The Ground Off Mexico Has Not Stopped Moving Since Friday — Here's What Comes Next

[00:00:00]3 days after magnitude 7.3 tore open the Middle America Trench, the aftershocks off southern Mexico are not fading. They are clustering.

[00:00:08]One 9-hour stretch this weekend produced at least eight earthquakes above magnitude 4.5, and the largest of them arrived a full day after the sequence had been assessed as winding down.

[00:00:17]Nearly every new significant earthquake logged anywhere on Earth this week came out of that single box of ocean floor.

[00:00:23]By the end of this, you will know the one measurement that would settle what is happening down there and why nobody has it yet.

[00:00:31]The stretch of seafloor off the coast of Chiapas, southern Mexico, has not gone quiet since Friday.

[00:00:36]3 days after magnitude 7.3 earthquake ruptured the Middle America Trench near the Mexico-Guatemala border, the aftershock sequence is still producing moderate earthquakes at a rate that was supposed to have collapsed by now. It has not collapsed.

[00:00:49]Over the weekend, it did something closer to the opposite. For the first day, the sequence behaved exactly the way the science says it should. The main shock hit Friday. Roughly 30 minutes later, a magnitude 6.0 arrived, which is almost precisely the largest aftershock the statistics predict for a main shock that size.

[00:01:07]A run of magnitude five events followed through Friday night, including a magnitude 5.2 shortly after midnight into Saturday.

[00:01:14]By Saturday's assessment, everyone tracking this was describing it as entirely ordinary. The expectation on the table was one or two more magnitude sixes, maybe a small chance of an upper six, and then a gradual fade into background.

[00:01:26]Then Saturday night happened, and the fade did not come.

[00:01:29]Across roughly 9 hours spanning Saturday evening into the early hours of Sunday, at least eight earthquakes of magnitude 4.5 or greater were cataloged inside a zone about 90 km across. Three of them were magnitude 5.3 or larger. And the single largest event of that entire burst, a magnitude 5.7, arrived a full 24 hours after the sequence had been called normal. That is the story, not a prediction, not an alarm. A sequence that was assessed as decaying, and then measurably did not.

[00:02:01]Here is what the catalogs actually recorded because the specifics matter more than the summary.

[00:02:06]The magnitude 5.7 struck at 17:26 Universal Time on Saturday, 72 km southwest of Puerto Madero, Mexico.

[00:02:14]It carried a green alert level from the United States Geological Survey and no tsunami flag.

[00:02:19]Just over 5 hours later, a magnitude 4.6. 49 minutes after that, a magnitude 4.5 about 39 km west-southwest of Brisas Barra de Suchiate. Then the pace tightened. At 1:47 Sunday morning, a magnitude 5.3, 85 km southwest of Puerto Madero.

[00:02:36]14 minutes later, a magnitude 4.6. 17 minutes after that, another magnitude 5.3, 59 km southwest of Puerto Madero.

[00:02:45]Additional cataloged companions in the same zone include another magnitude 5.3 and a magnitude 4.6 roughly 118 km west- southwest of Puerto Madero.

[00:02:56]Observational material recorded Saturday evening described a magnitude 4.7 earlier that evening and a magnitude 4.4 earlier that morning, as well, on top of everything already listed.

[00:03:06]One observer looking at a three-dimensional globe display described what they were seeing as a pancake of aftershocks, a flattened spreading smear of dots where there should have been a tightening cluster.

[00:03:16]Map that footprint out and it spans roughly 32 km north to south and roughly 83 km east to west. That is a large area for a sequence following a single magnitude 7.3 rupture.

[00:03:27]It is worth pausing on what those magnitude numbers actually represent because the scale is not intuitive. The magnitude scale is logarithmic, so a magnitude 5.7 releases something like 30 times the energy of a magnitude 4.7 and the main shock that started all of this released hundreds of times more than any individual event in the weekend burst.

[00:03:47]None of the weekend's earthquakes were damaging on their own.

[00:03:50]A A 5.3 80 km offshore is a firm shake on shore and very little else.

[00:03:56]What makes them significant is not their individual size, but their number, their timing, and their refusal to taper.

[00:04:02]There is also a caveat on the catalog itself that has to be stated plainly because it produces an apparent contradiction.

[00:04:09]The 7-day survey feed used to compile these events returned only 20 entries against a stated total of 144, which means the magnitude 7.3 main shock is missing from the returned slice.

[00:04:21]It is not missing from the catalog. It is a truncation artifact in the feed. If you pull that data yourself and wonder why the biggest earthquake in the region appears to be a 5.7, that is why.

[00:04:31]All of which raises the first real question of the story, which is not about how many earthquakes there were, but about where exactly they were happening.

[00:04:41]Before the depth question, there is one statistic that puts this weekend in proper scale, and it is the number that made the sequence impossible to file away as routine.

[00:04:50]The 7-day worldwide count of earthquakes at magnitude 4.5 and above rose from 137 to 144 across the same interval as this burst. Seven events globally week over week. That single patch of seafloor off Chiapas accounts for essentially the entire increase. Take a moment with that. On a planet with roughly 40,000 km of active plate boundary, the entire Ring of Fire, the whole Mid-Atlantic Ridge, every subduction zone from Kamchatka to Chile, nearly every new moderate earthquake logged anywhere on Earth this week came out of one 90-km box off the southern coast of Mexico.

[00:05:23]That is not a claim about danger. It is a measure of concentration, and it explains why a sequence that produced no casualties and no significant damage is still worth 20 minutes of anyone's attention.

[00:05:35]Now, to the part where the instruments stop agreeing with each other.

[00:05:39]Take that magnitude 5.7, the largest event of the burst. The United States Geological Survey places it at a depth of 10 km. The European Mediterranean Seismological Centre places the same earthquake at 29 km. That is not a rounding difference. That is a factor of nearly three on the single most consequential parameter in the entire data set for the single most important event in the sequence. And it is not isolated. Look at the depths cataloged across the whole burst and you get a scatter. 10 km, 10, 10, 16, 16, 18, 19.6, 29, 35.

[00:06:16]That is 25 km of vertical spread.

[00:06:18]A clean aftershock sequence releasing stress on one rupture patch does not usually distribute like that. It clusters because the thing generating the earthquakes is a specific surface at a specific depth. So, either the depths are badly constrained or the sequence is not happening on one structure. Both of those possibilities are live and which one is true changes what this sequence means.

[00:06:42]To understand why depth is the whole ballgame here, you have to understand that a subduction zone is not one fault.

[00:06:48]It is at least two entirely different earthquake machines stacked on top of each other at the same spot on the map.

[00:06:54]The Cocos plate is sliding beneath the North American plate off the coast of Chiapas converging at roughly 76 mm per year by the survey's measurement with regional observers putting some segments closer to 88. As it descends, it creates two separate populations of earthquakes.

[00:07:09]The first population lives on the interface, the contact surface where the descending slab is locked against the plate above it.

[00:07:15]These are the megathrust earthquakes.

[00:07:17]They are the ones that reach magnitude eight and magnitude nine. They are the ones that generate large tsunamis. They are, in short, the ones that matter most.

[00:07:26]The second population lives inside the descending slab itself below the interface where the plate bends and pulls apart under its own weight.

[00:07:34]These intra-slab earthquakes can be large and genuinely destructive, but they are a different physical process operating on a different set of faults.

[00:07:42]A sequence clustered at 10 to 16 km is plausibly on the interface. A sequence at 29 to 35 km is more likely inside the slab. A sequence scattered across both, which is what this appears to be, is either poorly measured or genuinely activating more than one structure at once.

[00:07:59]And here is why the distinction is not academic. Interface sequences on a locked mega thrust are the geometry in which foreshock sequences occur.

[00:08:08]Intra-slab sequences generally are not.

[00:08:10]The depth answer would tell you which kind of situation you're looking at.

[00:08:14]Right now, the depth answer does not exist.

[00:08:19]It would be reasonable to ask how two well-funded scientific agencies can disagree this badly about something as basic as how deep an earthquake was.

[00:08:28]The answer is genuinely boring, and it is worth understanding because it dissolves a lot of bad theories.

[00:08:33]Earthquake depth is measured largely by timing the difference between waves that travel directly to a seismometer and waves that bounce off the earth's surface first and arrive slightly later.

[00:08:43]That measurement works well when you have instruments distributed all around the epicenter. It degrades sharply when every seismometer you own sits on one side of it.

[00:08:51]An earthquake 80 km out to sea is exactly that problem. All the stations are on shore in one direction, and the geometry is one-sided, so the solution gets loose. There is a further wrinkle that explains a lot of these lists. When an automated depth solution will not converge on an answer, many catalogs default to a fixed 10 km value.

[00:09:10]Look back at the depth scatter in the sequence and notice how often 10 km appears. Some of those 10s are measurements. Some of them are almost certainly placeholders, the software's way of saying it does not know. On top of that, several events in this burst are still under automatic non-reviewed status. A human seismologist has not yet gone through the waveforms. When those reviewed solutions publish, the depths will move. Some of them may move a lot.

[00:09:34]It is also worth noting that the two agencies are not running the same experiment. They use different velocity models, different assumptions about how fast seismic waves travel through the particular rock beneath the particular piece of coast, and they draw on different sets of stations.

[00:09:48]Feed the same waveforms into two different models, and you get two different answers, and neither agency is hiding that. Both publish their solutions openly, including the disagreements.

[00:09:57]A gap of this size on an offshore event is routine, not remarkable, and it is a measurement limitation rather than evidence that anyone is being misled.

[00:10:08]Underneath all of this sits an uncomfortable fact that no amount of instrumentation has fixed, and it is the reason honest coverage of this sequence has to stay careful.

[00:10:17]Foreshocks are only identifiable in retrospect. Roughly 5 to 10% of moderate earthquakes are followed within days by something larger, which retroactively reclassifies them as foreshocks.

[00:10:28]At the same time, a substantial fraction of large subduction earthquakes are preceded by foreshock activity. Both of those statements are true simultaneously, and together they produce a situation that is operationally useless.

[00:10:40]The overwhelming majority of earthquake swarms are not foreshock sequences, and there is no measurable property that reliably tells you which ones are.

[00:10:47]The clearest illustration is also the most sobering one.

[00:10:50]The magnitude 9.0 Tohoku earthquake that struck Japan in 2011 was preceded 2 days earlier by a magnitude 7.3 foreshock.

[00:10:59]At the time, nobody called it a foreshock. It was treated as a significant earthquake with its own ordinary aftershock sequence, which is exactly how the July 17th event has been treated, and exactly how it should be treated, because in the overwhelming majority of cases, that treatment is correct. That is not a prediction about Mexico. It is the honest statement of what the science can and cannot do.

[00:11:20]Anyone telling you they know which category the sequence falls into is telling you something the instruments cannot support.

[00:11:29]There is, however, one genuinely useful thing to hold on to, and it is the reason this story has a checkable spine instead of just vibes.

[00:11:36]Saturday's assessment did not just say the sequence looked normal. It named a specific falsifiable criterion for when that judgment should change. If multiple magnitude 6.0 or greater events begin appearing in the sequence, that would indicate elevated potential for a further major earthquake. What was described as a major stress reliever, an event that would substantially discharge the strain accumulated along the trench.

[00:11:58]It was also noted, correctly, that the Middle America Trench is capable of producing earthquakes in excess of magnitude 8. So, there is the test, multiple magnitude 6's.

[00:12:08]As of the most recent catalog data, that threshold has not been crossed. The largest event in the entire weekend burst is the magnitude 5.7.

[00:12:16]No magnitude 6 has appeared in it at all. The only magnitude 6 in this whole sequence remains the one that arrived 30 minutes after the main shock on Friday.

[00:12:24]And that one is the textbook aftershock.

[00:12:27]Statistical expectation for a magnitude 7.3 main shock puts the largest aftershock right around magnitude 6.1, which is almost exactly what showed up.

[00:12:36]The reason that threshold carries weight on this particular stretch of coast is that the region has demonstrated what it can do, and recently.

[00:12:43]In September of 2017, a magnitude 8.2 struck the Tehuantepec region, the largest earthquake Mexico had experienced in a century, killing approximately 98 people. That event is instructive in two directions at once.

[00:12:55]It proves the segment is capable of going well beyond magnitude 8. It was also mechanically distinct from Friday's earthquake. It was an intra-slab normal faulting event at roughly 70 km depth, the slab tearing itself under its own weight, rather than a thrust on the locked interface.

[00:13:10]Seven magnitude 7+ earthquakes have occurred within 155 miles of this area since 1950.

[00:13:16]This is not a quiet piece of the planet, and it never has been.

[00:13:19]That is the honest current answer. The sequence is running hot, above expectation on rate, ambiguous on depth, and it is not crossed the one line that would change the assessment.

[00:13:30]And while everyone was watching that line, earthquakes started appearing somewhere else entirely.

[00:13:37]Within hours of the Chiapas burst, moderate earthquakes appeared along and beyond the boundary between the North American and Caribbean plates, a magnitude 5.1 south of Grand Cayman, magnitude 4.7 events off the Pacific coast of Nicaragua.

[00:13:50]That prompted an immediate reading among observers that the Mexican rupture was somehow pressurizing the neighboring plate boundary. The physics on that is worth walking through carefully because part of it holds up and part of it does not.

[00:14:02]When a fault slips, it changes the stress on nearby faults. Regions where stress increases get pushed closer to failure.

[00:14:09]This is called static stress transfer, and it is well established. It is also strongly limited by distance. The effect is meaningful within roughly one to two rupture lengths, and falls away rapidly beyond that.

[00:14:21]For magnitude 7.3 with a rupture on the order of 50 to 80 km, that reaches maybe 1 to 200 km. Grand Cayman is about 2,000 km away. Static transfer cannot explain it. There's a second mechanism that does operate at long range, dynamic triggering, the transient stress carried by the seismic waves themselves as they sweep past. It is real and documented.

[00:14:43]The 1992 Landers earthquake in California set off seismicity across the western United States at distances of hundreds to over 1,000 km.

[00:14:52]But dynamic triggering is weak. It acts almost exclusively on faults already sitting right at failure.

[00:14:58]So, the honest reading is layered. The Nicaragua events, at roughly 700 km on the same subduction system, are plausible candidates for a real if weak relationship. The Cayman event, on a different plate boundary 2,000 km out, is far more likely coincidence in a region that produces earthquakes every week. That said, the instinct to watch that boundary is sound on its own merits. It runs east from Guatemala through the Motagua and Polochic fault systems offshore into the Cayman Trough and along the Swan Islands transform, a left lateral strike-slip system moving roughly 20 mm per year.

[00:15:32]In 1976, it produced a magnitude 7.5 on the Motagua fault that killed approximately 23,000 people. And the Chiapas sequence sits very near the western end of that boundary at the complex junction where the Cocos, North American, and Caribbean plates all meet, which brings us to the question most people watching this actually came here to ask.

[00:15:53]The answer is a qualified no, and the qualification is the part worth staying for.

[00:15:58]The Mexican sequence cannot load California's faults.

[00:16:01]The same arithmetic that ruled out static stress transfer to Grand Cayman rules it out for California with room to spare.

[00:16:07]Static transfer is effectively zero at that range, and dynamic triggering from a magnitude 7.3 at roughly 2 and 1/2 thousand kilometers is far too weak to move anything that was not already on the edge.

[00:16:19]Nothing happening off Chiapas this weekend changed the probability of an earthquake in California. That is not reassurance, it is just physics, and it applies in both directions.

[00:16:28]What is true, independently and with no connection to Mexico whatsoever, is that the southern San Andreas and San Jacinto systems have been measured at their highest stress state in a thousand years.

[00:16:39]The southern San Andreas has not produced a major rupture since the Fort Tejon earthquake of 1857, an event estimated around magnitude 7.9.

[00:16:48]That is 169 years of accumulated strain on a system that does not stay quiet forever.

[00:16:54]And the San Francisco Bay Area has gone conspicuously quiet in a way that seismologists notice.

[00:16:59]Every one of those statements was true last week. Every one of them will be true next week.

[00:17:04]The earthquake in Mexico did not cause any of them. What it did was make a very large number of people go and look. That distinction is worth protecting because the loudest claims circulating right now depend on blurring it. The framing that agencies know something bigger is coming and are withholding it assumes there is a prediction to withhold. There is not.

[00:17:23]No agency on Earth possesses a validated short-term earthquake prediction method.

[00:17:27]Aftershock forecasts are published openly as continuously updated probability tables and every catalog cited in this video is a public feed anyone can pull.

[00:17:36]The stale tsunami record sitting in the data at the sequence coordinates with an empty summary field is an orphaned automated entry from Friday's main shock whose threat was stood down after a half-foot observation.

[00:17:47]It is a database artifact, not a suppressed warning.

[00:17:52]So, where does this actually stand and what would change it?

[00:17:55]Three things are worth tracking over the coming days and all three are publicly checkable. The first is the magnitude six threshold. It has been named out loud, it is specific, and it has not been crossed. If multiple magnitude six events start appearing in the sequence, the assessment changes from a productive aftershock sequence to something wanting far more concern. Right now, the ceiling is 5.7. The second is the revised depths. Several of the weekend's earthquakes are still automatic solutions that no human has reviewed.

[00:18:23]When those reviewed numbers publish, the sequence resolves in one of three directions: toward the interface, toward inside the slab, or toward genuinely multi-structure.

[00:18:32]Each answer means something different and each one narrows a question that is currently wide open.

[00:18:37]The third is whether the eastward activity continues.

[00:18:40]If more moderate earthquakes appear along the North America-Caribbean boundary, the Cayman Trough, the Swan Islands transform, the Motagua system, the pressurization reading gains real weight. If that magnitude 5.1 south of Grand Cayman stands alone, it was almost certainly coincidence.

[00:18:56]And the honest bottom line is this: the most likely explanation for everything in this video is that a productive subduction interface produced a vigorous but ordinary aftershock sequence, that decay is statistical rather than smooth, and that clusters separated by quiet hours are the norm rather than the exception.

[00:19:12]That reading is probably correct, but the largest event arrived a day after the sequence was called normal. The rate is on the high side of expectation, and the depth distribution is consistent with more than one structure being active on a trench that produced a magnitude 8.2 within living memory in 2017, and is capable of exceeding magnitude 8 again.

[00:19:30]Both of those readings are on the table, and the resolution does not exist yet.

[00:19:34]That is not a cliffhanger. That is the actual state of the science, and the instruments that would settle it are going to publish their answers in the next few days.