The following piece is part of the Quest’s new series featuring final projects of Minerva students. This piece was written for Minerva’s Geobiochemiphysics: Integrating Earth’s Systems course by Kayla Cohen, Minerva Class of 2019.  Kayla is a double-major student of Arts and Humanities and Natural Sciences. To view more final projects, click here. If you are a Minerva student and would like to have your final project published,  fill out this form

A Comprehensive Introduction to Plate Tectonics, Mantle Scars, and New Madrid Seismic Zone

Fifty-two years ago, geologist J. Tuzo Wilson made a discovery that would change the way scientists understand oceans, continents, and landforms forever. His work is distilled in the theory of “Plate Tectonics” and it describes how Earth’s mantle is constantly in motion. In Wilson’s seminal essay “Did the Atlantic Close and then Reopen?” he observes that “regions of similar fauna are separated by the whole width of the Atlantic Ocean… [and]… dissimilar faunas lie adjacent to one another”, which he then suggests is due to the opening and closing of the Atlantic Ocean (Wilson, 1966, p.676). He substantiates his theory by breaking down the lifespan of the ocean basin into several stages that occur over approximately 500 million years. The “Wilson Cycle” entails the spreading of the ocean floor, subduction of the ocean crust, contraction of the continents back together, their eventual collision and the formation of fold mountains. Thus, Earth’s continents, though they seem stationary, are actually “moving at around the speed at which our fingernails grow”, embedded in and drifting with mobile tectonic plates (Heron, 2016).

The influence of Plate Tectonic theory cannot be overstated. It gave way to all kinds of exciting new observations. Many mountain ranges were finally understood to have formed due to tectonic compression at convergent boundaries, dispelling Harold Jeffrey’s hypothesis once and for all that mountains were the wrinkles on Earth’s cooling and shrinking surface. Volcanoes, too, were newly appreciated as the result of subduction processes, whereby sinking ocean plates and subducted water reduce the mantle’s melting temperature and produce magma, eventually to erupt out of a volcano and onto the continental crust.

However, despite Wilson’s emphasis on the cyclical nature of the spread and contraction of oceanic basins, scientists largely failed to consider the impact of continental collisions in the long-term. Perhaps they thought that the compressional force at convergent boundaries was so powerful that colliding plate boundaries became totally enmeshed and indistinguishable. Indeed, present day continental interiors resemble a mosaic of pieces that have been stitched together during collision events (Heron, Pysklywec & Stephenson, 2016). These pieces of heterogeneous crust may appear welded together and non-distinct when viewed from Earth’s surface, but underneath the upper crust there can exist deep “scarring left over from ancient collisions of continents” in the lower crust and mantle (Heron, 2016).

In 2016, a team of Earth Scientists published a paper in Nature Communications , claiming that past continental collisions sometimes remain “quasi-plate boundaries” and continue to influence regional Earth surface processes and geologic activities (Heron, Pysklywec & Stephenson, 2016, p.1). They set up a series of thermo-mechanical finite-element experiments to compare strong mantle lithospheres (normal continental crust) with “weak zones” (associated with collisions and rock deformation). They found that “deep lithospheric anomalies can dominate shallow geological features in activating tectonics in plate interiors,” lending support to their titular hypothesis that “lasting mantle scars lead to perennial plate tectonics” (Heron, Pysklywec & Stephenson, 2016, p.1). Accordingly, they speculate as to whether these latent plate boundaries are responsible for intraplate earthquakes that continue to mystify geologists.

Intraplate earthquakes frequently take place at New Madrid Seismic Zone (NMSZ) in the southern and midwestern United States. Between the years 2000 and 2011, over 300 minor earthquakes rumbled the region (Missouri Department of Natural Resources, n.d.). Far from a spreading ridge or subduction zone, this geologic activity might have confused scientists had it not been for the Mantle Scar Hypothesis discussed above. New Madrid overlies an ancient divergent boundary that developed 500-750 million years ago during the splitting apart of supercontinent Rodinia. However, the rift failed to produce a new oceanic plate, instead stretching, thinning, and severely weakening the continental crust above.

Although the Earth Scientists were focused on ancient convergent boundaries, I believe that the weak zones at bygone divergent boundaries undergo similar dynamics. Indeed, intraplate earthquakes occur at both failed rift zones and ancient collision zones. I want to study the relationship between a region’s tectonic history and its present-day intraplate earthquakes. I hypothesise that earthquake depth is related to the crustal thickness of the region such that ancient collision sites have thicker crusts and thus deeper earthquakes. Accordingly, I anticipate shallow earthquakes at failed rift zones where the crust is thinner.



Figure 1: Crustal Thickening and Isostatic Rebound at Convergent Boundaries. This diagram shows that 1) the continental crust is not dense enough to subduct and the mountain is supported by “roots” that project into the mantle, increasing the crust’s overall thickness. 2) The mountain erodes and rocks are transported to surrounding areas, causing the roots to “uplift” in an act of isostatic rebound. Thus, crustal thickness decreases overtime. This diagram is based on a powerpoint published with Oxford University Press, at ​

Figure 2: Crustal Thinning at Divergent Boundaries. 1) This diagram shows that, overtime, upwelling in the asthenosphere leads to stretching and thinning of the overlying continental crust. 2) At successful divergent boundaries, rising magma flows through the lithospheric mantle and produces a new ocean crust, thinning the overlying continental plate until it breaks apart. At failed rifts, the continental crust fails to break apart, but has thinned significantly. This diagram is based on a powerpoint published with Oxford University Press, at ​.

Research Design

I propose an exploratory study that gathers key information about New Madrid and the Pyrenees mountain range. Both regions lie above mantle scars and experience intraplate earthquakes due to weaknesses in the underlying mantle and lower crust. However, the nature of their weak zones is different due to their varying geologic histories. New Madrid overlies a failed divergent boundary while the Pyrenees formed at a convergent boundary during the European Plate-Iberia collision of 55-25 million years ago. The respective underground “landscapes” beneath these regions mean that they respond differently to localised strain patterns. It would therefore be useful to conduct a strength profile analysis on both types of regions – ancient collision zones and failed rifts – to better understand the implications of tectonic history on present-day earthquake activity.

I propose measuring crustal composition, confining pressure, temperature, strain rate, and fluid content. These aspects affect the strength of the lower crust and mantle, in which pseudo subductions and fault reactivation occur. Composition refers to rock-types within the crust, their plasticity and rheology. Confining pressure refers to the compressional force acting on or around fault lines, and increases with depth. According to an article published by University of California, Berkeley, “ confining pressure in the Earth increases approximately by 26 MPa/km in the crust and by 35 MPa/km in the mantle” due to the higher temperature and pressure within the mantle, which is closer to Earth’s core and buried beneath the continental crust (“Strength and Deformation of the Earth’s Lithosphere”, 2017). Temperature varies with crust thickness such that the thinner the crust the higher the temperature (“Strength and Deformation of the Earth’s Lithosphere”, 2017). Th e strain rate refers to how the crust and mantle respond to pressure, and how that response varies with depth, extent of compressive force, and the velocity at which it is applied. The relationship between rock strength and strain rate is nonlinear and “at confinements below 20 MPa, the strength of the material increases faster at the higher strain rate, but at confinements higher than this, the effect… is stronger at the lower strain rate” (Hokka, et al. 2016). Finally, fluid content considers the amount of water within the crust, since water weakens rocks through diffusive processes.

In order to determine the strength profiles of New Madrid and Pyrenees crusts, I suggest deep drilling to discover the water content, strain rate, and distribution of rock-types. Reflection seismology could be used to investigate the structure of the crust and the confining pressure acting upon it. I would then use the well-established geothermal gradient to determine the relative temperatures of the regions.

Defining the strength of the crust at New Madrid and Pyrenees sites would help us to better understand the cause of intraplate earthquakes and how tectonic history affects earthquake depth in complex ways.

Preliminary Results

I examine the magnitude and depth of the last twenty earthquakes to strike New Madrid and the Pyrenees respectively. I anticipated that earthquake depth would be greater at the Pyrenees’s ancient convergent boundary due to its thick crust, than at New Madrid’s failed rift where the crust is thin. By mapping magnitude as a function of depth, I could roughly discern whether the size of an earthquake will be affected by the depth of its focal point, such that greater depth leads to bigger earthquakes.

I found a greater variance of earthquake depths at the Pyrenees, ranging from 2-25 kms. New Madrid earthquakes were more consistent, occurring at depths between 1.6-7.2 kms. The average magnitude of Pyrenees earthquakes was more than tenfold the average strength of New Madrid earthquakes, at 3.65 and 2.405 respectively (accounting for the logarithmic scale of Richter magnitudes). More than half of the Pyrenees earthquakes scaled between 4 and 5 in magnitude, in contrast to the “minor” earthquakes of New Madrid.

Figure 3: ​Pyrenees versus New Madrid. ​This scatterplot compiles data on the magnitude and depth of the twenty most recent earthquakes to have occured in these regions. The graph plots magnitude as a function of depth. The
Pyrenees earthquakes are marked by orange triangles and the New Madrid earthquakes are displayed as green squares. My datasets for New Madrid and Pyrenees earthquakes were found at ​ and ​ respectively.

Figure 4: Table of averages. This table compares the average depth, as well as the average magnitude, of earthquakes occurring in the Pyrenees mountain ranges versus those in New Madrid. The results are based off of the same data as the scatterplot above.

Interpretation of Preliminary Results and Alternative Explanations:

Although my sample size is too small to draw founded conclusions, my preliminary data analysis corroborates my expected results and lends support to my hypothesis. The fact that, on average, earthquakes occur at greater depths and magnitudes at the Pyrenees mountain range than in New Madrid suggests the significance of crustal thickness. On the most basic level, if the crust is thicker, it extends deeper, and there are more opportunities for weak zones to form or an ancient fault line to reactivate at greater depths.

Interestingly, the twenty most recent earthquakes at New Madrid all occurred between April 12-17th, 2018, whereas earthquakes in the Pyrenees happen much less frequently, spanning March 10th, 2017 to December 2nd, 2008. This is likely due to differences in strain rate sensitivity and confining pressure across the regions. It is also possible that temperature plays an important role. In 2008, a group of Spanish scientists found that “ the probability of earthquakes is significantly lower in areas of higher crust temperature” (“Earth’s temperature linked to earthquakes”, 2008). This begs the question: is the relative frequency of earthquakes in these regions due to their difference in crust temperature? While one might suppose a thinner crust is associated with higher temperatures due to its proximity to the mantle, it may be that the thicker the crust, the higher its temperature due to increased pressure. Thus, we would expect New Madrid to be cooler than the Pyrenees, and therefore more prone to earthquakes. Indeed, in 2007, researchers modelling the heat-flow of New Madrid found “no compelling case for assuming that the New Madrid seismic zone is significantly hotter and weaker than its surroundings” (McKenna, Stein & Stein, 2007, p.168). Furthermore, in 1990 the Pyrenean  Variscan crust was described as “under high-temperature metamorphic conditions”, which further corroborates my burgeoning temperature-based hypothesis (Blanquat, Lardeaux, & Brunel, 1990, p.259). However, the Spanish scientists’ claim is contested, and so I pose this explanation only tentatively.

An alternative hypothesis explaining my results might entail rock compositions and how they are distinct across regions. The presence of mineral “talc” would be significant, for example, because of its “very low shear strength in the temperature range 100–400 °C ” (Moore & Rymer, 2007, p.797). Scientists Diane Moore and Michael Rymer explain that “for a given mineral to control the behaviour of the creeping section, it must be very weak as well as characterized by stable shear” and thus Talc-content is a good indicator of a region’s weakness and susceptibility (Moore & Rymer, 2007, p.796). They found an association between “serpentine” and talc such that “talc… forms along the foliation in sheared serpentinite grains… [and] replaces serpentine minerals along the vein walls” (Moore & Rymer, 2007, p.796). Thus, the presence of serpentine would suggest weakness by association.

Another insightful detail to measure is grain-size. Scientists have discovered very fine grained minerals called “mylonite” at many plate boundaries (Wendel, 2015). They argue that the swelling of mineral grains “forces the grains into smaller and smaller sizes” and creates weak zones (Wendel, 2015). They argue that bands of mylonite were responsible for the initial breaking up of Earth’s surface into various plates. While, mylonite has been found at the Pyrenean Variscan crust, it is unclear whether mylonite is present within failed rift crusts (Blanquat, Lardeaux, & Brunel, 1990). Thus, my proposal to drill into New Madrid and the Pyrenees could provide a range of answers to longstanding questions about rock compositions and their embedded minerals. Certainly, if we observed a significant difference between grain-size, mylonite-content, and talc-content across the regions, we could speculate as to how the different tectonic histories of divergent and convergent boundaries have led to distinct rock compositions that in turn affect earthquake depth and magnitude.

Broader Impacts

My proposal would advance the discovery of features at ancient convergent and divergent boundaries most relevant to understanding why intraplate earthquakes occur along their perennial faults. Earthquakes are dangerous, especially intraplate ones, where people are often less suspecting. By studying mantle weaknesses, and how they differ in strength profiles among sites with various tectonic histories, we can better predict which kinds of earthquakes will take place where and prepare local communities accordingly. My preliminary results indicate a relationship between earthquake depth and magnitude. Thus, my proposal has an ethical component – by comparing the crusts of convergent versus failed rift zones scientists could better predict where intraplate earthquakes are likely to cause most damage and perhaps even save lives.