Abstract
Abstract. A new low-order coupled ocean–atmosphere model for midlatitudes is derived. It is based on quasi-geostrophic equations for both the ocean and the atmosphere, coupled through momentum transfer at the interface. The systematic reduction of the number of modes describing the dynamics leads to an atmospheric low-order component of 20 ordinary differential equations, already discussed in Reinhold and Pierrehumbert (1982), and an oceanic low-order component of four ordinary differential equations, as proposed by Pierini (2011). The coupling terms for both components are derived and all the coefficients of the ocean model are provided. Its dynamics is then briefly explored, through the analysis of its mean field, its variability and its instability properties. The wind-driven ocean displays a decadal variability induced by the atmospheric chaotic wind forcing. The chaotic behavior of the coupled system is highly sensitive to the ocean–atmosphere coupling for low values of the thermal forcing affecting the atmosphere (corresponding to a weakly chaotic coupled system). But it is less sensitive for large values of the thermal forcing (corresponding to a highly chaotic coupled system). In all the cases explored, the number of positive exponents is increasing with the coupling. Two codes in Fortran and Lua of the model integration are provided as Supplement.
Highlights
Low-order models were originally developed to isolate key aspects of the atmospheric and climate dynamics (Stommel, 1961; Saltzman, 1962; Lorenz, 1963; Veronis, 1963)
The systematic reduction of the number of modes describing the dynamics leads to an atmospheric low-order component of 20 ordinary differential equations, already discussed in Reinhold and Pierrehumbert (1982), and an oceanic low-order component of four ordinary differential equations, as proposed by Pierini (2011)
Many low-order models were proposed in various fields of science (e.g., Sprott, 2010), and in particular in climate science (Charney and DeVore, 1979; Nicolis and Nicolis, 1979; Vallis, 1988; Yoden, 1997; Imkeller and Monahan, 2002; Crucifix, 2012). These models allow clarifying important aspects of the underlying structure of the atmospheric and climate dynamics, such as the possibility of multiple stable equilibria (e.g., Simonnet and Dijkstra, 2002; Dijkstra and Ghil, 2005), the possibility of catastrophic events (e.g., Paillard, 1998), or the intrinsic property of sensitivity to initial conditions that led to the development of new approaches for forecasting (Lorenz, 1963; Nicolis, 1992; Palmer, 1993; Trevisan, 1995; Nicolis and Nicolis, 2012)
Summary
Low-order models were originally developed to isolate key aspects of the atmospheric and climate dynamics (Stommel, 1961; Saltzman, 1962; Lorenz, 1963; Veronis, 1963). Building on the latter stream of ideas, Vannitsem (2014) proposed to couple two low-order models for the atmosphere and the ocean, derived from quasi-geostrophic equations. This model is intermediate between the “very loworder” coupled models proposed by van Veen (2003), and the more sophisticated process-oriented low-order coupled models of Lorenz (1984b) and Nese and Dutton (1993).
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
