Volcanic explosions are the most critical replenishing mechanism of the stratospheric aerosol Junge layer. A fresh volcanic cloud comprises mostly sulfur-bearing gases, volcanic ash, and water vapor. It is commonly assumed that only sulfate aerosols remain in an aged volcanic cloud. Accurate simulation of the initial evolution of multicomponent fresh volcanic clouds is largely missing due to insufficient spatial resolution and a lack of relevant physics in global climate models. However, this initial stage is essential, as the vertical structure, composition, and altitude of a freshly developed volcanic cloud affect its long-term evolution. To fill this gap, we modified a regional WRF-Chem model to study the dispersion of a Pinatubo-size volcanic cloud in the equatorial belt with a 25 km grid spacing explicitly accounting for the SO2, ash, sulfate, water vapor, and hydrometeors radiative effects. The model best reproduces the observed evolution of the Pinatubo optical depth when eruptive products are injected above the cold tropical tropopause at 17 km. During the first week, the volcanic cloud in our simulations rises 1 km/day. Ash is primarily responsible for the heating and lofting of the volcanic products. Radiative heating of SO2 is weaker than that of ash and sulfate but is sufficient to position the core of the SO2 layer 1–2 km above the sulfate layer. Utilizing a more realistic description of the volcanic cloud's initial stage potentially improves overall volcanic cloud predictability. It might also be essential to designing geoengineering technologies based on the injection of aerosol precursors in the lower stratosphere.
|Original language||English (US)|
|Journal||Journal of Geophysical Research: Atmospheres|
|State||Published - May 14 2021|
Bibliographical noteKAUST Repository Item: Exported on 2021-06-10
Acknowledged KAUST grant number(s): BAS/1/1309-01-01.
Acknowledgements: The study reported in this publication was supported by funding from King Abdullah University of Science and Technology (KAUST) through Grant BAS/1/1309-01-01. The authors thank the KAUST Supercomputing Laboratory for providing computer resources.