European Solar “Dimming” and “Brightening” as a Diagnostic for Aerosol Simulation in GCMs

1. European Solar “Dimming” and “Brightening” as a Diagnostic for Aerosol Simulation in GCMs Joel Norris Scripps Institution of Oceanography 20 March 2006 Special thanks to Martin Wild

2. Motivation • Historical trends in direct and indirect radiative forcing by anthropogenic aerosols are highly uncertain. • GCMs have previously estimated the magnitude of aerosol radiative forcing by matching the observed 20th century temperature record. • Unfortunately GCMs cannot obtain a unique solution due to a trade-off between aerosol radiative forcing and climate sensitivity. • Quantification of observed changes in regional aerosol radiative forcing would provide additional constraints on GCM aerosol/cloud parameterizations.

3. Datasets • Monthly downward all-sky SW radiation flux from Global Energy Budget Archive (GEBA) stations during ~1965-2003 (clear-sky not available) • Monthly gridded daytime total cloud amount from ISCCP D2 during 1983-2004 • Monthly gridded downward all-sky and clear-sky SW radiation flux at the surface from ISCCP Flux Dataset (FD) during 1983-2004 • Monthly daytime total cloud cover from synoptic reports at WMO stations during 1971-1996 (Hahn and Warren NDP026D)

4. ×= GEBA station ● = cloud station − = ISCCP grid box

5. SW Flux Estimated from Synoptic Reports • Use synoptic cloud reports to bridge between satellite observations and pre-1983 GEBA measurements estimated SW CRF anomaly = (mean SW CRF / mean cloud cover) × total cloud cover anomaly ) × thickness factor The “thickness factor” accounts for the positive correlation between cloud cover and cloud thickness anomalies

6. SW Flux Estimated from Synoptic Reports correlation = 0.87 no adjustment for thickness

7. SW Flux Estimated from Synoptic Reports correlation = 0.87 ×1.8 thickness factor

8. “dimming” “brightening”

9. “dimming” “brightening”

10. Observational Summary • Cloud cover and related thickness variations dominate surface solar radiation flux on monthly to subdecadal time scales. • Once cloud cover and related effects are removed, the residual time series shows a deceasing trend before 1984 and an increasing trend after 1992. • The residual trend represents changes in clear-sky flux and changes in cloud optical thickness that are uncorrelated to changes in cloud cover. • The most likely cause of the trends is direct and indirect forcing (first effect) by anthropogenic aerosol.

11. CAM3 IPCC AMIP Ensemble • Five simulations for 1950-2000 with prescribed historical SST and same 20th century radiative forcing used in coupled IPCC runs • Time-varying aerosol, but direct radiative forcing only (no influence on cloud microphysical properties) • Monthly output was processed the same way as in the observational study CLDTOT (total cloud) FSDS (all-sky downward solar flux at surface) FSDSC (clear-sky downward solar flux at surface)

12. += CAM3 grid box center − = ISCCP grid box

13. “dimming” “brightening”

14. cyan = individual runs “dimming” “brightening”

15. Conclusions • CAM3 overproduces interannual variations in surface downward solar flux over Europe by a factor of two. • CAM3 does not reproduce the observed solar “dimming” and “brightening” trends over Europe when best guess 20th century anthropogenic aerosol radiative forcing is applied. • Incorrect aerosol input? • Poor representation of transport, sources, and sinks? • Poor representation of clear-sky radiative forcing? • Lack of aerosol influence on cloud albedo?