Tropical and extra-tropical forecast sensitivity to sub-tropical observational enhancement

Tropical circulations have been found to significantly affect the evolution of weather phenomena outside of the tropics (e.g., MJO, El Niño), and therefore provide a potential source of predictability at higher latitudes if correctly depicted. Advances in understanding and predicting tropical convection and associated wind fields have been sought, for example, through special field programs and weather models that better represent tropical-extra-tropical interactions. One way to determine the value of new atmospheric observations in the tropics is to determine whether they improve weather forecasts, or whether new observations have any impact on the forecasts. This study presents a series of real-data forecast experiments based upon the South American Low-level Jet Experiment (SALLJEX), conducted over sub- tropical South America during the Summer of 2002-2003 (Vera et al 2006). A compressible, nonhydrostatic global Euler model that retains little explicit horizontal diffusion is compared to predictions performed by a global primitive equations model and by the regional WRF model for selected cases. The two global research models are run stably out to 2-weeks and allow variable spatial resolution for the case of spherical coordinates. The examined cases include active mesoscale convective events. Preliminary results suggest that:

1) the nonyhydrostatic Euler model displays at least twice as much 48 hour forecast sensitivity to the special field observations as do the regional WRF model or the global PE model; 2) although the field experiment observations were centered in the sub- tropics, some of the strongest shorter term impact on forecast winds occurs in the tropics and then spreads to the extra-tropics later in the forecast; 3) model forecast sensitivity to initial state perturbations is strongly influenced by model diffusion; and 4) forecast sensitivity is strongly affected by moist processes.