February 8, 2012

Legendary phenomena in the 1811-1812 New Madrid earthquakes

Posted by Austin Elliott

200 years ago this February 7, on the western frontier of European settlement in North America, the pioneering westward expanders and the natives whose land they were colonizing were thrown from their sleep in the deep wee hours of a winter night by the culminating temblor of a harrowing, months-long sequence of major earthquakes, aftershocks of which continue to this day.

Map of shaking intensity interpolated from historic accounts of the 2:15am mainshock of the New Madrid sequence. Map courtesy Susan Hough, USGS.

The so-called New Madrid earthquakes–named for a small Missouri settlement near the modern-day borders of Kentucky, Tennessee, Illinois, Indiana, and Arkansas that lay nearest the center of this cataclysmic seismic sequence–are the largest to have struck the eastern United States since well before they became the United States. In the recorded history of western settlement of North America, no quakes outside of the mountainous west match them in size and scope, and only a few come close.

Plenty of people have been and will be reporting on these earthquakes as we celebrate their bicentennial, including the organizers of the Great Central U.S. ShakeOut, which took place this morning to commemorate the massive culminating temblor of the sequence that started in December 1811. Even mapping software purveyor ESRI has put together a commemorative compilation of informative and beautiful interactive maps about the quakes (super cool compilation! If you click on one link in this post, let it be that one).  It is worth reading some of these syntheses and reviews because the earthquake series itself makes a captivating narrative. It’s nearly impossible to imagine the terror with which these relentless temblors must have stricken the settlers, who were already braving the “wild” frontier of a foreign continent. Even the mid-continent’s native inhabitants had not experienced such a thing in scores of generations, and in the early 19th century no one would have had any reasonable framework in which to explain the occurrence of massive earthquakes.

Because the New Madrid quakes occurred so early in our country’s recorded and geographic history, piecing together the events with a modern understanding of earthquakes and plate tectonics has required a great deal of sleuthing, and some of the details gleaned about them remain controversial, most notably their magnitudes (were they more like M7 or more like M8?). The uncertainty regarding the exact size of these earthquakes compounds the issue of determining the seismic hazard posed by recurrence of major earthquakes in the New Madrid Seismic Zone. To understand how seismicity may continue in southeastern Missouri we can look for patterns in the prehistoric record of earthquakes, but ideally we would like some idea of what forces caused these earthquakes to happen here. This remains an open question, and one in particular for which the question of the quakes’ magnitudes may be a crucial bit of information. Researchers have tried to use modern seismicity to constrain the behavior of large earthquakes in the New Madrid Seismic Zone, and some have interpreted the ongoing small quakes there as the tail end of an unsurprising aftershock sequence, suggesting that they don’t represent heightened seismic risk, but that in fact New Madrid is as likely as any number of other places in the eastern U.S. to have more major temblors.

The ongoing scientific controversy over ambiguous interpretation of details of these quakes stems from the nature of the data. Researching “pre-instrumental” earthquakes is a pursuit that fuses seismology, history, and social science, in an effort to understand historic written accounts of the earthquakes in the context of their time and cultural setting. A somewhat recent article in Seismological Research Letters describes the endeavor of anecdotal seismology, and through some colorful examples illustrates how historical reports can be translated into seismological data, clarifying the sources of interpretive ambiguity. The marriage of historical and seismological research to inform our model of seismic events in the eastern U.S. could be and has been the subject of many volumes, so I can’t hope to cover it here.

Instead I’ll draw analogy to this incredible sequence of earthquakes through videos and pictures from recent events, hopefully grounding some of the legendary accounts in footage of real and recognizable phenomena.

To the extent that people have learned about the New Madrid earthquakes of 1811-1812, they have often heard of them referred to as the largest quakes to ever strike the U.S.  Ask California [1857 & 1906] and Alaska [too many to name] and you’ll find this claim is far from true. Along with this hyperbolic appraisal comes the legendary confluence of phenomena eyewitnesses allegedly reported: the Mississippi running backwards, giant fountains of water issuing from the Earth, trees being thrown to the ground, and land sinking into the river. The unimaginable chaos of these phenomena all occurring in the midst of violent shaking defies belief, but contemporary earthquakes and modern video recording technology allow us to ground them in reality, and perhaps to understand them as more modest individual events that have been amplified in intensity by their conflation and coincidence in legend. We can see examples of all four in much more modest earthquakes:

1. The Mississippi running backwards

It’s difficult to imagine what possible physical phenomenon could have led to this observation/claim… unless you understand that the New Madrid quakes–just like all other large temblors–resulted from slip along several geologic faults. At the surface, fault slip breaks and displaces the ground, moving one side in a direction opposite the other. In the case of the causative Reelfoot Fault, the surface trace cut right across the Mississippi River channel, dropping an upstream portion of the river relative to the adjacent reach downstream. This warping has been thoroughly investigated and modeled, and thanks to the September 4, 2010 Darfield earthquake–a M7.1 event that ripped across rivers on New Zealand’s flat Canterbury Plain–we have a beautiful modern analog of the occurrence.

Aerial view of the Horata River spilling off of the fault scarp formed by the September 24, 2010 Darfield earthquake in New Zealand. Image courtesy Dr. Mark Quigley, University of Canterbury, Christchurch, NZ.

Where the 2010 NZ rupture fault sliced across the Horata River, it diverted the water into surrounding farmland, effectively changing the course of the flow. This is precisely analogous to the diversion of the Mississippi that led to both the damming and formation of Reelfoot Lake, and the temporary diversion of river flow back upstream.

2. Fountains of water issuing from the Earth

There are a few processes that may combine to produce this effect. In the past year we’ve seen plenty of examples of sand volcanoes, the eruptive results of shaking-induced soil liquefaction. When subjected to seismic waves (as in this New Zealand aftershock, or the Tohoku quake below), these sand blows can be squeezed into fountains of substantial height. The force of a larger and longer earthquake would undoubtedly increase the height these reach.


Extrusion of liquefied sediment by seismic waves isn’t the only coseismic phenomenon that may throw water high into the air: seiching–harmonic oscillation–of small bodies of water may throw water against their banks and up into the air. We’ve seen this dramatically demonstrated in swimming pools during a M7.2 earthquake, but natural ponds don’t necessarily have the splashing power of sharp corners and hard edges in concrete-walled pools. Nonetheless, with these two phenomena operating in tandem, the amount of water being thrown into the air by the quake would certainly be fodder for tales–legendary or not–of high fountains from the Earth.

3. Trees being thrown to the ground

Videos from several modest (M ~6) earthquakes in the past few years have revealed just how much trees can be wrenched around during shaking. Under the accelerations of earthquakes, trees’ own weight can be a more powerful force than high winds. Here a stand of neighborhood trees sways in a mere 4.4 earthquake in Christchurch:


In a M6 we see through the windows the same effect:


Finally, video the USGS captured at practically the epicenter of the M6.0 2004 Parkfield earthquake shows fairly violent lashing of late summer oaks in the California Coast Ranges.


A tree along the San Andreas Fault in Wrightwood, CA, had its top snapped off in an 1812 earthquake, from which it grew two new crowns. The tree no longer exists, but others like it can be found along the 1906 rupture near Point Arena in NorCal. Image from "Mixed Matters"

Though the effects shown above do not amount to trees being thrown to the ground, the earthquakes that produced them were much smaller than the ones that struck Missouri. We have clear evidence along the San Andreas Fault of trees whose tops were snapped off during the 1906 earthquake. This is a common effect in the epicentral region of large quakes.

4. Land sinking into the river

This phenomenon is akin to but distinct from the Mississippi being diverted and running backwards. In fact the underlying process is more closely related to the processes that give rise to sand blows. Shaking liquefies water-saturated soils and they lose their shear strength, rendering them unable to support gravitational loads. Thus the land slumps, under its own weight or the weight of trees, houses, or riverboat moorings, downhill towards unencumbered free edges like river banks. This “lateral spreading” is commonly observed along river banks shaken by earthquakes, and results in lowering and inundation of the ground surface. Examples abound from earthquakes as geographically and tectonically various as the 1964 Good Friday event in Alaska, the 1906 San Francisco earthquake, the 2010 Haiti earthquake, and the 2010 El Mayor-Cucapah earthquake. In all of these events vast swaths of land shook loose and slumped ocean- or river-ward, and effectively “sank”.


The video examples compiled above may not match the apparent drama of those recounted from 1811-12 Missouri, but I find it easy to imagine the cumulative results of decades and decades of re-telling on the details of these accounts. In any case, large earthquakes produce remarkable effects, and although many people around the world witness or experience earthquakes, still relatively few witness the truly violent shaking that occurs near an earthquake’s source. Written and oral accounts give us the most thorough picture, even if we have to take them with a grain of salt. Video may gradually be replacing verbal accounts in objectivity (no relying second-hand information!), but it has yet to become as widely distributed and available as individual eye-witnesses.

Next time you strike up a conversation about these earthquakes, consider yourself informed about many of the features that defined them, but by all means gather more information on your own. My two favorite informative links are the following:

ESRI maps of the 1811-1812 New Madrid quakes:

CERI compendium of New Madrid primary sources:

Happy Bicentennial!