What’s going on beneath Mexico?

Something weird is happening in Mexico these days. Two powerful earthquakes have hit the country in two weeks. The second one occurred exactly on the same date of the thirty-second anniversary of the most damaging quake known in the country and the early warning system did not work properly. Even more: two hurricanes are causing devastation through the Gulf of Mexico. Yes, definitely something really weird is going on in Mexico these days… or maybe not.

Earthquakes are no surprise in Mexico. As part of the infamous Pacific ring of fire, the country suffers often from earthquakes and volcanic eruptions. Both are natural manifestations of the motion of two oceanic plates -named Rivera and Cocos- under the North American and Caribbean continental plates, a process called subduction (Fig. 1). As an oceanic plate dives, friction with the overriding plate eventually stops its motion. The interface between both plates is said to be “coupled and locked”, so the oceanic plate cannot continue descending. However, gravity still pulls the slab of oceanic plate below the boundary, increasingly stretching the plate. This situation continues till a breaking point, when a fracture relieves part of the stress accumulated in the descending plate. This type of earthquake is called intraslab. Both the magnitude 8.1 September 8th event under the Gulf of Tehuantepec and the magnitude 7.1 September 19th one beneath Puebla are of this kind and ruptured within the Cocos plate.

Figure 1. simplified sketch showing the subduction of the Cocos oceanic plate under the North America continental plate in Central Mexico. Thick orange line: plate interface. Orange circles at the interface: shallow interface earthquakes. Red explosions: deep intraslab earthquakes. Modified from Kostoglodov and Pacheco (1999).

Does it mean the second quake was an aftershock of the first one? Not really. Aftershocks reflect somehow the adjustment of the Earth’s interior in the vicinity of the area fractured by a larger shock. In this case the event in Puebla took place more than 600 km away from the Tehuantepec rupture. A large distance, too far to consider the former event as an aftershock of the latter one. This does not mean that both earthquakes could not be related. Both events have taken place in the Cocos plate, at similar depths (40 to 80 km) and relatively close, plus the first event was strong enough to induce changes in the forces acting within that plate. There is ample evidence worldwide of seismicity triggered by strong earthquakes at even further distances, so future studies will say.

What about the coincidence of dates (September 19th) of the 2017 Puebla earthquake and the catastrophic Michoacán shock of 1985? Well, that is just… a sad coincidence. Or not that sad. The 1985 earthquake is the more damaging event in Mexican history, with a death toll ranging from 9500 to 40000 depending on the source. Most of them died in Mexico City, 450 km away from the coastal epicentre. As such, the Michoacán quake was a watershed in Mexico’s seismic preparedness. It spurred the update of Mexico City’s building code, the birth of the national Civil Protection and the development of an earthquake early warning (EEW) that was the first public system in operation worldwide. It also established the tradition of performing a national seismic drill every 19th of September to remind Mexicans about the impending threat of earthquakes. The same drill that has taken place since 1986. The same drill that ended a couple of hours before the waves of the 2017 quake hit the ground. The same drill that has probably saved thousands of lives thanks to people’s training. So perhaps such a coincidence has not been that bad in the end.

The 1985 earthquake was also a watershed for seismology. At that time seismologists did not know if that portion of the Mexican coast was capable of causing major quakes. When the boundary between the oceanic and the continental plates is locked, stresses build up within the descending plate and along the interface. Once a critical point is reached, the oceanic plate slips beneath the upper plate producing a powerful earthquake: an interplate event (Fig. 1).

Historical records did not place any significant event along the Michoacán shore, depicting a seismic gap: an area with a lack of seismicity at the interface. So experts thought the Michoacán gap might be aseismic, unable to produce large quakes. The 1985 magnitude 8.0 earthquake revealed it wasn’t. It also showed how interplate events generate strong seismic wave trains that travel inland close to the surface and get trapped in the Valley of Mexico, where they are dramatically amplified. This effect is caused by two factors. First, interplate ruptures are long, shallow, and smooth due to the cyclic slip of the plates along the same interface. These long ruptures efficiently generate long-period surface waves. Second, the Valley of Mexico is an old lake basin, mostly dried up after the Spanish arrival. The sediments on top of this basin are water-saturated clays that act like jelly under the shaking of seismic waves, especially amplifying those with periods between 0.5 and 1 s. These periods match to the resonance periods of 5- to 10-story buildings, which were the most damaged in 1985.

In contrast, the recent earthquakes in Tehuantepec and Puebla are totally different beasts. They occur deep within the descending plate, where rocks are tougher than at the plate boundary. These quakes thus need more energy to break, and their ruptures are shorter and more powerful. So intraslab quakes tend to be smaller than interplate ones, but proportionally release more energy. This energy is mainly carried by fast seismic waves called P(-rima) and S(-econda), since they arrive in the first and second place. These waves shake the ground at shorter periods than surface waves, usually damaging lower buildings wherever they hit. If we add that the Puebla earthquake happened only at 120 km from Mexico City, we will understand why this magnitude 7.1 event has caused such level of damage despite being far smaller than the 1985 earthquake. Additionally, deep events pose a tricky challenge for any warning system. Since they happen below the continent, the first and more energetic waves travel directly upwards towards inhabited places. In other words, there is no warning signal carried by weak waves before the strong waves arrive. This is why the Mexican EEW system issued an alert only a few seconds in advance during the Puebla earthquake.

Determining the true size of the disaster will take time. Unfortunately the number of fatalities and disappeared people is still on the rise. However, we do know why the two recent deep earthquakes happened and why they caused the reported devastation. What we do not know yet, and might be more concerning, is what will happen next. The common seismic cycle in subduction zones is that intraslab earthquakes precede large interplate events. Both the Tehuantepec and the Puebla earthquakes have happened downwards of known seismic gaps: the Tehuantepec and Guerrero gaps (Fig. 2), which have not seen major interplate events in more than a century.

Figure 2. Large historical earthquakes in Central Mexico and their dates. Red explosions: deep intraslab earthquakes. Coloured areas: shallow interplate earthquakes. Blue asterisks: other relevant events. Green stars: 2017 Tehuantepec and Puebla intraslab earthquakes. Note the prominent seismic gaps along the coast in Guerrero and Tehuantepec. Modified from Kostoglodov and Pacheco (1999).

Does it mean that large events along the coast are coming in the near future? If so, when and how? We do not have answers yet to these questions, but we know that the best recipe is making our societies and buildings more resistant and resilient to these (not that weird) natural hazards. Exactly as we do for hurricanes that, by the way, happen every year during the hurricane season.


Campillo, M., J.C. Gariel, K. Aki, and F.J. Sánchez-Sesma (1989). Destructive Strong Ground Motion in Mexico City: Source, Path, and Site Effects during Great 1985 Michoacan Earthquake. Bull. Seism. Soc. Am. 79(6), 1718-1735.

Choy, G.L., and S.H. Kirby (2004). Apparent Stress, Fault Maturity and Seismic Hazard for Normal-Fault Earthquakes at Subduction Zones. Geophys. J. Int. 159(3) 991–1012. doi:10.1111/j.1365-246X.2004.02449.x

García, D., S.K. Singh, M. Herraiz, M. Ordaz, and J.F. Pacheco (2005). Inslab Earthquakes of Central Mexico: Peak Ground-Motion Parameters and Response Spectra. Bull. Seism. Soc. Am. 95(6), 2272–2282. doi:10.1785/0120050072

Kanamori, H. (1986). Rupture Process of Subduction-Zone Earthquakes. Ann. Review of Earth & Planetary Sc. 14, 293-322. doi:10.1146/annurev.ea.14.050186.001453

Kostoglodov and Pacheco (1999). Cien Años de Sismicidad en México. Catálogo de Sismos Moderados y Grandes Ocurridos en México durante el Siglo XX. Tech. Report (in Spanish), Institute of Geophysics, UNAM, México D.F.

Lay, T., L. Astiz, H. Kanamori, and D.H. Christensen (1989). Temporal Variation of Large Intraplate Earthquakes in Coupled Subduction Zones. Phys. Earth & Pl. Interiors 54(3-4), 258-312. doi:10.1016/0031-9201(89)90247-1

Rosenblueth, E., M. Ordaz, F.J. Sánchez-Sesma, and S.K. Singh (1989). The Mexico Earthquake of September 19, 1985 – Design Spectra for Mexico’s Federal District. Earthq. Spectra 5(1), 273-291. doi:10.1193/1.1585523.

Sánchez‐Sesma, F.J., S. Chávez‐Pérez, M. Suárez, M.A. Bravo, and L.E. Pérez‐Rocha (1988). The Mexico Earthquake of September 19, 1985 – On the Seismic Response of the Valley of Mexico. Earthq. Spectra 4(3), 569-589.

Singh, S.K., L. Astiz, and J. Havskov (1983). Seismic Gaps and Recurrence Periods of Large Earthquakes along the Mexican Subduction Zone: A Reexamination. Bull. Seism. Soc. Am. 71(3), 827–843.

Written by

Leave a Reply

Your email address will not be published.Required fields are marked *