Local magnetic anomalies in rugged volcanic terrain explain bias in paleomagnetic data: consequences for sampling

Volcanic rocks are believed to be reliable recorders of changes in the Earths magnetic field in the past. Paleomagnetic data from volcanic edifices is used to make reconstructions of the behavior of the Earth’s magnetic field through time. Recently, however, it became evident that volcanic rocks may not always record the ambient magnetic field accurately. Therefore, we set out to test the accuracy of paleomagnetic data recorded by Mt. Etna lavas by (1) directly measuring the magnetic field above the current topography, i.e. the field that would be recorded by a future flow in that location; and (2) by assessing the paleomagnetic information in historical (1850 – today) flows.

Mt. Etna is characterized by an irregular topography with ridges and gullies that may give rise to local magnetic anomalies that a new flow would record. We measured the ambient geomagnetic field on Mt. Etna with a three-component fluxgate magnetometer at five sites along the length of three paths with an irregular surface. Paths were walked perpendicular to ridges and gullies and measurements were made every meter at different heights above the ground. We found that the declination varies between -17.5 and 18°, and on average differs -3.3° from the expected geomagnetic field. The inclination has a range between 44.4 and 59.5° and is on average 52.3°, while 53.4° is expected. Lastly, the intensity varies between 37.2 and 50.4 µT and is on average 44 µT, with an expected value of 45.2 µT. The deviation with respect to the expected value decreases as function of height above the flow for the inclination and intensity, while the variation in declination does not improve. Most importantly, the variations in inclination and intensity correlate with topographic features: both inclination and intensity are higher above ridges and lower in gullies.

The second part of our study consisted of compiling an overview of literature data on historical Mt. Etna flows and combining this with newly measured paleomagnetic data from 12 sites from 9 different flows with ages between 1865 and 2002. We observed the same trends in this data compilation as in our field observations: the reported inclinations and intensities are often too low.

Our observations have consequences for paleomagnetic sampling strategies in rugged volcanic terrain. To avoid sampling a local magnetic anomaly, samples should be taken spread out over a larger area, preferably meters apart and from different parts of the flow. While this will lead to a higher degree in scatter in paleodirections and a lower precision parameter k, they will better represent the Earth’s magnetic field at the time of cooling. Paleodirections with a high k or paleointensities with a low uncertainty most likely sample a local magnetic anomaly arising from the underlying terrain.