Hemispheric Coupling: Comparing Dynamo Simulations and Observations

Numerical simulations that reproduce solar-like magnetic cycles can be used to generate long-term statistics. The variations in north-south hemispheric solar cycle synchronicity and amplitude produced in simulations has not been widely compared to observations. The observed limits on solar cycle amplitude and phase asymmetry show that hemispheric sunspot area production is no more than 20 % asymmetric for cycles 17-23 and that phase lags do not exceed 20 % (or two years) of the total cycle period, as determined from Royal Greenwich Observatory sunspot data. Several independent studies have found a long-term trend in phase values as one hemisphere leads the other for, on average, four cycles. Such persistence in phase is not indicative of a stochastic phenomenon. We compare these observational findings to the magnetic cycle found in a numerical simulation of solar convection recently produced with the EULAG-MHD model. This long "millennium simulation" spans more than 1600 years and generated 40 regular, sunspot-like cycles. While the simulated cycle length is too long (∼40 yrs) and the toroidal bands remain at too high of latitudes (>30°), some solar-like aspects of hemispheric asymmetry are reproduced. The model is successful at reproducing the synchrony of polarity inversions and onset of cycle as the simulated phase lags do not exceed 20 % of the cycle period. The simulated amplitude variations between the north and south hemispheres are larger than those observed in the Sun, some up to 40 %. An interesting note is that the simulations also show that one hemisphere can persistently lead the other for several successive cycles, placing an upper bound on the efficiency of transequatorial magnetic coupling mechanisms. These include magnetic diffusion, cross-equatorial mixing within latitudinally-elongated convective rolls (a.k.a. "banana cells") and transequatorial meridional flow cells. One or more of these processes may lead to magnetic flux cancellation whereby the oppositely directed fields come in close proximity and cancel each other across the magnetic equator late in the solar cycle. We discuss the discrepancies between model and observations and the constraints they pose on possible mechanisms of hemispheric coupling.