Investigation of volcanic effects on the particle size of stratospheric aerosols - Thesis

[Geoengineering News]

https://epub.ub.uni-greifswald.de/frontdoor/index/index/docId/14588

Authors: Felix Wrana

24 April 2026

Abstract 

In this work, a new algorithm to retrieve the size distribution of stratospheric aerosol from the SAGE III instrument's satellite solar occultation measurements was introduced. Utilizing the instrument's wide spectral range, all three parameters of an assumed monomodal lognormal distribution could be derived, eliminating a large source of ambiguity compared to previous satellite retrievals. This results in one of the most reliable satellite-based stratospheric aerosol size data sets to date.

Utilizing the balance this data set strikes between the size resolution of the retrieved PSD and the spatial and temporal coverage, an unexpected signal was found in both the SAGE III/M3M (2002-2005) and the SAGE III/ISS (2017 - ongoing) data sets. Some volcanic eruptions were found to lead to a decrease of the average size of stratospheric aerosol rather than an increase, which is what would have been expected based on past studies, e.g. of the 1991 Pinatubo eruption. Furthermore, this size decrease was found to last up to 9 months, i.e. longer than expected. These were all eruptions of tropical volcanoes with relatively small amounts of SO2 (less than 1 Tg) injected into the stratosphere, namely Ruang and Reventador in 2002, Manam in 2005, Ambae in 2018, Ulawun in 2019 and La Soufrière in 2021. Model simulations with the MAECHAM5-HAM model could reproduce the qualitative signal found, but not its longevity. On the other hand, the 2019 Raikoke eruption (48°N, 1.37 Tg SO2) and the 2022 Hunga eruption (20.5°S, ~ 0.4-0.5 Tg SO2) were found to have led to an increase of the average aerosol size.

This raises the question of which factors are needed for a volcanic eruption to lead to a size decrease. The ambient temperature likely plays an important role (and therefore season of the eruption, volcano latitude and injection altitude) as low temperatures favor the nucleation of new particles over the condensation onto existing particles. However, the coagulation rate must also be an important factor, especially since the size reduction was found to last so long in some cases. Furthermore, the injected SO2 amount and the background PSD will also play a role. A lot of further research is needed in order to better understand the interaction of all these factors, both by modeling and observational experts.

The presented retrieval data sets were compared and validated against other observational aerosol size data sets, such as from balloon-borne in situ measurements and satellite limb scattering observations. While such comparisons are in detail very difficult to interpret due to differences in viewing geometry, observed air volumes, quality of collocations and differences in assumptions and retrieval algorithms, ultimately good agreements were found. This is encouraging especially in the case of the in situ measurements, that represent the most reliable aerosol size data available to date.

Finally, a general issue raised before by von Savigny and Hoffmann (2020), which is an important source of uncertainty, was applied to the presented method: The assumption of a simplified analytical function, in this case a monomodal lognormal distribution, to describe the stratospheric aerosol PSD. Since the larger particles of an aerosol population will individually scatter more strongly than the smaller particles, a divergence of the true PSD from the assumed PSD shape can introduce large errors into the aerosol size retrievals from any remote sensing instrument. In this work, this affects the number density the strongest.

The presented results are relevant for climate modeling and for the better understanding of the physical processes in our atmosphere in general. Furthermore, they add two major uncertainties to the predictability of possible geoengineering experiments in the form of solar radiation management through the injection of sulfur species into the stratosphere: Firstly, we currently do not sufficiently understand under which circumstances the aerosol size will increase or decrease, which will affect their lifetime and radiative effect. Secondly, all remote sensing observations that we have used in the past to understand the stratospheric aerosol PSD include major uncertainties due to the limited size resolution of the size retrievals and due to the smaller particles contributing very little to the measured signals.

Source: Universität Greifswald, Mathematisch-Naturwissenschaftliche Fakultät