Assessment of Stratospheric Aerosol Properties (ASAP)
Report on the Assessment Kick-Off Workshop, Paris, France, 4-6 November 2001

Larry W. Thomason, NASA Langley Research Center, USA (l.w.thomason@larc.nasa.gov)
Thomas Peter, Swiss Federal Institute of Technology, Switzerland (Thomas.Peter@ethz.ch)
Co-Chairs of ASAP

ASAP Structure (please contact the SPARC Office, for a better resolution of the image)

Aerosol is one of the most variable components of stratospheric air. As the result of episodic volcanic eruptions, the mass of aerosol in the stratosphere has varied by about 2 orders of magnitude over the last 25 years. Aerosol in the upper troposphere/lower stratosphere (UT/LS) can have a significant impact on climate through radiative effects and on stratospheric chemistry, particularly impacting ozone. The magnitude of these effects is also highly variable and, as a result, can complicate isolating human-derived changes in the stratosphere. Accounting for aerosol effects properly is a key component of modelling climate/chemistry effects properly.

By mass stratospheric aerosol consists of almost exclusively aqueous sulphuric acid droplets. However, its precise characterisation is nonetheless complex. The details of composition and size distribution strongly influence important aerosol properties like surface area density and the spectral dependence of aerosol scattering and absorption. While most gas species measurements, such as those of ozone, are essentially measurements of the species molecular density and hence by itself useful, aerosol instruments employ a variety of measurement strategies, each determining only some aspects of the underlying aerosol distribution. As a result, the conversion of measurements to practical parameters generally requires model-based transformations that may themselves be subject to substantial uncertainty.

In recognition of the importance of stratospheric aerosol, SPARC has initiated a project to evaluate the scientific understanding of UT/LS aerosol and aerosol measurements. The assessment was initiated with a workshop in Paris, France on 4-6 November 2001, supported by the SPARC Office. Thirty aerosol scientists attended the workshop and identified key scientific issues that should be addressed in the assessment. These include:

• How representative are satellite-based aerosol records and derived climatologies?
• How have key aerosol properties (e.g., surface area density) and precursor gas loadings varied with time?
• What is the non-volcanic background for stratospheric aerosol? Can a trend in the background level be inferred?
• How well does the aerosol record/climatology reflect our understanding of aerosol processes, in particular in regions of or at times of enhanced aerosol nucleation, coagulation, growth and evaporation?
• How well can models using gaseous precursors to produce observed aerosol properties under non-volcanic and volcanically perturbed conditions?

The assessment will attempt to address these questions in five working groups: processes, aerosol precursors, aerosol record and climatology, trends, and modelling. In parallel with the construction of the aerosol record, that we intend to become the most comprehensive one compiled so far, the assessment of precursor gases, aerosol trend detection, and aerosol modelling will proceed with regular feedback between groups to ensure consistency. These activities are also hoped to merge into an aerosol climatology reflecting stratospheric aerosol properties during volcanically quiescent periods.

The aerosol processes group will review our knowledge of the aerosol lifecycle including their formation in and removal from the stratosphere, thereby providing context to all sections of the assessment. Specific issues to be addressed include the nucleation of aerosols at the tropical tropopause, and in the free stratosphere following volcanic events. Processes that control clouds in the vicinity of the tropopause and polar stratospheric clouds will also be reviewed. The aerosol precursors group will quantify what is known about the non-volcanic gaseous precursors to stratospheric aerosol formation and produce a database for further experimentation. This includes natural and human-derived source species such as OCS, tropospheric SO2, and DMS. This group will also address whether there are discernible trends in the gas concentrations, possibly of anthropogenic origin.

The primary products of the aerosol record will be satellite-based and as such, the primary parameters will be aerosol extinction at a variety of wavelengths. Key secondary products (e.g., aerosol surface area density and effective radius) will be derived and supplemented with additional parameters (e.g., conversion factors between backscatter and extinction at a few wavelengths). We expect that the long records from SAGE II, HALOE, and other space-based instruments will play a prominent role in building a comprehensive data record, while balloon-borne and lidar records will further serve to validate the remotely estimated values. The construction of the climatology for volcanically quiescent periods will depend crucially on in situ observations like the University of Wyoming optical particle counter (OPC) and on long-term lidar measurements like the Garmisch record. Since there is evidence for significant systematic differences between some satellite and in situ measurements in derived quantities, like surface area density, we will address the potential for bias and limitations in general.

The record will be compiled as a function of latitude, altitude and time, and possibly also as a function of equivalent latitude, potential temperature and time. The altitude range of interest extends from 2-3 km below the tropopause to the stratopause. Hence, it will include the scientifically interesting and for satellite instruments easily accessible region of the uppermost troposphere. This spans a region in which ice clouds frequently occur. We intend to quantify the frequency of occurrence of cirrus events as a function of time and location but otherwise exclude them from aerosol data products. The final products will be electronically available through the SPARC Data Center.

The evaluation of trends will concentrate on the volcanically quiescent periods, though much can be learned about natural aerosol variability in volcanically perturbed periods as well. Underlying data include aerosol mixing ratios from the OPC, backscatter ratios from the lidars, and aerosol extinction for HALOE and SAGE II. The trends group will determine what periods can be identified as being at non-volcanic levels, if any. The cleanest periods, for which substantial data are available, are the late 1970’s, the late 1980’s and the late 1990’s through the current time. These periods will be compared using a rigorous statistical approach, and the significance of any observed differences will be evaluated.

The modelling group will bring together the measured aerosol source gas budgets, the observed volcanic and background aerosol levels, the derived trends in the latter periods, and the modelling of transport, oxidation, nucleation and condensation processes. While this group is not intended to perform a comprehensive comparison between global aerosol models, the agreement between the observed aerosol and that produced by various models is expected to give significant insight in the how close we are to understanding the processes that control aerosol in the stratosphere.

Acknowledgements:

We wish to thank the many colleagues who have accepted to serve as authors in this assessment and whose enthusiasm for the topic was evident throughout the Paris workshop. We would further like to thank CNES for hosting the meeting, and most prominently the SPARC Office, whose nonpareil support in organising this workshop was a key element in its success.