“…The boundary condition for the velocity v in (2) is the free boundary condition of the Navier boundary condition. Furthermore, if we add suitable initial data (v 0 , θ 0 ) satisfying divv 0 = 0 and the compatibilty condition (2), then we can prove, see Appendix A and [15,23,24,26,27,29,30,31,35,36], that the initial boundary value problem corresponding to the system (1) possesses a unique solution (v, θ, π) such that v ∈ C([0, T ]; H 3 (Ω)) ∩ L 2 (0, T ; H 4 (Ω)), v t ∈ L 2 (0, T ; L 2 (Ω)), θ ∈ C([0, T ]; H 3 (Ω)), π ∈ L 2 (0, T ; H 1 (Ω)).…”