The goal of this research was to test our claim that the direction of earth’s magnetic field at a site can change very rapidly (several degrees per month) at times during a reversal in geomagnetic polarity.  Our claim was based on a lava flow (which we call “Flow 20”) that erupted and became magnetized parallel to earth’s magnetic field 15.2 million years ago.  At that time, the earth’s magnetic field was in between its two stable states: “normal polarity” like today when compasses point north and “reverse polarity” when compasses point south. By analyzing the way the magnetization of the lava flow was removed by laboratory heating and numerically simulating the flow’s cooling history, we inferred that Flow 20 erupted, cooled for about a year, and then was buried by another lava (Flow 21).  Heat from Flow 21 partially re-magnetized Flow 20 in a direction that differed by 53 degrees from its orientation when Flow 20 was erupted. This finding implied that the geomagnetic field direction was changing at rate of several degrees per month when these two flows were forming.

We tested our interpretation in several ways.  First, we improved our numerical cooling model and conducted many experiments to make sure that our result was robust.  We also obtained measurements of the thermal properties of the rock to confirm the values we used in the numerical model were correct.  Our current estimate of the rate of field is half what we published in 2010 but still orders of magnitude faster than the field has changed in historic times.

We also attempted to examine the magnetization of Flow 20 at other places to determine whether our results from the first study area (“Big Talus”) were typical. We resampled Flow 20 at two spots (20E and 20W) within a few tens of meters of our first profile (20A) and also at third spot (MY) about 2 km to the SE where the lavas are ~35% thicker.  The same two components of magnetization are recognizable at all 3 spots. The simple pattern diagnostic of reheating from above is present at both the 20W and MY profiles, although it is not well enough expressed at MY to permit a convincing comparison to the numerical model. At 20E, the top of Flow 20 appears to have escaped the severe reheating and complete remagnetization observed at 20A, a result that is mystifying. We interpret these results to mean that there are significant lateral variations in the way that Flow 20 cooled.

A third element of the study focused on the nature of the volcanic eruptions that produced these lava flows which were all erupted during the several thousand years it takes earth’s magnetic field to switch polarity. In outcrop, the appearance of the flows varies between thick, reddish and massive to thin, black, and vesicular.  To better understand this difference, we obtained major and trace element analyses of a set of 10 flows distributed through the stack. What these analyses show is that that the silica content of the flows increased or decreased by 2-3% at least seven times during the several thousand years it took the stack to form.  The difference in appearance between Flows 20 and 21 coincides with one such compositional change. Because the variation occurred so frequently as the lava stack was forming, we conclude there is no reason to presume that (in contrast to our hypothesis) that it implies a long interval between the eruption of Flows 20 and 21.

Finally, we worked with a colleague at Scripps Institution of Oceanography to obtain estimates of how strong the geomagnetic field was during or immediately following the episode of rapid field change.  The experimental results were of very low quality, but with hints that the field may have been very weak (perhaps 15% of normal) before the change. Earlier work using a ...

Please report errors in award information by writing to: awardsearch@nsf.gov.