A remark on finite volume methods for 2D shallow water equations over irregular bottom topography