Speaker
Description
Physics and chemistry fundamentally govern stratospheric aerosol processing and dispersion following an injection event. While a self-evident statement, the exact course of those processes which govern residence time and final yields of chemical products like sulfuric acid can have significant uncertainty in the time period shortly following these events. Forecast models have a degree of power validated by empirical observations including satellite observations. The accuracy and precision of those measurements and their underlying utility is a product of the spatiotemporal sampling frequency and the sampling technique used by each instrument. The more accurate techniques often trade sampling frequency for that increased accuracy or direct measurement of the relevant quantities as opposed to inference. Thomason et al. (2021) showed evidence for the predictability of stratospheric aerosol evolution given the initial injection conditions, and the Global Space-based Stratospheric Aerosol Climatology (GloSSAC) provides a model for stratospheric aerosol perturbations as a posterior analysis. Robust predictive tools that can fill gaps between instrument measurements or entire records can improve estimates for aerosol climatologies as well as forecasts for long-term aerosol burdens and fallout rates. This research will assess if there is promise in additional analysis tools such as machine learning techniques to improve existing understanding and modelling assimilations for the prediction of aerosol properties based on previous satellite observations.
| Topic | Aerosols and clouds |
|---|
