1:45 PM - 3:15 PM
[SCG57-P03] Numerical simulation of sulfate formation through the reaction between volcanic ash and gas: Evaluation of sulfur losses from eruption clouds
Keywords:gas-ash reactions, diffusion, pyroclastic flow, SO2 scavenging, sulfate, Pinatubo 1991 eruption
Explosive volcanic eruptions inject sulfur dioxide (SO2) gases into the stratosphere. The SO2 gas can be converted into H2SO4 aerosols through chemical reactions in the stratosphere, which scatter solar radiations and decrease the temperature of the Earth’s surface. Therefore, estimating amounts of SO2 injected into the stratosphere is important to evaluate the impacts of volcanic eruptions on the Earth’s climate. Fresh volcanic ashes adhere SO2 gases onto their surfaces in eruption plume and in the conduit after magma fragmentation, resulting in the formation of CaSO4 through the reaction between Ca in the ash and SO2 gas molecules at high temperatures. When the volcanic ashes settle, the sulfate salts are removed from the eruption plume, that is, the volcanic ashes scavenge SO2 gases from the eruption plume through the formation of CaSO4. The growth of CaSO4, i.e., the process of SO2 scavenging depends on Ca diffusion at high temperatures. Here, we estimated the amount and efficiency of SO2 scavenged by volcanic ash in the eruption plume and in the conduit, combining the 3D numerical simulation of the volcanic cloud and the diffusion model.
The 3D simulation combines a pseudo-model for fluid motion and Lagrangian model for particle motion. The diffusion model assumes that the rate-limited process of the CaSO4 formation is Ca diffusion in a silicate melt and its diffusivity depends on only temperature. We obtained the temperature history of particles from the 3D simulation of the cloud and the size distribution of particles in the eruption plume was estimated based on pyroclast deposits. This model is applied to the pyroclastic flow and eruption column observed during the Pinatubo 1991 eruption. We assume the initial concentration of Ci = 4.5 wt%, and the boundary condition of C = 0 for time t > 0. Magma discharge rate was set to be 1.0×109 kg s-1 and the magma temperature was set to be 780°C.
Our calculation results showed that the amount of SO2 scavenging in the eruption plume is comparable to or slightly larger than that in the conduit. This result is quite different from the previous prediction that the amount of SO2 scavenging in the eruption plume must be low. The amount of SO2 scavenging in pyroclastic flow is larger than that in the eruption column in the case of Pinatubo. This is because pyroclastic flow forms a larger hot region just above the vent and thermal energy cannot be consumed easily by mixing volcanic gas and air compared to the eruption column. Our results also indicate that fine ash particles remove efficiently SO2 because of their high ratio of surface area to mass. An individual particle with a large size can uptake large amounts of SO2, but the contribution to the total amount of SO2scavenging is small compared to fine particles. These results are explained by considering that the large hot region formed in the eruption plume (especially in the pyroclastic flow) can trap fine particles for a long time while large particles are quickly removed from the eruption plume. Finally, we estimated the actual efficiency of SO2 scavenging during the Pinatubo 1991 eruption (9 h), using our calculation results and the satellite data of SO2 injected into the stratosphere (20 Mt), resulting in 62–81%. This estimate suggests that the large amounts of SO2 might have been scavenged in the case of the Pinatubo. However, it should be noted that our estimate includes large uncertainties.
We also found that the efficiency of the SO2 scavenging is low and high in the case of small-scale and large-scale eruptions, respectively, because the hot region in the eruption plume is more likely to be formed by large-scale eruptions. In the case of small-scale eruptions, scavenging at high temperatures can be dominant in the post-fragmentation conduit rather than in the eruption plume.
The 3D simulation combines a pseudo-model for fluid motion and Lagrangian model for particle motion. The diffusion model assumes that the rate-limited process of the CaSO4 formation is Ca diffusion in a silicate melt and its diffusivity depends on only temperature. We obtained the temperature history of particles from the 3D simulation of the cloud and the size distribution of particles in the eruption plume was estimated based on pyroclast deposits. This model is applied to the pyroclastic flow and eruption column observed during the Pinatubo 1991 eruption. We assume the initial concentration of Ci = 4.5 wt%, and the boundary condition of C = 0 for time t > 0. Magma discharge rate was set to be 1.0×109 kg s-1 and the magma temperature was set to be 780°C.
Our calculation results showed that the amount of SO2 scavenging in the eruption plume is comparable to or slightly larger than that in the conduit. This result is quite different from the previous prediction that the amount of SO2 scavenging in the eruption plume must be low. The amount of SO2 scavenging in pyroclastic flow is larger than that in the eruption column in the case of Pinatubo. This is because pyroclastic flow forms a larger hot region just above the vent and thermal energy cannot be consumed easily by mixing volcanic gas and air compared to the eruption column. Our results also indicate that fine ash particles remove efficiently SO2 because of their high ratio of surface area to mass. An individual particle with a large size can uptake large amounts of SO2, but the contribution to the total amount of SO2scavenging is small compared to fine particles. These results are explained by considering that the large hot region formed in the eruption plume (especially in the pyroclastic flow) can trap fine particles for a long time while large particles are quickly removed from the eruption plume. Finally, we estimated the actual efficiency of SO2 scavenging during the Pinatubo 1991 eruption (9 h), using our calculation results and the satellite data of SO2 injected into the stratosphere (20 Mt), resulting in 62–81%. This estimate suggests that the large amounts of SO2 might have been scavenged in the case of the Pinatubo. However, it should be noted that our estimate includes large uncertainties.
We also found that the efficiency of the SO2 scavenging is low and high in the case of small-scale and large-scale eruptions, respectively, because the hot region in the eruption plume is more likely to be formed by large-scale eruptions. In the case of small-scale eruptions, scavenging at high temperatures can be dominant in the post-fragmentation conduit rather than in the eruption plume.