Topographic Influence on the Predictability of Baroclinic Cyclones

The fundamental role that topography plays on forcing a wide range of phenomena in the Earth's climate system cannot be understated. Despite this importance, the overall influence of topography on the predictability of weather systems is largely unknown. While it has been argued that topography is a known forcing that may constrain errors from growing in forecast models, recent studies have linked terrain to uncertainty in forecasts with scales ranging from the sub-mesoscale to synoptic scale.

In this study, we examine the impact of topography on forecast errors associated with initial condition uncertainties in an idealized baroclinic fluid on an f-plane evolving in a channel. A series of experiments is conducted with the Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS) to: (1) determine the role that topography plays on the storm track, in particular, the process of lee-cyclogenesis, and (2) to characterize the fundamental role that topography plays in shaping the evolution of initial condition perturbations. To address the first objective, a set of 400-day idealized simulations with and without topography is performed with COAMPS for a baroclinic fluid in a channel. In the simulation with topography, processes associated with lee cyclogenesis lead to an enhancement of the storm track intensity well downstream of the barrier and overall greater variability of the baroclinicity. To address the second objective, a series of ensemble forecasts are performed in which the 400-day free run is considered ‘truth' and a cycling ensemble Kalman filter data assimilation system is used to construct the initial perturbations. Observations are drawn from the truth run implying that forecast errors are solely associated with the initial condition perturbations. Every two days over the 400-day period, the ensemble is forecast to 120-hours and the error-growth statistics are analyzed. The implications of these results for predictability in full numerical weather prediction models will be discussed.