%0 Article in Journal/Newspaper %A Sprain, CJ %I Elsevier %D 2019 %G English %T An assessment of long duration geodynamo simulations using new paleomagnetic modeling criteria (Q PM) %U https://eprints.whiterose.ac.uk/151558/ %U https://eprints.whiterose.ac.uk/151558/1/1-s2.0-S0012821X19304509-main.pdf %X Long-term temporal variations of the magnetic field (timescales >10 Myr), characterized from paleomagnetic data, have been hypothesized to reflect the evolution of Earth's deep interior and couplings between the core and mantle. By tying observed changes in the paleomagnetic record to mechanisms predicted from numerical geodynamo simulations, we have a unique tool for assessing changes in the deep interior back in time. However, numerical simulations are not run in an Earth-like parameter regime and assessing how well they reproduce the geomagnetic field is difficult. Criteria have been proposed to determine the level of spatial and temporal agreement between simulations and observations spanning historical and Holocene timescales, but no such criteria exist for longer timescales. Here we present a new set of five criteria (Quality of Paleomagnetic Modeling criteria, QPM) that assess the degree of semblance between a simulated dynamo and the temporal and spatial variations of the long-term (∼10 Myr) paleomagnetic field. These criteria measure inclination anomaly, virtual geomagnetic pole dispersion at the equator, latitudinal variation in virtual geomagnetic pole dispersion, normalized width of virtual dipole moment distribution, and dipole field reversals. We have assessed 46 geodynamo simulations using the QPM criteria. The simulations have each been run for the equivalent of at least ∼300 kyr, span reversing and non-reversing regimes, and include either homogeneous or heterogeneous heat flux boundary conditions. We find that none of our simulations reproduce all salient aspects of the long-term paleomagnetic field behavior for the past 10 Myr. Nevertheless, our simulations bracket Earth values, suggesting that an Earth-like simulation is feasible within the available computationally accessible parameter space. This new set of criteria can inform future simulations that aim to reproduce all aspects of Earth's long-term magnetic field behavior.