The TRR 181 seminar is held by Lesley de Cruz (Royal Meteorological Institute of Belgium), followed by a Career Talk Lunchon November 21, 11 am at Universität Hamburg, Bundesstr. 53, room 22/23.
Understanding the effect of ocean-atmosphere interactions on predictability using the Modular Arbitrary-Order Ocean-Atmosphere Model (MAOOAM)
Numerical weather prediction has traditionally focused on accurately describing the atmospheric dynamics while considering the ocean as a nearly-constant, independent backdrop. However, this approach has changed radically in recent years: the European Centre for Medium-range Weather Forecasts (ECMWF), for example, has switched to a fully coupled ocean-atmosphere operational model last year, and is investigating fully-coupled data assimilation. This shift in focus is due to the great added value from an accurate description of the ocean-atmosphere coupling. Indeed, the atmosphere's coupling with the ocean strongly affects its predictability.
There are many approaches to understanding the dynamics of the coupled
ocean-atmosphere system, ranging from the analysis of observations and
full-blown earth system models to reduced-order models. In the reduced-order
modelling approach, one aims to reduce the system to the key ingredients that
reproduce the observed variability, while respecting the governing equations
dictated by the physics.
The oceanic influence is apparent from the presence of long-term variability in large-scale weather patterns. This low-frequency variability occurs on time scales between those of typical weather events (days) and those on which climate is defined (~30 years). The ocean is well known to play an important role in some of these variations, such as El Niño - Southern Oscillation (ENSO). While the source of some other variability patterns such as the North-Atlantic Oscillation (NAO) is less clear, recent work with a low-order model has indicated that this kind of low-frequency variability can occur purely as emergent behaviour due to the coupling between the ocean and the atmosphere (Vannitsem et al, 2015).
To investigate the predictability properties in different model configurations, we have developed a flexible reduced-order model called MAOOAM. It features a two-layer atmosphere over a shallow-water ocean layer with passively advected temperature. It incorporates both frictional coupling and an energy balance scheme which accounts for radiative and heat fluxes between ocean and atmosphere. The availability of the tangent linear and adjoint models makes MAOOAM an ideal testbed for data-assimilation experiments in the coupled ocean-atmosphere system, and to study its stability properties. Together with researchers from Universität Hamburg, we have used MAOOAM to investigate the Lyapunov exponents, which determine the growth/decay rates of errors in the coupled system.
Changing the basis functions of MAOOAM also allows us to modify the boundary
conditions, changing the geometry from an ocean basin (cfr. the Atlantic) to a channel ocean (cfr. the Southern Ocean). I will present our recent study of the influence of these boundary conditions, which has shown that there are interesting alternative paths to long-term predictability.