5:15 PM - 6:45 PM
[MIS15-P01] Application of a 2-D two-layer depth-averaged model for large-scale pyroclastic density currents to the Pinatubo 1991 eruption: Effects of uncertainty in source location
Keywords:Pyroclastic Density Current, Ignimbrite, Progressive Aggradation, 2-D Two-Layer Depth-Averaged Model, Mt. Pinatubo, Gravity Current
During explosive volcanic eruptions, a mixture of volcanic particles and gas ejects from the vent as an eruption column, and it can collapse and flow along the ground as a pyroclastic density current (PDC). PDCs are hazardous phenomena due to their rapid propagation along the ground, large dynamic pressures, and high temperatures. The processes of flow and deposition of PDCs are complex, which makes it difficult to relate the run-out area of PDCs with the conditions of eruption column (source) and topography. We aim to understand the relationship and develop a numerical model to predict the run-out area.
Shimizu and Koyaguchi (2022, JpGU meeting, SVC32-11) developed a numerical large-scale PDC model and applied it to the Pinatubo 1991 eruption. A PDC generally consists of an upper dilute turbulent suspension current (particle concentration <~1 vol.%) and a lower dense fluidized granular current (particle concentration ~50 vol.%). For large-scale PDCs, dilute currents with thickness of 102-103 m can flow without much influence from topography, whereas dense currents with thickness of 10-1-100 m flow in a complex manner along valley. To explain this feature, we formulated a two-layer PDC model consisting of a one-dimensional axisymmetric depth-averaged dilute current expanding radially from the eruption column and a two-dimensional depth-averaged dense current depending on topographic height. The numerical simulation for the Pinatubo 1991 PDC used a digital elevation model based on the measurement in 2007 and set the center of the eruption column to those of Mt. Pinatubo. The simulation result explains the global features of the PDC deposits, i.e., the axisymmetric distribution near the source (within ~5 km from Mt. Pinatubo) and the valley-fill distribution farther from the source (from ~5 km to the maximum run-out distance (~12 km)). On the other hand, it failed to reproduce the following two field observations: the wide distribution of the valley-filling deposits to the northwest of Mt. Pinatubo and the presence of the deposits along the Sacobia River east-northeast of Mt. Pinatubo.
To consider the cause of the above two discrepancies, we performed a parametric study for the effect of uncertainty in the location of the eruption column (source). Generally, the central coordinates of the source of PDCs can vary due to the directionality of eruption columns. To consider this effect, we shifted the central coordinates of the source by 2 km (~ the caldera diameter) in each of eight directions from Mt. Pinatubo. The inability to reproduce the wide valley-fill deposits to northwest of Mt. Pinatubo is improved by shifting the source location 2 km northwest of Mt. Pinatubo. However, the simulation under this condition results in underestimating the run-out area in the opposite (southeast) direction. The inability to reproduce the observed distribution of deposits along the Sacobia River on the east-northeast side is not improved even if the source location is shifted 2 km each in the northeast or east direction.
The above results suggest that, in addition to the directionality of eruption columns, the temporal changes in eruption conditions and topography play an important role in the deposit distribution. For example, to form the deposits along the Sacobia River, the dense current must overcome the ridge located 7-8 km east of Mt. Pinatubo. According to field observations just after the eruption, the valleys near the source were buried by PDC deposits during the eruption, and the dense current might flow over the ridge. The current numerical model considers the mass loss of the dense current and the mass gain of the deposits due to progressive aggradation, but does not consider the temporal change in the topographic height due to the deposition. In the future, it may be possible to explain more detailed features of the deposit distribution by considering the temporal changes in eruption conditions and topography in the model.
Shimizu and Koyaguchi (2022, JpGU meeting, SVC32-11) developed a numerical large-scale PDC model and applied it to the Pinatubo 1991 eruption. A PDC generally consists of an upper dilute turbulent suspension current (particle concentration <~1 vol.%) and a lower dense fluidized granular current (particle concentration ~50 vol.%). For large-scale PDCs, dilute currents with thickness of 102-103 m can flow without much influence from topography, whereas dense currents with thickness of 10-1-100 m flow in a complex manner along valley. To explain this feature, we formulated a two-layer PDC model consisting of a one-dimensional axisymmetric depth-averaged dilute current expanding radially from the eruption column and a two-dimensional depth-averaged dense current depending on topographic height. The numerical simulation for the Pinatubo 1991 PDC used a digital elevation model based on the measurement in 2007 and set the center of the eruption column to those of Mt. Pinatubo. The simulation result explains the global features of the PDC deposits, i.e., the axisymmetric distribution near the source (within ~5 km from Mt. Pinatubo) and the valley-fill distribution farther from the source (from ~5 km to the maximum run-out distance (~12 km)). On the other hand, it failed to reproduce the following two field observations: the wide distribution of the valley-filling deposits to the northwest of Mt. Pinatubo and the presence of the deposits along the Sacobia River east-northeast of Mt. Pinatubo.
To consider the cause of the above two discrepancies, we performed a parametric study for the effect of uncertainty in the location of the eruption column (source). Generally, the central coordinates of the source of PDCs can vary due to the directionality of eruption columns. To consider this effect, we shifted the central coordinates of the source by 2 km (~ the caldera diameter) in each of eight directions from Mt. Pinatubo. The inability to reproduce the wide valley-fill deposits to northwest of Mt. Pinatubo is improved by shifting the source location 2 km northwest of Mt. Pinatubo. However, the simulation under this condition results in underestimating the run-out area in the opposite (southeast) direction. The inability to reproduce the observed distribution of deposits along the Sacobia River on the east-northeast side is not improved even if the source location is shifted 2 km each in the northeast or east direction.
The above results suggest that, in addition to the directionality of eruption columns, the temporal changes in eruption conditions and topography play an important role in the deposit distribution. For example, to form the deposits along the Sacobia River, the dense current must overcome the ridge located 7-8 km east of Mt. Pinatubo. According to field observations just after the eruption, the valleys near the source were buried by PDC deposits during the eruption, and the dense current might flow over the ridge. The current numerical model considers the mass loss of the dense current and the mass gain of the deposits due to progressive aggradation, but does not consider the temporal change in the topographic height due to the deposition. In the future, it may be possible to explain more detailed features of the deposit distribution by considering the temporal changes in eruption conditions and topography in the model.