Yes; the CESM-H run included aerosols (and all other non-CO₂ GHGs); however, it was run for 100-years (see extracted summary below), and while there are questions of spin-up and model bias, it is a projection if the world had stopped emitting CO₂ at 367ppm. However, to get a full ensemble projection we will likely need to wait another 2.5 years when the first phase of the ACME project ends)
http://onlinelibrary.wiley.com/doi/10.1002/2014MS000363/fullExtracted Summary:
"A new high-resolution CESM simulation has been performed for 100 years, the longest such simulation with CAM5. The main highlights of the run are as follows:
1. Equatorial Pacific SST biases are small (mostly <0.2°C), and globally the SST bias relative to observations is quite low compared to a standard resolution case.
2. The power spectrum of Niño3.4 SST time series reveals that the frequency and amplitude of ENSO is comparable to observations, with much less variance than previous CESM/CCSM runs. The typical generation and decay of El-Niño events compares well with the observed record, but the La-Niña events do not show the observed second-year reemergence.
3. There is a notably small SST bias in the equatorial Atlantic where a realistic cold tongue in the eastern basin develops in JJA and the ITCZ keeps mostly north of the Equator. Most climate models (including the CMIP5 generation and CCSM4) have a weak cold tongue or even a reversal of the zonal SST gradient. In this case, much of the improvement is also seen in the lower-resolution version of the model indicating a fundamental change from CCSM4 to CESM.
4. In the ocean, the model eddy field is very rich and comparable to observations in terms of SSH variability: however the SST variability is too high. Ocean fronts are quite well represented: long-standing issues of the Gulf Stream path overshooting at Cape Hatteras and being too zonal in its extension still exist, but the consequent SST biases are considerably reduced compared to standard resolution models.
5. A consequence of having strong SST gradients is that the overlying atmospheric boundary layer and storm track is modified, such that some of the strongest near-surface wind variability in the Globe occurs over the warm side of ocean fronts. In addition, huge amounts of heat are passed from ocean to atmosphere as cold winter continental air passes over western boundary currents. The climatological latent heat flux is, however, too strong over the Gulf Stream in the coupled model. Comparison with an atmosphere-only model reveals that the SST bias of the coupled model leads to the error.
6. Mesoscale Convective Systems over the Rocky Mountains, which contribute to many of the important summertime weather events in the central U.S. are also seen in this simulation, but tend to occur earlier in the year, in spring, and have slower propagation speeds than most observed systems.
7. In common with some previous high-resolution runs [e.g., McClean et al., 2011; Sakamoto et al., 2012], Tropical Cyclones are permitted. The tracks of these extreme events are somewhat realistic except for a striking lack of storms in the N. Atlantic and E. Pacific basins, whilst there are too many strong systems in the western Pacific.
8. The substantial overestimation of Southern Hemisphere summer sea-ice seen in previous CCSM/CESM models is not seen; however, there is a general underestimation of summer sea ice in both high-resolution and standard resolution CESM in the perpetual year 2000 conditions.
Despite these improvements, some limitations of the model simulations persist, such as substantial deep ocean model drift at both high and standard resolutions, excessive precipitation in the ITCZ and wind stress in the extratropical storm tracks, and an overdiffuse Equatorial thermocline. Current model development with CESM is aiming to address these issues.
Current and future plans with these simulations include (i) the analysis of Kuroshio Extension variability on interannual to decadal time scales, and the atmospheric circulation response, (ii) investigation of possible compensation between ocean eddy heat transport and atmosphere heat transport, (iii) Southern Ocean variability, including response of the ocean to changing wind stress in a high-resolution coupled system, (iv) ocean near-inertial wave response to strong winds, (v) relationship of extreme events to large-scale modes of variability, (vi) analysis of Tropical Atlantic climate, and (vii) improvement of regridding of winds onto coastal ocean cells for upwelling studies. In addition, different resolution models are being analyzed (including one with 0.25° resolution in atmosphere, 1° in ocean) to more clearly distinguish the role of resolving ocean features versus resolving atmosphere features in the model improvements seen in CESM-H."