Improvement of Inorganic Aerosol Treatments in CESM/CAM5

Aerosol can affect the earth's climate by scattering and absorption of solar and infrared radiation, and also by serving as cloud condensation nuclei (CCN). Aerosol size distribution and chemical composition are affected by various atmospheric processes such as nucleation, condensation, coagulation, thermodynamic equilibrium, and deposition. New particle formation via nucleation is an important source of CCN, which can affect aerosol indirect radiative forcing. Atmospheric ions can make significant contribution to new particle formation, especially in troposphere and upper stratosphere. The 2005 version of the Carbon Bond Mechanism for Global Extension (CB05_GE) has been previously incorporated into CESM/CAM5 and coupled with the 3- and 7-mode prognostic Modal Aerosol Model (MAM3 and MAM7). There are three nucleation parameterizations used in the 5.1 version of Community Atmosphere Model (CAM5) in the Community Earth System Model (CESM): the empirical power law of Wang and Penner (2009) (referred to as WA09) for nucleation in the planetary boundary layer (PBL) in both MAM3 and MAM7 and the binary H₂SO₄-H₂O homogeneous nucleation of Vehkamaki et al. (2002) (referred to as VE02) in MAM3 and ternary H₂SO₄-NH₃-H₂O homogeneous nucleation of Merikanto et al. ( 2007) (referred to as ME07) in MAM7 above PBL. MAM3/MAM7 includes simplified inorganic aerosol thermodynamics that only involves sulfate and ammonium. In this work, new treatments for new particle formation and aerosol thermodynamics are incorporated into CESM/CAM5 to improve existing inorganic aerosol treatments. An ion-mediated nucleation mechanism based on Yu (2010) (referred to as YU10) has been incorporatered into MAM3 in CAM5, and intercompared with default nucleation parameterizations in CAM5. CESM/CAM5 simulations are conducted with several combinations of nucleation parameterizations (e.g., VE02/WA09, YU10/WA09, and YU10) for 2001 at a horizontal resolution of 1̊× 1̊ and a vertical resolution of 30 layers. The simulations are also evaluated with satellite observations of column CCN and aerosol optical depth (AOD), and land-, ship-, and aircraft-based measurements for new particle formation rates in terms of spatial distribution and performance statistics.

The preliminary results with CAM5/MAM3 show that the nucleation rate in the PBL predicted by YU10 and WA09 is in the range of 10^-2-10^-1 cm³s^-1 in the latitudes of 30-60̊ N. Compared with VE02, YU10 can predict nucleation rates higher by about 2-3 orders of magnitude and aerosol number concentration higher by factors of 2-7 in the upper troposphere and stratosphere. Compared with YU10, WA09 can predict nucleation rates by about 1-3 orders higher and aerosol number concentrations by factors of 1.2-20 higher in the PBL. All these simulations underpredict nucleation rates by ~3 orders of magnitudes, CCN by ~1 order of magnitude, and AOD by ~10%.

The model biases may be attributed to uncertainties in the nucleation parameterizations and early growth mechanism, and simplification of model treatments for cloud microphysics and atmospheric condition in MAM3 (e.g., cloud fraction and meteorology condition). Limited observations of nucleation rates and aerosol number concentrations also introduce some uncertainties in model validation. Additional work will include the incorporation of YU10 and several other parameterizations into MAM7 in CAM5 and intercompare results with those with MAM3. An aerosol thermodynamics module, ISORROPIA II, will also be incorporated into MAM3 and MAM7 to improve aerosol thermodynamic equilibrium treatments.