1: \begin{abstract}

2:   We revisit the issue of sensitivity to initial flow and intrinsic

3:   variability in hot-Jupiter atmospheric flow simulations, originally

4:   investigated by \citet{Choetal08} and \citet{ThraCho10}.  The flow

5:   in the lower region ($\sim$1 to $20$~MPa) `dragged' to immobility

6:   and uniform temperature on a very short timescale, as in

7:   \citet{LiuShow13}, leads to effectively a complete cessation of

8:   variability as well as sensitivity in three-dimensional (3D)

9:   simulations with traditional primitive equations.  Such momentum

10:   (Rayleigh) and thermal (Newtonian) drags are, however, ad hoc for 3D

11:   giant planet simulations.  For 3D hot-Jupiter simulations, which

12:   typically already employ strong Newtonian drag in the upper region,

13:   sensitivity is not quenched if only the Newtonian drag is applied in

14:   the lower region, without the strong Rayleigh drag: in general, both

15:   sensitivity and variability persist if the two drags are not applied

16:   concurrently in the lower region.  However, even when the drags are

17:   applied concurrently, vertically-propagating planetary waves give

18:   rise to significant variability in the $\sim$0.05 to $0.5$~MPa

19:   region, if the vertical resolution of the lower region is increased

20:   (e.g. here with 1000 layers for the entire domain).  New

21:   observations on the effects of the physical setup and model

22:   convergence in `deep' atmosphere simulations are also presented.

23: \end{abstract}

24: