Jon Aurnou (UCLA) & the Geodynamo Team

Over the past 20 years, researchers have made great strides in simulating convection-driven dynamo action.  They have modeled convection of electrically-conducting fluid in a rotating spherical shell by solving the full set of dynamo equations.  However, the parameter values employed in these dynamo models are very far from being Earth-like; they do not adequately approach the extreme values that describe planetary settings. Typical present day core dynamics simulations can resolve `quasi-laminar' convection at moderate rotation rates. In contrast, in Earth's core turbulent convective motions are strongly modified by the effects of planetary rotation. Because the core's extreme parameter values are presently inaccessible in dynamo models, the rapidly rotating turbulent magnetoconvection that exists in Earth's core has never been simulated. The discrepancy between the turbulence that exists in the core and the quasi-laminar flows that exist in geodynamo simulations raises the question: Are the current generation of dynamo simulations modeling convection that is relevant to actual core processes?

To address this question, CIG's Dynamo Working Group has been developing two new codes, Calypso and Rayleigh, in order to provide the broader community with the ability to carry out massively-parallelized dynamo simulation that can more accurately simulate Earth's core conditions.  In addition, we are carrying out a community-wide accuracy-performace dynamo benchmark exercise to determine which dynamo simulation methods are best suited for the development of next generation codes that will be well-suited to peta- and exa-scale computing platforms.

[All 2013-14 webinars] [YouTube]

1. Aurnou
2. webinar
3. 2014