We study the quasigeostrophic merging dynamics of axisymmetric barocli
nic vortices to understand how baroclinicity affects merging rates and
the development of the nonlinear cascade of enstrophy. The initial vo
rtices are taken to simulate closely the horizontal and vertical struc
ture of Gulf Stream rings. A quasigeostrophic model is set with a hori
zontal resolution of 9 km and 6 vertical levels to resolve the mean st
ratification of the Gulf Stream region. The results show that the baro
clinic merging is slower than the purely barotropic process. The mergi
ng is shown to occur in two phases: the first, which produces clove-sh
aped vortices and diffusive mixing of vorticity contours; and the seco
nd, which consists of the sliding of the remaining vorticity cores wit
h a second diffusive mixing of the internal vorticity field. Compariso
n among Nof, Cushman-Roisin, Polvani et al., and Dewar and Killworth m
erging events indicates a substantial agreement in the kinematics of t
he process. Parameter sensitivity experiments show that the decrease o
f the baroclinicity parameter of the system, GAMMA2 [defined as GAMMA2
= (D2f0(2))/(N0(2)H2)], increases the speed of merging while its incr
ease slows down the merging. However, the halting effect of baroclinic
ity (large GAMMA2 or small Rossby radii of deformation) reaches a satu
ration level where the merging becomes insensitive to larger GAMMA2 va
lues. Furthermore, we show that a regime of small GAMMA2 exists at whi
ch the merged baroclinic vortex is unstable (metastable) and breaks ag
ain into two new vortices. Thus, in the baroclinic case the range of G
AMMA2 determines the stability of the merged vortex. We analyze these
results by local energy and vorticity balances, showing that the horiz
ontal divergence of pressure work term [del . (pv)] and the relative-v
orticity advection term (v . deldel2psi) trigger the merging during th
e first phase. Due to this horizontal redistribution process, a net ki
netic to gravitational energy conversion occurs via buovancy work in t
he region external to the cores of the vortices. The second phase of m
erging is dominated by a direct baroclinic conversion of available gra
vitational energy into kinetic energy, which in tum triggers a horizon
tal energy redistribution producing the final fusion of the vortex cen
ters. This energy and vorticity analysis supports the hypothesis that
merging is an internal mixing process triggered by a horizontal redist
ribution of kinetic energy.