I also proposed that geomagnetic jerks observed at the Earth's surface be viewed as singularities in the time behaviour of the geomagnetic field. Such a signal should have suffered some distortion when diffusing from the top of the core through the conducting mantle. Assuming the upper mantle as an insulator, and given the electromagnetic time constant of the mantle and a thickness of 2000 km for a uniform lower conducting mantle, I proposed a value for the mantle electrical conductivity smaller than , still a considerable value.
When computing the fluid flow at the CMB over the last four centuries, the flow pattern conserves some broad features over this whole time-span. However, coming to a more accurate description of the last decades, the geometry of the acceleration fields corresponding to the well documented geomagnetic jerks are similar. Furthermore, over the last decade covered by satellite data, I have been interested by the very rapid changes in the magnetic field acceleration, a new source of information on the core flows. I have also shown that connections between geomagnetic jerks detected in the core field acceleration, torsional oscillations and/or core-mantle coupling are important, but I note that all those phenomena must somehow be connected to each other.