High-resolution volcanic SO2 emissions in WACCM produce more realistic AODs

Volcanic climate impact is strongly correlated with how high its effluents reach in the atmosphere. Volcanic SO₂ injected into the stratosphere can have a residence time of several years, whereas injections in the troposphere only have a residence time of weeks.
We have performed simulations of the 2009 Sarychev eruption with the Community Earth System Model version 2 (CESM2), Whole Atmosphere Community Climate Model version 6 (WACCM6). We have compared the standard model dataset with two high-resolution datasets developed in our group to investigate the importance of utilizing highly vertically and horizontally resolved SO₂ datasets. The default dataset M16 [1] used in WACCM has a vertical resolution of 1 km and is released in one latitude longitude gridbox. Our two high-vertical resolution datasets, S21-3D, and S21-1D, were created from the dataset of Sandvik et al. [2]. These datasets have a vertical resolution of 200 m, spanning from 7.6 to 18.6 km. S21-3D is distributed over several latitudes and longitudes, whereas S21-1D releases all SO₂ in one latitude longitude gridbox to mimic the default dataset in WACCM.

The SO₂ from S21-3D, and S21-1D is injected at a higher altitude than the M16, leading to a longer residence time for both the SO₂ and the formed stratospheric aerosol. The S21-3D dataset has the highest stratospheric SO₂ and SO₄ concentrations of the three simulations and the concentrations peak later than the other two simulations. The results from the S21-1D dataset are similar to those from the S21-3D simulation but with slightly lower concentrations.

The simulated S21-3D AOD agrees with AOD from the space-borne lidar instrument CALIOP. In the simulation with the M16 dataset, the AOD is underestimated by >50%. The volcanic radiative forcing from the Sarychev eruption was 31% lower at the end of 2009 in the simulation with M16 compared to the simulation with S21-3D.

Our results show that the vertical resolution of SO₂ injections substantially impacts the model’s ability to correctly simulate the climate effects of volcanic eruptions, especially if the SO₂ is injected in the vicinity of the tropopause.

Future work will involve simulating a high vertically resolved dataset of SO₂ of other volcanic eruptions since 2006.

References

[1] Mills, M. J., A. Schmidt, R. Easter, S. Solomon, D. E. Kinnison, S. J. Ghan, R. R. III Neely, D. R. Marsh, A. Conley, C. G. Bardeen, et al. (2016), Global volcanic aerosol properties derived from emissions, 1990–2014, using CESM1(WACCM), J. Geophys. Res. Atmos., 121, 2332–2348, doi:10.1002/2015JD024290.

[2] Sandvik, O. S., Friberg, J., Sporre, M. K., and Martinsson, B. G.: Methodology to obtain highly resolved SO₂ vertical profiles for representation of volcanic emissions in climate models, Atmos. Meas. Tech., 14, 7153–7165, https://doi.org/10.5194/amt-14-7153-2021, 2021.