Impact of a super-volcanic eruption on general circulation and chemistry in the middle atmosphere

The Toba super-volcanic eruption, which was occurred about 74,000 years ago, was the largest explosive eruption at least within the last 100,000 years. The estimated injection of sulfur dioxide (SO2) into the atmosphere by the event was about 300 times larger than that of the 1991 Mount Pinatubo eruption which was the largest ones in the 20th century. Robock et al. (2009) investigate climate responses to such a super-volcanic eruption by analyzing a set of climate model simulations. They suggest that the super-volcanic eruption could certainly produce a volcanic winter for about a decade, with serious effects on plant and animal life. In this study, we also estimated quantitatively impacts of a Toba-like super-volcanic eruption on the Earth's climate and environment by using an earth system model. We investigated particularly impacts of the Toba-like eruption on the Brewer-Dobson circulation and chemistry process in the middle atmosphere, both of which affect stratospheric residence time of volcanic sulfate aerosols. In addition, we also examined radiative effects of large increase in stratospheric water vapor induced by the Toba-like perturbation, causing larger cooling in the stratosphere and significant changes in the Brewer-Dobson circulation.

The earth system model used in this study has been developed at Meteorological Research Institute (MRI) of Japan Meteorological Agency (JMA), called MRI-ESM1 (Adachi et al., 2013). MRI-ESM1 is an extended version of MRI-CGCM3 (Yukimoto et al., 2012). MRI-CGCM3 consists of three component models: the atmospheric GCM (AGCM), called MRI-AGCM3, which includes a land module; the aerosol model, MASINGAR mk-2; and the ocean GCM (OGCM), MRI.COM3. Along with the three component models of MRI-CGCM3, MRI-ESM1 includes an ozone chemistry model, MRI-CCM2 (Deushi and Shibata, 2011), and carbon cycle modules for land and ocean. Two long-term runs were conducted using MRI-ESM1 for extracting climate responses to the Toba-like eruption: the preindustrial (hereafter PI) control run under the condition of 1850 AD with no large volcanic eruption, and the 300 times Pinatubo run with the stating date of August 1, in which 6 Gt of SO2 (300 times larger than the Pinatubo amount of 20 Mt) was injected in the stratosphere on August 26 in the first year of the simulation. The boundary conditions and initial values in the 300 times Pinatubo run were the same as those of the PI control run except for the super-volcanic SO2 injection. The climate responses caused by the super-volcanic eruption were calculated by subtracting the PI control run from the 300 times Pinatubo run. We also performed an additional sensitivity run which is basically the same as the 300 times Pinatubo run except that the same stratospheric water vapor distribution as that of the PI control run was forced when calculating the atmospheric radiation process. Comparing the climate responses of the 300 times Pinatubo run with those of the sensitivity run, we can estimate impacts of changes in stratospheric water vapor distribution due to the super-volcanic eruption on temperature in the middle atmosphere and the Brewer-Dobson circulation.

The simulation results of the 300 times Pinatubo run show that for the first 5 years the global-mean temperature anomaly at the height of 30 hPa is positive with its maximum of about 30K due to the large volcanic aerosol loading in the stratosphere. After that, the global temperature rapidly decreases and its anomaly turns to be negative for about 10 years. On the other hand, this negative anomaly is not seen in the sensitivity run. Therefore, this negative anomaly seen in the 300 times Pinatubo run is possibly caused by the larger radiative cooling due to rapid increase in the stratospheric water vapor. This stratospheric cooling also induces the stronger Brewer-Dobson circulation, leading to the shorter stratospheric residence time of sulfate volcanic aerosols by several years, compared to the sensitivity run. Although both the 300 times Pinatubo run and the sensitivity run show that the surface global-mean temperature is cooled by about 12K, the cooling period in the 300 times Pinatubo run is shorter by several years compared to the sensitivity run because of the shorter residence time of stratospheric volcanic aerosols.