Aerosol Workshop — June 2-3, 1997
Panel B: Indirect Aerosol Forcing
P. Hobbs (Facilitator), S. Ghan (Recorder), T. Novakov, S. Schwartz
Panel B Summary by
Uncertainty in estimates of indirect radiative forcing is far greater than uncertainty in any other anthropogenic forcing. This great uncertainty is due to the complexity of the mechanisms involved. Given the potential for a large forcing, we should not shirk the issue, but rather work hard to reduce the uncertainty.
To reduce the uncertainty, an experiment should be designed involving a signal that is characteristic of anthropogenic aerosol (e.g. a small town). A well-designed experiment of this type, involving satellite as well as in situ measurements, and closely linked to improving and evaluating both process models and GCMs, offers the potential to obtain for the first time a quantitative estimate of indirect forcing by anthropogenic aerosols. The experiment should isolate the forcing into at least three factors: (a) the sensitivity of aerosol concentration, composition, and size distribution to emissions, (b) the sensitivity of droplet number to aerosol concentration, composition, and size distribution, and (c) the sensitivity of cloud radiative forcing to cloud microphysical parameters.
We envisage a highly focused experiment, similar in scope to the highly successful (and relatively cheap) TARFOX. Satellite data would be used to identify a geographical location where anthropogenic emissions commonly give rise to the situation depicted schematically in the box on the following page. That is, emissions from a town or city cause stratiform clouds downwind to have a greater albedo than the surrounding clouds, due to the effects of the emissions on cloud microstructure (similar to the "ship track" effect, but on a larger scale). Airborne studies would then be carried out at the site to quantify the nature of the anthropogenic emissions and their effects on cloud structure. Based on these observations, parameterizations of the effects of aerosols on cloud albedo would be developed for inclusion in GCMs.