term societal management, as demonstrated by decades of observations and ongoing interventions after the deposition of the 1980 debris avalanche of Mount St. Helens (United States). There, post-emplacement sediment yields caused by deposit erosion remain several times above estimated background yield and lakes impounded by the deposit still pose threats of downstream flooding. Despite the length and quality of measurements of the geomorphic evolution and consequent sediment release at Mount St. Helens, the long-term trajectory of drainage network evolution across, and the associated sediment release from, large volcanic debris avalanches remains uncertain. Observations and modeling at Mount St. Helens, however, indicate channel instability can persist many decades and may persist for centuries to millennia. We examined potential influences on the erosion and preservation of volcanic debris avalanche deposits (VDADs) by mapping valley networks developed on 89 VDADs selected from volcanic arcs across the world and spanning a variety of topographic settings and climate regimes. Using the best available topographic data (1 m lidar to 30 m radar-derived data depending on location) and aerial imagery, we estimated the areas of deposits that have been reworked relative to initial deposit footprints as a proxy for post-emplacement erosion. We found that a primary influence on reworking is the topographic confinement of the VDAD: confined, valley-filling deposits are systematically more reworked than unconfined deposits. There is no apparent relationship between deposit age and reworking for valley-filling deposits, indicating that drainage networks on deposits in confined topographic settings like at Mount St. Helens reform rapidly after emplacement. In contrast, our data indicate that the reworking of unconfined deposits has a monotonic positive relationship with age. This observation agrees with a conceptual model of channel formation at Mount Taranaki (New Zealand), which posits that an unconfined VDAD created a topographic high that initially (e.g., 2–8 ky for the Pungarehu formation at Taranaki) diverted erosion to the deposit margins. We found only a weak to moderate relationship between reworking and modern precipitation regimes, which may reflect differences between modern and paleo-precipitation conditions at many of our study sites. We also found no correlation between the size (surface area or volume) of deposits and the degree of reworking. The work presented here implies that downstream cascading sediment hazards from landscape-resetting processes like VDADs (such as thick, extensive pyroclastic flow deposits) depend on the relief and organization of the surrounding landscape.