GEM-AQ on the global scale

The model was exercised for five years (2001-2005) to evaluate its ability to simulate seasonal variations and regional distributions of trace gases such as ozone nitrogen dioxide and carbon monoxide on the global scale. The model results have been compared with observations from satellites, aircraft measurement campaigns, surface station measurements and ballon sondes. Below are some of the results from this evaluation. (Kaminski et al., to be submitted to ACPD)

(Click to enlarge)

Ozone: 2001 time series of surface measurements from Algoma, Canada (black line) and GEM-AQ (red line). Summer episodes are captured well by the model. There is a slight low bias in winter, indicating the dry deposition rate over snow may be too high.

(Click to enlarge)

Carbon monoxide: comparison of GEM-AQ (left) MOPITT CO (right) at 500 hPa level for April 2002 (top) and October 2002 (bottom). GEM-AQ captures the overall pattern, but misses the biomass burning signatures in Amazonia in April and in Indonesia in October. This is a result of using climatological biomass burning emissions, not year-specific emissions based on fire counts.

GEM-AQ on a regional scale

The "multiscale" aspect of GEM-AQ is demonstrated by simulations done on a global variable resolution grid (an example of this type of grid is shown the figure below). This type of grid was used to study the transport of forest fire smoke in July of 2002 (O'Neill et al., Amos. Env., 2006).

A portion of the global variable resolution grid with ~24km resolution core over eastern US and Canada. The global variable resolution grid allows for high resolution regional simulations without the concerns of domain boundary conditions as in a limited area grid and reduced computational time compared with a global uniform resolution.

(Click to enlarge)

Comparison of MODIS satellite visible image with modelled aerosol optical depth.

Click here for an animation of the simulated smoke plume at 850 hPa, for July 6-15, 2002.

GEM-AQ in Europe

GEM-AQ was also used to study the European heat wave in July of 2006. Modelled ozone and nitrogen dioxide as well as meteorological parameters were compared with surface station measurements during this period. The model was run with a 15 km resolution core over Europe. The figures below are some results from this simulation.
(Struzewska and Kaminski, submitted to ACPD, 2007)

(Click to enlarge)

Modelled vs. observed surface ozone at Rochester, UK.

(Click to enlarge)

Scatter plot of modelled vs. observed maximum temperature.