Venezuela's Quake Lifted Trinidad's Coast 20 Feet. AI Could Warn the Caribbean Next Time

On 24 June 2026, twin earthquakes measuring magnitude 7.2 and 7.5 struck Venezuela's Yaracuy state and killed more than 5,100 people. The shaking travelled far enough to lift a stretch of Trinidad's southwestern coastline at Galfa Point, Cedros, by up to 20 feet in a matter of minutes, a shift UWI's Seismic Research Centre has confirmed was directly triggered by the quakes nearly 500 kilometres away. The event is a warning the Caribbean cannot afford to file away, and AI already has the tools to turn that warning into preparation.

What Happened on 24 June, and Why Trinidad's Coast Moved

At 18:04 Venezuelan time, a magnitude 7.2 foreshock struck near Veroes Municipality in Yaracuy state, west of the city of San Felipe. Thirty-nine seconds later, a magnitude 7.5 mainshock ruptured along the same San Sebastian fault system. UWI's Seismic Research Centre later assessed that the two events shared similar focal mechanisms and occurred close enough in time and space to be treated as one massive rupture, with combined shaking that lasted more than 10 minutes.

The damage in Venezuela was severe. Authorities confirmed over 5,100 deaths and 16,740 injuries, with estimates of missing persons ranging as high as 50,000 in the days after the quake. Nearly 800 buildings collapsed nationwide, 189 of them completely levelled, and an estimated 590,000 structures sustained damage. In La Guaira state, the hardest hit, roughly 80 percent of buildings collapsed. The World Bank's early estimate put direct damages at around US$37 billion.

La Guaira sits roughly 300 miles from Trinidad's southwestern peninsula. Even at that distance, the rupture along the San Sebastian fault, which runs close to the Caribbean-South American plate boundary and connects geologically to fault systems that extend under Trinidad, transmitted enough stress to trigger a seabed shift on the Trinidadian side. Geologist Xavier Moonan, who conducted the initial field assessment for UWI, estimated the Galfa Point coastline rose as much as 20 feet almost instantaneously. Hundreds of stranded sea creatures and newly exposed natural oil seeps were reported along the affected shoreline in the days that followed. UWI geologist Kafele Reddock led the follow-up investigation into whether the tremors had triggered a broader landslip along the seabed in the area.

The same earthquake sequence sent tremors as far as Aruba, Bonaire, Curacao, and the Dominican Republic, and triggered a small tsunami of around 4 centimetres recorded at monitoring stations in Puerto Rico. None of that reads like a headline event on its own. Taken together, it describes a single earthquake sequence that touched six or more Caribbean jurisdictions in one afternoon.

The Caribbean Is More Exposed to Earthquakes Than the Postcards Suggest

Hurricanes get the region's disaster planning budget, and understandably so. But the Caribbean plate sits wedged between the North American and South American plates, threaded with active fault systems that have already produced some of the deadliest earthquakes in the hemisphere's history. Haiti's 2010 earthquake, magnitude 7.0 on the Enriquillo-Plantain Garden fault, killed an estimated 220,000 to 300,000 people and remains one of the worst natural disasters ever recorded in the Western Hemisphere. The Puerto Rico Trench, one of the deepest points in the Atlantic Ocean basin, sits along a subduction zone capable of producing both major earthquakes and tsunamis. Jamaica's Enriquillo-Plantain Garden and Wagwater fault systems have produced destructive quakes in 1692 and 1907, the latter levelling much of Kingston.

Trinidad and Tobago sits at the southeastern edge of this plate boundary, close enough to the El Pilar and San Sebastian fault systems in northern Venezuela that seismologists have long treated the two countries as geologically connected. What the June earthquakes demonstrated is not a new risk. It is a risk most Caribbean governments have under-planned for because the last major, close-range event was recent enough to remember and distant enough to feel unlikely to repeat soon.

That is precisely the wrong lesson to draw. A rupture nearly 500 kilometres away moved solid ground on a different island by six metres. A rupture on Trinidad's own fault systems, or a repeat of Haiti's 2010 event elsewhere in the region, would not offer the buffer of distance.

Five Places AI Can Close the Gap

The Caribbean does not need to invent new science to prepare for the next event. The AI tools already exist and are operating in other seismically active regions. What is missing is deployment.

1. AI Earthquake Early Warning

Earthquake early warning systems exploit a simple physical fact: the fast, low-damage P-wave of an earthquake arrives seconds to tens of seconds before the slower, destructive S-wave. Machine learning models trained on dense seismic sensor networks, and increasingly on crowdsourced smartphone accelerometer data through systems like California's MyShake, can detect that P-wave signature and push an alert to phones, schools, and critical infrastructure before the shaking that causes damage arrives. Mexico's SASMEX system has delivered public warnings of 60 seconds or more ahead of major quakes reaching Mexico City.

The Caribbean does not currently operate a public AI-driven early warning network. It does have the foundation to build one. UWI's Seismic Research Centre, headquartered in Trinidad, already operates the region's core seismic monitoring infrastructure across the Eastern Caribbean. Layering an AI detection and alert model onto that existing network, rather than building a new one from scratch, is a matter of funding and integration, not invention.

2. AI-Powered Ground Deformation Monitoring

Satellite radar known as InSAR, Interferometric Synthetic Aperture Radar, can measure ground movement down to a few millimetres by comparing radar images of the same location taken weeks or months apart. AI models trained on InSAR time series can flag unusual strain accumulation along a fault line before it ruptures, effectively mapping which stretches of coastline or which fault segments are storing the most stress. This will not tell a government the exact date of the next earthquake. It will tell a government where to prioritise monitoring, retrofitting, and evacuation planning, information that is not currently built into Caribbean disaster planning in any systematic way.

3. AI Tsunami Detection and Coastal Alerting

The 4-centimetre tsunami recorded in Puerto Rico after the June quakes was minor, but it confirms the detection systems worked and that the pathway from a Venezuelan earthquake to a Caribbean coastline is real and measurable. AI models that fuse real-time ocean buoy data, seismic readings, and coastal bathymetry can generate inundation forecasts for specific bays and beaches within minutes of an offshore quake, rather than issuing a single blanket warning across an entire coastline. For tourism-dependent economies with dense coastal development, from Trinidad's southwest to the north coasts of Jamaica and the Dominican Republic, that specificity is the difference between an orderly evacuation and a false-alarm fatigue that gets ignored the next time.

4. AI Structural Risk Scoring

Most Caribbean building stock, particularly schools, clinics, and older hotel properties, has never had a formal seismic risk assessment. Computer vision models trained on satellite and street-level imagery can perform rapid visual screening at scale, flagging buildings with construction types, ages, or configurations associated with higher collapse risk, the same triage approach engineers use manually in earthquake-prone countries like Japan and Chile but at a fraction of the cost and time. For a region with a shortage of structural engineers, that speed matters as much as the accuracy.

5. AI Damage Assessment and Faster Recovery Financing

After Haiti's 2021 earthquake and Turkey's 2023 earthquakes, AI models comparing before-and-after satellite and drone imagery identified collapsed and structurally compromised buildings within hours, work that previously took manual survey teams days or weeks. That speed feeds directly into recovery. CCRIF SPC, the Caribbean Catastrophe Risk Insurance Facility headquartered in the Cayman Islands, already pays out to member governments automatically once an earthquake or hurricane crosses a pre-set parametric threshold, without waiting for a full damage survey. AI can sharpen the models behind those triggers by combining real-time seismic data, satellite imagery, and historical loss patterns, shortening the gap between a disaster and the moment cash reaches the people who need it.

What Has to Happen in the Next 12 Months

Three actions would move the region from reactive to prepared, and none of them require new technology to be invented.

First, fund the AI layer on top of UWI's existing seismic network so that early warning capability extends to every CARICOM member state, not just the islands closest to the Seismic Research Centre's headquarters. The sensors and the institutional expertise already exist. The gap is the detection model and the alert distribution system.

Second, CDEMA, the Caribbean Disaster Emergency Management Agency, should integrate satellite-based ground deformation monitoring into its regional risk planning alongside its existing hurricane and flood models. Seismic risk has been treated as a secondary concern behind climate hazards for too long, and the June earthquakes are a direct argument against that ordering.

Third, governments should require AI-assisted structural risk scoring for schools, hospitals, and hotels in the highest-risk coastal zones, starting with Trinidad's southwest peninsula, Jamaica's Kingston basin, and the northern coastlines of Haiti and the Dominican Republic. A risk score is not a retrofit, but it is the precondition for knowing which buildings need one first.

The Bigger Picture

A Caribbean coastline moved six metres in minutes because of an earthquake that happened in another country. That fact alone should end any assumption that geographic distance is a form of protection. The region has spent decades building disaster infrastructure around hurricanes, for good reason, given how often they arrive. Earthquakes arrive without a forecast track and without days of warning on the news. What AI offers is not a way to stop the next one. It is a way to know it is coming seconds or minutes earlier, to know which buildings will not survive it before it happens, and to get resources moving within hours instead of days once it does.

UWI's Seismic Research Centre did its job in June. It identified the cause of the Galfa Point uplift and confirmed the link to Venezuela within days. The next step is making sure that kind of analysis happens before an event affects Caribbean soil directly, not after.

Frequently Asked Questions