Aerosol-climate interactions in the CAM-Oslo atmospheric GCM and investigation of associated basic shortcomings

The paper discusses some challenges in aerosol-climate modelling. CAM-Oslo, extended from NCAR-CAM3, employs an aerosol module for sea-salt, dust, sulphate, black carbon (BC) and particulate organic matter (OM). Primary aerosol size-distributions are modified by condensation, coagulation and wet-phase processes. Aerosol optics and cloud droplet numbers use look-up tables constructed from first principles. Ground level sulphate and sea-salt are generally well modelled, BC and OM are slightly underestimated (uncertain), and dust is considerably (factor ∼2) underestimated. Since non-desert dust, nitrate, anthropogenic secondary organics, and biological particles are omitted, aerosol optical depths (0.12) are underestimated by 10–25%. The underestimates are large in areas with biomass burning and soil dust. The direct and indirect forcing of aerosol increments since pre-industrial time are estimated at +0.031 Wm−2 and −1.78 Wm−2, respectively. Although the total absorption AOD probably is slightly underestimated, the BC contributes to DRF with double strength compared to the AeroCom average. Main reasons for this include: internal BC-mixing (+0.2 Wm−2), accumulation mode BC-agglomerates (+0.05 Wm−2), assumed aitken-mode OM-BC mixture (+0.1 Wm−2), large BC fraction (36%) above 500 hPa, and high low-level cloudiness. Using a prognostic CDNC and process parametrized CCN activation instead of assuming CDNC are equal to CCN, the indirect forcing is 36% smaller.