Simulation of Rock Avalanches and Pyroclastic Flows

Sheridan EAR-0087665 The risk of potential rock avalanches and pyroclastic flows is a problem that public safety authorities around the world face several times a year (Tilling, 1989a). We propose to develop a 3-D computer code to simulate the mass and momentum flow from dangerous geologic avalanches, based on estimates of original volume of the starting materials and their properties and an accurate digital terrain model of the surface. The model will forecast lateral and longitudinal thickness variations in both the moving flow and the resulting deposits. This effort will greatly extend the forecasting products now provided by the successful Flow3D code of Sheridan (Kover, 1994; Sheridan and Kover, 1997). This project will contain several elements: 1. a mathematical model for dense grain flows with an air (avalanche) or hot gas (pyroclastic flow) matrix; 2. data constraint based on field studies of several recent avalanches in the Cascade Range and pyroclastic flows at Mount St. Helens and Colima, Mexico; 3. flume and inclined slope laboratory experiments to test the model at a small scales; and 4. simulation of scenarios of potential future events at the study sites. The simulation code will be developed in close concert with laboratory and field studies, providing fidelity of the computer modeling with experimental results. The basis of our simulations will be estimation of source volumes and runout data (deposit profile, thickness, etc.) of actual avalanches and flows and accurate DEMs of current topography. The models will provide velocity histories, flow paths, flow thickness, and planimetric areas of the simulations, all in three dimensions. The output display format will be suitable for interpretation by public safety officials as well as scientists. The results will be back-fitted to pyroclastic flow data from Unzen, Soufriere Hills, Mt. St. Helens, and Volcano Colima and avalanche data from Cascade volcanoes. Based on our benchmarking, the code will allow us to forecast future local hazards at these and other sites including Pico de Orizaba, Mount Rainier, and Mount Adams. The resulting model should help authorities design mitigation plans for specific dangerous areas.