Direct and indirect effects of aerosol on climate simulation

Abstract

The aerosol direct and indirect effect studies are complemented with simulations using the GEOS-5 GCM recently upgraded with double moment cloud microphysics (i.e., the ability to predict both cloud water content and particle numbers), interactive GOCART aerosol model, advance radiative transfer package RRTMG with Monte Carlo Independent Column Approximation modes, and CFMIP Observation Simulator Package (COSP). Comparisons of GEOS-5 integrations using the schemes allow us to identify and separate consistent (and therefore more robust) responses of monsoonal circulations, clouds and precipitation to aerosol rather than dynamics, and meteorological seasonality. Especially biomass burning (BB) intensity is strongly correlated with large-scale circulation change, so GCM experimental designs that will isolate the aerosol effects are necessary. A baseline comparison is first performed between near decade-long GEOS-5 runs using the two different single (GEOS-5 Standard) and double moment (McRAS-AC) cloud microphysical schemes and prescribed observed SSTs for the period 2002-2011 which has extensive multi-satellite coverage. Two 10-year long simulations bring to light discernible improvements in the TOA zonal radiation budget when the latter scheme is implemented. These and other encouraging results from the simulation suggest that McRAS-AC exhibits significant skills and has the potential to be further improved with targeted enhancements. And we conducted a similar segregation between anomalously high and climatological emission days for GEOS-5 in order to determine, with the aid of the COSP satellite simulator package, which cloud scheme can better reproduce observed cloud response to enhanced BB aerosol emissions. Equipped with this knowledge, we could proceed to a new series of experiments for a more in-depth analysis of cloud and precipitation response to extreme BB emission scenarios Taking appropriate differences between AGCM experiment sets we find that BB aerosols affect liquid clouds in statistically significantly ways, increasing cloud droplet number concentrations, decreasing droplet effective radii (i.e., a classic aerosol indirect effect), and locally suppressing precipitation due to a decelerate of the autoconversion process, with the latter effect apparently also leading to cloud condensate increases. Geographical re-arrangements of precipitation patterns, with precipitation increases downwind of aerosol sources are also seen, most likely because of advection of weakly precipitating cloud fields. Somewhat unexpectedly, the change in cloud radiative effects (cloud forcing) is in the direction of less cooling because of decreases in cloud fraction. Overall, however, because of direct radiative effect contributions, aerosols exert a negative forcing at both the top of the atmosphere and, perhaps most importantly, the surface, where decreased evaporation triggers feedbacks that further reduce precipitation. Invoking the approximation that direct and indirect aerosol effects are additive, we estimate that the overall precipitation reduction is about 40% due to the direct effects of absorbing aerosols which stabilize the atmosphere and reduce surface latent heat fluxes via cooler land surface temperatures. Further refinements of our two-moment cloud microphysics scheme are needed for a more complete examination of the role of aerosol-convection interactions in the seasonal development of the SE Asia monsoon.