New toy problem of a convectively propagate flame, where rotation
keeps it localized.

There are several parameters that affect the dynamics of this problem:

 -- atmosphere

    * rho_base
    * p_base
    * g

 -- reaction rate, simple powerlaw: r = r0 rho X T**nu
    with eps = Q_burn r / rho

   * nu
   * Q_burn
   * r0

 -- rotation:

   * omega



Numbers / quantities of interest:

   * Rossby,  Ro ~ U/(L Omega)
   
   * scale height, H ~ p_base / (rho_base g)

   * burning front speed? thickness?

   * sound speed: c ~ \sqrt(p_base / rho_base)

     do we want t_cs ~ H/c ~ 1?

  