Repository logo

Improvements to the limb scattering stratospheric aerosol record



Journal Title

Journal ISSN

Volume Title






Degree Level



In the last decade stratospheric aerosols have gained considerable attention due to the influence of a series of moderate volcanic eruptions. The eruptions have been explosive enough to inject aerosols and precursors into the stratosphere and cause minor but important radiative and chemical effects, impacting projections and modelling of the global climate. Improved understanding of these effects requires accurate measurements of aerosol levels at spatial and temporal scales that resolve the rapidly changing conditions after events such as volcanic eruptions while also providing global information. This has been enabled by the advent of satellite profiling observations beginning in the 1980s that are able to produce global, vertically resolved measurements of stratospheric aerosols. These records have helped improve estimates of radiative forcing but remain uncertain in key aspects, including the magnitude of the biases between different measurement systems, errors in records due to retrieval assumptions, and aerosol levels in the upper troposphere and lower stratosphere. This work quantifies and addresses these limitations using three studies. First, biases are explored between the two longest satellite-based stratospheric aerosol records: SAGE II from 1984-2005 and OSIRIS from 2001-present. Biases are found to be relatively small, approximately 10\%, in the majority of the stratosphere, and a merged aerosol record spanning 35 years is produced by adjusting for these measurement biases. This work produced an aerosol climatology suitable for use in climate models, but did not determine the reasons for the biases. The second study compares two instruments and their retrievals, OSIRIS and SCIAMACHY, to investigate the major sources of error. It is found that errors in the a priori assumptions including particle size and the aerosol profile at high altitudes cause the majority of biases, while differences in the retrieval techniques and the radiative transfer models have mostly negligible impacts. The final study uses these results to develop a new multi-wavelength retrieval for OSIRIS measurements that aims to minimize the errors from a priori assumptions and improve retrieval sensitivity in the upper troposphere and lower stratosphere. This is used to produce the publicly available version 7 OSIRIS aerosol product, and is validated using comparisons with SAGE measurements as well as satellite lidar observations. Significant reductions in particle size biases are found with this new algorithm, and an updated cloud filter allows for retrievals at lower altitudes than previously possible.



Aerosol, Stratosphere, Remote Sensing



Doctor of Philosophy (Ph.D.)


Physics and Engineering Physics




Part Of