Quasi-bi-Hamiltonian structures of the 2-dimensional Kepler problem (Q908279)

The Kepler problem is one of the most important integrable models and many underlying geometrical structures already have been studied. In this paper, the superintegrability of the system is interpreted by means of closed 2-forms \(\Omega_1\), \(\Omega_2\) and a nowhere-vanishing function \(\lambda\), such that the Kepler system vector field \(\Gamma_K\) is conformally Hamiltonian: \[ i(\Gamma_K)\Omega_1=-\lambda dJ_4, \qquad i(\Gamma_K)\Omega_2=\lambda dJ_3, \] where \(J_3\), \(J_4\) are analogous to components of the Runge-Lenz vector. This shows that the Kepler problem is a quasi-bi-Hamiltonian system (see [\textit{R. Brouzet} et al., J. Phys. A, Math. Gen. 29, No. 9, 2069--2076 (1996; Zbl 0911.70012)]). The construction is based on computations of Poisson brackets of appropriate complex functions. For example, in a complex notation, the above equations take the form \[ i(\Gamma_K)\Omega=\sqrt{-1}\lambda dJ, \qquad \Omega=\Omega_1+\sqrt{-1}\Omega_2, \qquad J=J_3+\sqrt{-1}J_4. \] A similar construction in parabolic coordinates leads to quasi-bi-Hamiltonian structures with respect to another two closed 2-forms and integrals \(K_3\) (a component of the Runge-Lenz vector) and \(K_4\) (proportional to the square of the angular momentum and the Hamiltonian function).