Magnetic dynamos powered by white dwarf superficial convection

When the effective temperature of a cooling white dwarf $T_{\rm eff}$ drops below the ionization limit, it develops a surface convection zone that may generate a magnetic field $B$ through one of several dynamo mechanisms. We revisit this possibility systematically using detailed stellar evolution computations, as well as a simple analytical model that tracks the expansion of the convection zone. The magnetic field reaches a maximum of several kG (for a hydrogen atmosphere) shortly after a convection zone is established at a cooling time $t=t_{\rm conv}$. The field then declines as $B\propto T_{\rm eff}\propto t^{-7/20}$ until the convective envelope couples to the degenerate core at $t=t_{\rm coup}$. We compare the onset of convection $t_{\rm conv}\propto M^{25/21}$ to the crystallization of the white dwarf's core $t_{\rm cryst}\propto M^{-5/3}$, and find that in the mass range $0.5\,{\rm M}_\odot<M<0.9\,{\rm M}_\odot$ the order of events is $t_{\rm conv}<t_{\rm cryst}<t_{\rm coup}$. Specifically, surface dynamos are active for a period $\Delta t\approx t_{\rm cryst}-t_{\rm conv}$ of about a Gyr (shorter for higher masses), before the convection zone is overrun by a stronger magnetic field emanating from the crystallizing core. Our predicted magnetic fields are at the current detection limit, and we do not find any observed candidates that fit the theory. None the less, surface dynamos may be an inevitable outcome of white dwarf cooling, significantly affecting white dwarf accretion and seismology.