Nonlinear alpha (2)Omega -dynamo waves are considered in a thin turbulent,
differentially rotating convective stellar shell. Nonlinearity arises from
cu-quenching, while an asymptotic solution is based on the small aspect rat
io of the shell. Wave modulation is linked to a latitudinal-dependent local
cu-effect and zonal shear flow magnetic Reynolds numbers R(alpha)f(theta)
and R(Omega)g(theta) respectively; here B is the latitude. The study is a d
irect extension of that of Meunier et al. (1997) for alpha Omega -dynamo wa
ves which corresponds to finite dynamo number R alpha R Omega in the limit
R alpha --> 0.
The essential picture developed is that of a modulated dynamo wave whose am
plitude varies spatially with theta. The linear solution is controlled by t
he properties of the double turning point B, of the ordinary differential e
quation for the mode amplitude. Significantly, though B, is real and is loc
ated at the local dynamo number maximum in the alpha Omega -dynamo limit R-
alpha --> 0, it migrates into the complex theta -plane once R-alpha not equ
al 0. Linear and weakly nonlinear solutions are found over a limited range
of R, and their qualitative properties are found to be largely similar to t
hose for the alpha Omega -dynamo limit. One significant astrophysical diffe
rence is the fact that the frequency generally decreases with increasing R-
alpha. Thus alpha (2)Omega -stellar dynamos may occur with alpha Omega -dyn
amo wave characteristics but exhibit significantly longer cycle times incre
ased by a factor roughly two or more.
Finite amplitude dynamo waves, like those when R-alpha --> 0, are modulated
by an envelope which evaporates smoothly at some low latitude but is termi
nated abruptly by a front at a high latitude BF Significantly, for given no
n-zero R,, these frontal solutions are subcritical (a property linked to th
e complex-value taken by B,). For sufficiently large R,, however, new low f
requency modes emerge that are more closely related to steady alpha (2)-dyn
amos localised near the pole theta = pi /2. In these circumstances, up to f
our distinct finite amplitude states are identified; they may be loosely ch
aracterised as alpha Omega -high frequency, alpha (2)Omega -medium frequenc
y, alpha (2)-low frequency and alpha (2)-steady modes. In view of the possi
ble mode competition, we comment on the likely realise physical state.