Reflecting on the September 2017 Mexico Earthquakes

Mike Floyd for EAPS News
September 25, 2017

Michael Floyd, a Research Scientist at EAPS in the Geodesy and Geodynamics Group, offers some thoughts after the two recent earthquakes affecting southern Mexico

In the last two weeks, there have been two unusual and destructive earthquakes that have rocked Mexico. The first earthquake occurred on September 8th and the second occurred on September 19th, cumulatively resulting in several hundred deaths to date. For earthquakes to occur along the west coast of Mexico is not unusual but both these earthquakes were not as simple as they first seem and their occurrence emphasizes an often underappreciated hazard. The earthquakes occurred near a subduction zone off of the southwest coast of Mexico but their depth and style of faulting is uncommon for such a tectonic setting. So what was unusual about these earthquakes and what can we learn from them?

In the last decade or so, we have had our attention drawn repeatedly to classic subduction earthquakes. Subduction zones are known for producing the very largest earthquakes, often termed “megathrusts” due to their size and style of fault motion, where the overlying plate rides up over the subducting plate. Not only can such megathrusts produce earthquakes of magnitude 9 and above, but the thrusting motion lifts the sea floor and the water above it, which spreads out from the source and produces the ultimate destructive force, a tsunami. This type of plate boundary exists in two places within the United States: along the arc of Aleutian Islands in Alaska and off the coast of the Pacific Northwest. The west and south coasts of the major Indonesian islands of Sumatra and Java, west coast of South America, and the east coast of Japan are other well-known examples of subduction zones that have produced megathrust earthquakes. Respectively, the December 2004 Sumatra, February 2010 Chile, and March 2011 Japan earthquakes were all megathrusts that had accompanying local and ocean-crossing tsunami, spectacularly recorded by modern media. Over the same recent period, we have seen the 50th anniversaries of two infamous megathrust events: the May 1960 Chile earthquake, the largest in recorded history; and the March 1964 Alaska earthquake.

Subduction zones strike fear into most populations that live among them and occupy a large proportion of active research focus and interest. Along much of Mexico’s southwest coast there is also a subduction zone, in this case where the Cocos tectonic plate, lying beneath the Pacific Ocean, converges with and descends below the continental plate of North America.

In the case of the latest pair of earthquakes in Mexico, however, they were neither thrusts nor shallow. They occurred at depths likely below the contact between the oceanic and continental plates, at about 70 km (45 miles) and 50 km (30 miles), respectively. Furthermore, they were normal faulting earthquakes, not thrusts. Normal faults are the opposite of thrusts, where the two sides of a fault slip to produce extension. So how do we get extensional earthquakes next to a convergent plate boundary? It’s all about where they occurred, within the subducting plate itself as it descends, bends and breaks apart in the mantle. Such intra-slab extensional earthquakes are rare, and normal faulting earthquakes in general rarely reach such large magnitudes. Other destructive examples are the June 1977 Tonga and February 2001 Nisqually earthquakes, the latter of which occurred directly below western Washington and severely damaged the Seattle-Tacoma-Olympia urban corridor.

These deep earthquakes compel us to delve deeper, beyond our immediate reactions, to gain a full understanding of their significance. On the one hand, one of the reasons for such earthquake’s large size is likely their depth—the higher pressures enable larger stresses to build up before being released suddenly—but this depth also distances our populations on the surface from some of the power of the seismic waves. Conversely, whereas megathrust earthquakes generally occur tens to hundreds of kilometers offshore where the oceanic and continental plates meet, the intra-slab events take place directly underneath the land. The earthquake may be deep but we might live right at the epicenter, the seismic waves’ bullseye at the surface. Therefore there are a number of competing factors when it comes to the effects of such earthquakes, some of which lessen their impacts but others of which compound them compared to their more infamous and spectacular cousins, the megathrusts and tsunamis. In the case of Mexico City specifically, the local geology of the underlying ground also exacerbates shaking.

This opens a whole new realm of questions regarding the frequency of such earthquakes and the hazard they pose, which we may become distracted from appreciating while focusing our attention on the upper “megathrust” part of the subduction zone. From a standpoint of resources and research success, it is understandable to focus on areas that are probably going to have the greatest impact on our societies. However, this is of little comfort to those individuals who have been affected by these latest earthquakes outside others’ region of interest. This proves how valuable science is and will be even if the immediate reward is not obvious. It is all very well channeling resources into the most potentially catastrophic threats but there is always a need for expertise in what may be perceived as less important or unrelated regions or topics.

From a scientific perspective, the major problem with these recent types of earthquakes is that they are exceptionally difficult to study. They are deep down in the Earth, not at the surface where we can set up instruments to measure the gradual tectonic motions between earthquakes as well as during them. They do not produce any observable signs except for the shaking at the surface at the time they occur. Even this limited knowledge can be put to good use, however. Mexico has had an implementation of an earthquake early warning system for years, developed after the 1985 earthquake that devastated Mexico City. The system is widely reported to have detected the impending shaking near the epicenter and provided invaluable seconds to take appropriate actions in areas further afield. It doubtlessly saved lives. The United States is now in the latter stages of implementing its own earthquake early warning system along much of the west coast, provided that funding remains invested towards its practical completion.

Ultimately, as always, the best protection is education, knowledge and appreciation of the hazard, and preparation for its potential occurrence. An earthquake early warning system is useless unless those whom it is intended to warn understand with mortal seriousness what it is, what it means and what to do. This is a fundamental intention of the “Great ShakeOut” earthquake drill [], an annual exercise which is due to take place again next month, at 10:19 local time on 10/19. The Great ShakeOut started as an education and outreach initiative of the Southern California Earthquake Center (SCEC) []. It was so successful that it expanded to the whole of California and eventually beyond. Now 46 US states officially take part, as well as five US overseas territories, three Canadian provinces, Japan, New Zealand and a few other regions around the world. However, while this drill originated in the United States, it is important to remember that each country, each region and even each city may have different local amplification effects, current building codes, history of retrofitting and advice on best first response during an event, which may necessitate different awareness and practices accordingly. Whether or not one lives in a known earthquake zone, travel for business and vacation means that we may find ourselves in such a situation that requires a personal response to stay safe.

The occurrence of these two unusual and, dare I say, unexpected earthquakes just one month before 2017’s Great Shakeout is the perfect opportunity to reinforce this attitude toward regional hazard preparedness and mitigation, to both the general populace and local, regional and national governments. As scientists, we have to be honest about the limitations of our current knowledge but offer educated, useful and actionable solutions nevertheless.

Story image credit: Photographs by the Presidency of Mexico CC-BY-2.0 via Wikimedia


Michael Floyd is a Research Scientist who works with Tom Herring, Bob King and Rob Reilinger in the Geodesy and Geodynamics Group. He gained his undergraduate degree and doctorate in the Department of Earth Sciences at the University of Oxford. He was then a postdoc in the Department of Earth Sciences at the University of California, Riverside, for two years before joining EAPS in April 2011. His research interests concentrate on using geodetic observations, mainly Global Navigation Satellite Systems (GNSS), including GPS, to study solid Earth phenomena, specifically fault motions throughout the earthquake cycle, continental deformation, and geothermal field monitoring. He has worked in many areas of the world towards this purpose, including the Hellenic subduction zone and overlying Aegean continental crust, Anatolia, the Caucasus, and the broad San Andreas Fault system in the North San Francisco Bay Area, including The Geysers geothermal field. He also currently serves on UNAVCO’s Geodetic Data Services Advisory Committee.

Recent News

What Happened After the 2014 Napa Quake? MIT News

Mind the Gap MIT News

Seismic Strain in Azerbaijan MIT News