A dynamical system derived from a conceptual climate model is investigated. The model is often referred to as the Budyko-Sellers-type energy balance model (EBM), or in short, the Budyko model. The focus of this study is on a mathematical formulation that further develops the model by capturing the ice line dynamics and thereby places the Budyko model within the dynamical systems framework. The classical Budyko EBM is represented here as an integro-difference equation, then coupled with an ice line equation. The resulting infinite dimensional system is shown to possess an attracting one dimensional invariant manifold. The novel formulation captures solutions that others have previously obtained while also making the formulation and the analysis of the model more precise.
Introduction.We investigate a dynamical system derived from a conceptual climate model. The model is a version of what is known in climate science as the Budyko-Sellers-type energy balance model (EBM). The result of this study is purely mathematical, with the main focus on the mathematical formulation that further develops the model and places it in the framework of dynamical systems. In particular, we represent the classical Budyko model as an integro-difference equation, then couple it with an auxiliary equation that prescribes the ice line dynamics. We analyze the stability of the coupled system and prove that, under certain conditions, the infinite dimensional system has a one dimensional attracting invariant manifold.Budyko-Sellers-type EBMs model the annual average temperature distribution and rest on the idea that the energy received by the planet is balanced by the energy it emits. Any imbalance in the system will result in a temperature change. The state variable of a BudykoSellers-type EBM is a latitudinally averaged temperature distribution, often referred to as a temperature profile. One feedback admitted by a Budyko-Sellers-type EBM is the ice albedo feedback, which is a positive feedback caused by the difference between the reflectivity of ice cover and that of the ocean. Much effort has been devoted to the steady state analysis of differential equations based on the two conceptual models pioneered independently by Budyko [3] and Sellers [22] in late 1960s; see, e.g., [5,17,23,9]. Here, we are concerned with