A new mechanism of vortex intensification by convectively forced vortex
Rossby waves was proposed by Montgomery and Kallenbach (1997, henceforth MK).
The symmetrization process is dependent both on radius and on azimuthal
wavenumber (Smith and Montgomery 1995). As demonstrated by MK, the
axisymmetrization process is described by vortex Rossby-waves which eventually
propagate outward before their symmetrization. MK were able to relate these
waves to intensity changes in hurricanes.
Moeller and Montgomery (1998) verified the linear and quasi-linear
predictions of MK with a nonlinear semispectral shallow-water AB model,
which is based on the AB theory derived by Shapiro and Montgomery (1993).
Much of the observed asymmetries are convective in nature. In our formulation
convection is represented to the extent that the prescribed initial potential
vorticity (PV) anomalies could be convectively forced. We examined the
barotropic axisymmetrization process of different disturbances. In the case
of an initial wavenumber-2 PV asymmetry that was 20% of the basic--state PV we
found an acceleration inside and a deceleration outside the radius of maximum
wind. The PV asymmetries circulate cyclonically around the vortex center and
radiate outward. When the asymmetry was 40% of the basic--state PV we found in
addition a wave-induced eigenmode which interacted with the vortex. In
this case not all of the PV asymmetries are symmetrized away.
The work of Moeller and Montgomery (1998) is being extended to three
dimensions. The fully nonlinear prognostic model can be used to examine
the effect of convectively-induced disturbances on the evolution of a
hurricane-like vortex. This investigation generalizes that of Montgomery
and Enagonio (1998, henceforth ME), who studied tropical cyclogenesis using a
quasigeostrophic balance model, to larger Rossby number. An initially barotropic
or midlevel tropical storm is initialized with a convectively-induced
double-cluster PV asymmetry (negative at upper levels, positive at lower
levels). The symmetrizing PV anomaly induces changes in the primary circulation
analogous to those in the barotropic experiments and is strongest at the top
and bottom of the domain. In addition, the baroclinic disturbance induces a
secondary circulation that counteracts the changes in the primary one. Depending
on the strength of this double cluster the upper-level anticyclonic PV anomaly
is expelled outward (stronger anomaly), which complements the cyclogenesis
results of ME, or is symmetrized (weaker anomaly) similar to the lower-level
positive PV anomaly. Similar to ME, we also simulate the ongoing process of
convection by adding PV anomalies to the PV fields every hour (so-called
"pulsing").
Vortex Rossby waves can propagate not only radially but also vertically. The
model has been also initialized with monochromatic azimuthal wavenumbers 1-3.
Following the dispersion relationship the asymmetries propagate faster upward
the lower the azimuthal wavenumber. While the vortex Rossby waves propagate
they transport energy, momentum and PV, whose dynamical impact has yet to be
clarified.
Results of the three-dimensional experiments will be presented.