Multiple-pulsed debris avalanche emplacement at Mount St. Helens in 1980: Evidence from numerical continuum flow simulations

The complex 1980 Mount St. Helens debris avalanche is modeled numerically as a transient biviscous fluid flow. Several approaches are considered, including two-step rheologically-distinct models for avalanche I and combined avalanches II/III, and a composite flow model consisting of retrogressive slides of identical rheology successively accreted to the main avalanche flow. For the two-step situation, flow rheologies are evaluated separately for the initial avalanche, comprising the debris avalanche block facies, and an ensuing explosive-influenced flow. Strengths (normalized by density) as high as 250 m2/s2 and apparent Newtonian viscosities as much as 275 m2/s were established for the block facies. These parameters for the explosively-influenced flow are an order of magnitude lower. The distribution of stratigraphic units within flowing model debris, compared with field distributions, suggests that the higher-strength emplacement models are appropriate for debris deposited on Johnston Ridge and in the upper parts and flanks of the North Fork Toutle River valley. In general, models for which constant rheology is assumed throughout the flow process provide lower-bound emplacement times, and excessive early velocities, as compared to the prototype event. Because model calibration is based on matching runout by trial and error, it is therefore biased toward the rheologic parameters essential to achieving that runout. These values characterize the flow in its latter stages, whereas the actual strength and viscosity may have substantially decreased as a function of displacement. Two-dimensional models predict debris accumulation about twice as thick as that observed at the foot of Mount St. Helens, where flow divergence was significant. This discrepancy is lessened with a quasi-three-dimensional modification of the flow model. Accretionary composite flow models with homogeneous rheology simulate the overriding of early avalanche debris by later debris pulses. The accretionary composite model also predicts a thick flow snout down-valley. Since this feature is not indicated in the debris deposit, the model lends support to the concept of flow separation by rheology. The two-step and accretionary composite flow models complement one another and provide model-based support for the Harry Glicken hypothesis for complex emplacement at Mount St. Helens.