aqua020.brian

1. Control Run

To construct the control run SST, the observed SST climatology* was averaged over the seasonal cycle and across longitudes. The resulting zonal-mean SST was assigned to the whole grid for all times. Then the hemispheres were averaged to remove hemispheric asymmetry. Sea-ice concentration is set to zero everywhere. Latitudes with all missing values (Antarctica) were set to the minimum of the zonal mean climatological SST. The result is shown in this figure: [png][eps].

To compare with the standard aqua planet SST configurations (and those of the APE), here is a figure that shows the built-in, symmetric SST distributions for the CAM aqua planet mode along with this one (black, dotted curve) [png][eps]. To compare with the real climatology, this plot shows the SST climatology (used by default in CAM), this aqua planet SST, and the difference [png][eps].

The model is then started as before. It uses the symmetric ozone data set, but the default initial conditions. That should adjust relatively quickly.

The early results show a net surface imbalance of about 10 Wm2 globally. These two figures show the surface energy balance components (zonal means) for each of the first 57 months of the run [png][eps] and the time series of Fnet [png][eps]. (The last panel and the last point in the figures is the climatological values.)

2. Calculating Q-flux

To make a first guess at the q-flux, I take about 2 years of the control run (years 3-5) and calculate the net surface flux, Fnet. Then I take the zonal average and then average the hemispheres. Then I use an iterative process to slowly adjust toward a vanishing global average value. This figure [png][eps] shows the two curves; the dashed curve is the one with approximately zero mean. The red, dotted curve shows a smoothed version of the solid curve, using the first 3 Fourier coefficients.