Stratospheric Aerosols: New Tricks for Old Dogs

The background stratospheric aerosol layer was first described by Christian Junge and others in 1961. Volcanic stratospheric aerosols have been remarked upon for centuries, though it was not until the eruption of Mt. St. Helens in 1980 that it was widely recognized that the volcanic particles were largely sulfuric acid rather than pulverized volcanic rock. Likewise, polar stratospheric clouds have been observed for more than a century, but it was only in the 1980s that it was realized that some of them are made of nitric acid solutions rather than purely water ice. Given this long history it might be thought that there was little left to discover, but the past few years have turned up a host of new discoveries and challenges which I will overview in this talk. The background stratospheric aerosol layer has long been thought to be dominated by sulfuric acid, but recent studies show that about 50% of the extinction below 20 km is instead due to organic aerosols. In addition, in parts of the world and during certain times, nitrates are abundant and black carbon from fires make important contributions. Above 30 km micrometeorites provide significant extinction up to the mesosphere. They compose about 10% of the stratospheric aerosol by mass, but their impacts have been little studied. Pyrocumulus clearly are responsible for injecting black carbon into the stratosphere, but the magnitude of the injection if not well quantified. One injection in 2017 was so large that the smoke layer was observed by satellites for more than 8 months, and the smoke rose to above 20 km due to radiative heating. Smaller volcanic eruptions have been found to be more common than recognized in the past, and to be regularly producing small perturbations to the climate. Some models have not been able to reproduce the conversion times between volcanic SO₂and sulfate suggesting a problem, perhaps in photochemical schemes which do not treat aerosols. More broadly, a comparison of volcanic cloud simulations shows large differences between models. These differences may primarily relate to models having incomplete physics, particularly related to not computing photochemistry of OH interactively with the volcanic aerosol. However, other problems may be related to how size distributions are resolved, treatment or lack thereof for volcanic ash and water injections, and treatment of the initial few months of particle evolution when particle sizes are determined by growth, and coagulation.