An Intraplate Mystery

So far, I have failed to stress the incomparable value of using models to simulate and predict geospheric responses to climate. A study of particular importance is that of Grollimund and Zoback (2001) whose model was designed to explore the melting of the Laurentide ice sheet as a cause of intraplate seismicity in New Madrid. The Laurentide ice sheet was a prominent feature of the last ice age covering most of Canada and much of North America. At its peak it covered 5 million sq miles and has been a major influence on climate throughout the Quaternary. The map below shows the seismic zone - incorporating the background seismicity and the extent of Reelfoot Ridge.

Paleoliquefaction data have shown that the New Madrid seismic zone experienced three major earthquakes in 1811-12 which appeared to occur on a 200-900 yr cycle prior to the seventeenth century event. The paleo reconstructions contain evidence of severe liquefaction likely to have been caused by a magnitude 7.5 quake or greater. Considering the region does not sit above a plate boundary, a significant trigger must have been present. Accordingly, there have been several proposed theories seeking to explain this phenomenon: a stress concentration from the nearby rift valley (Grana & Richardson, 1996), locally elevated heat flow inducing strain in the lower crust/upper mantle (Liu & Zoback, 1997) and a weak subhorizontal fault above the rift pillow  (Stuart et al., 1997). Nevertheless, these aforementioned hypotheses are unable to explain the sudden seismic upsurge during the Holocene.

The temporal match between the melting of the Laurentide ice sheet (between 19-8 ka) and the escalation of seismic activity prompted James and Bent (1994) to first suggest deglaciation as an alternative theory. Using simple ice-sheet geometry, they noticed that the loss of glacial cover could modify levels of strain several hundred kilometres away. In an attempt to further examine how the Laurentide ice sheet dynamics could have affected seismicity in New Mexico, Grollimun and Zoback (2001) created a three-dimensional finite element model to more accurately simulate lithospheric properties and interactions between large-scale tectonics, deglaciation and the geology.

The main findings of the study concluded that “the perturbation caused by ice loading is to suppress seismicity, whereas ice melting enhances seismic strain release”. The bending of the lithosphere under the weight of ice causes north-south stress beneath the New Madrid seismic zone. This bend generates a shortening of the lower lithosphere in a north-south direction, and an east-west extension. It is the deformation of the lower lithosphere that transmits stress to the upper crust and promotes brittle failure observed in the Holocene seismic record. Overall, the model used in this study has allowed Grollimun and Zoback to say with confidence that the melting Laurentide ice sheet was at least partly responsible for intraplate seismicity in 1811-12. They also conclude that rates of seismic strain in the anthropocene are likely to match high rates of the early Holocene and therefore warn of the high risk of anomalous seismic hazards in the future.