Cloud-Sulfate Correlations as a Constraint on Heterogeneous Sulfate Production

Statistical analysis of observed daily cloud cover (from ISCCP) and sulfate surface concentrations in Europe and North America indicates a significant negative correlation between clouds and sulfate. These anti-correlations are at long time scales (typically 8-64 days) and occur for many months of the 3 years of data. A corresponding analysis of modeled sulfate and cloud cover (from the Goddard Institute for Space Studies GCM) fails to produce significant correlation. However, if we separate heterogeneous (cloud-produced) and gas phase sulfate, there is some significant negative correlation between clouds and gas phase sulfate. If we turn off gas phase sulfate production beneath clouds (as should happen since the oxidant OH is photochemically generated), this negative correlation becomes stronger; however the total sulfate-cloud correlation remains insignificant. This version (`2-prime', 9-layer) of the GISS model, like most global models, does not have a separate budget for dissolved species. Instead, after each cloud timestep, the sulfate generated in-cloud is returned to the gridbox rather than retained in the cloud droplets. We perform another simulation which has a dissolved tracer budget for sulfate, so that sulfate generated in a cloud is not released to the gridbox unless the cloud evaporates. The result is that more of this sulfate is rained out and therefore sulfate production decreases (relative to the standard simulation): the in-cloud sulfate production drops from 32 to 10 Tg S/yr and total sulfate burden (including gas and heterogeneous phases) drops from 0.71 to 0.54 Tg S. In this simulation, we do achieve significant negative correlation between clouds and sulfate (although still not as much as observed). This study implies that most sulfate is generated in the gas phase, and that global models have excessive heterogeneous sulfate production. We will also present results from the new (`modelE', 16 layer) version of the GISS model that has dissolved species budgets for for precursers H₂O₂ and SO₂ as well as sulfate. Finally we will show how the reduction in the heterogeneous sulfate production affects the indirect radiative forcing.