Transforming our understanding of the geodynamo

Earth's magnetic field is essential to life on our planet because it provides protection from harmful cosmic rays and the solar wind that bombard the upper atmosphere. But this invisible protective field is far from constant. Its intensity and direction are always changing both temporally and spatially. On decadal timescales and longer these changes can be attributed to the interaction between Earth’s rotation about its axis and convective motion of its molten interior– its geodynamo. Yet our understanding of the geodynamo and the mechanisms by which it drives changes in our magnetic field is primitive. The main problem preventing progress is a lack of a robust data set predating the Holocene epoch (last ~10,000 years, 10kyr). For the Holocene, we have a relatively well dated compilation of records, largely from archaeological and lake archives, showing millennial timescale changes in magnetic intensity and direction from a network of sites around the world. This permits development of time varying geomagnetic field models on a global scale. These exciting developments provide a way to transform our understanding of the geodynamo but a Holocene focus is severely limiting because we need a longer perspective to study a fuller range in field variability. Earth’s magnetic field may have lost half its strength since Roman times but much larger changes in intensity occurred during the Pleistocene epoch. Furthermore, there has not been an undisputed full (180 deg.) change in field direction or polarity excursion since the Laschamp excursion (~40 kyrs ago) and that event is just one of at least five well-documented excursions that took place during the last ~500 kyrs. We propose to transform our understanding of the geodynamo and the cause of changes in Earth’s magnetic field by turning to this Late Pleistocene record of changes in field intensity and direction. Marine sediment drillcores are well-suited to providing the long undisturbed records needed for this purpose but, until now, it has not been possible to date them sufficiently robustly to work globally at the resolution needed.