Radial vorticity constraint in core flow modeling

We present a new method for estimating core surface flows by relaxing the tangentially geostrophic (TG) constraint. Ageostrophic flows are allowed if they are consistent with the radial component of the vorticity equation under assumptions of the magnetostrophic force balance and an insulating mantle. We thus derive a tangentially magnetostrophic (TM) constraint for flows in the spherical harmonic domain and implement it in a least squares inversion of GRIMM-2, a recently proposed core field model, for temporally continuous core flow models (2000.0-2010.0). Comparing the flows calculated using the TG and TM constraints, we show that the number of degrees of freedom for the poloidal flows is notably increased by admitting ageostrophic flows compatible with the TM constraint. We find a significantly improved fit to the GRIMM-2 secular variation (SV) by including zonal poloidal flow in TM flow models. Correlations between the predicted and observed length-of-day variations are equally good under the TG and TM constraints. In addition, we estimate flow models by imposing the TM constraint together with other dynamical constraints: either purely toroidal (PT) flow or helical flow constraint. For the PT case we cannot find any flow which explains the observed SV, while for the helical case the SV can be fitted. The poor compatibility between the TM and PT constraints seems to arise from the absence of zonal poloidal flows. The PT flow assumption is likely to be negated when the radial magnetostrophic vorticity balance is taken into account, even if otherwise consistent with magnetic observations.