The impact of atmosphere model resolution on mid-latitude storm track variability

Tuesday, 16 December 2014
Xuelei Feng, George Mason University Fairfax, Fairfax, VA, United States and Bohua Huang, George Mason University Fairfax, Fairfax, Fairfax, VA, United States
The impact of horizontal resolution in an atmospheric general circulation model (AGCM) on mid-latitude storm track simulation is investigated using the European Centre for Medium-Range Weather Forecasts’ (ECMWF) Integrated Forecast System (IFS). In one set of experiments, the IFS is forced with prescribed monthly sea surface temperature (SST) observations for 1960-2007 at three resolutions with T159 (~126km), T511 (~39km) and T1279 (~16km). To assess the effect of air-sea feedback, we have also examined three seasonal hindcast experiments of 7-months long initialized in every May and November for 1982-2010. The hindcasts are conducted using the ECMWF seasonal prediction system with IFS as its atmospheric component, which is set at resolutions of T319 (~62km) T639 (~31km) and T1279 for the three hindcasts respectively.

The climatologically annual cycle of the large-scale precipitation over the Kuroshio and Gulf Stream regions is well simulated by IFS. The mean precipitation is enhanced with increasing atmospheric resolution in both the AGCM and coupled runs. Monthly averaged variance of high-pass filtered daily 250hPa meridional wind is used as an indicator for storm track intensity. In comparison with ERA-40 reanalysis, all simulations overestimate the mean strength of the storm tracks in the North Hemisphere. An EOF analysis of the high-pass filtered variance fields further shows that the two leading modes of the T159 run are characterized by a simultaneous strengthening/weakening in both North Atlantic and Pacific and a seesaw between the two basins respectively, consistent with those from the ERA-40 reanalysis. As the resolution increases, however, the storm track variations in the two basins become more independent. Further analyses are being conducted to connect the storm track variability with the global and regional atmospheric and SST patterns in these simulations.