
The morning of Wednesday, January 31, 1906 was calm along the coast of Ecuador and southern Colombia - palm-lined shoreline, fishing boats, the slow coastal life of the Pacific tropics. At 10:36 local time, the sea floor off Esmeraldas gave way. A section of the North Andes plate boundary 500 to 600 kilometres long slipped in a single great rupture, releasing more seismic energy than all the major earthquakes of the following seventy years on that same coastline combined. The moment magnitude has been calculated at 8.8. It is one of the largest earthquakes ever measured instrumentally, and the tsunami it launched would still be recorded in Japan sixteen hours later.
This coastline is one of the most seismically active on Earth for a specific geological reason. Offshore, the Pacific's Nazca plate - together with smaller fragments now known as the Malpelo and Coiba plates - slides eastward beneath the North Andes plate and the South American continent. The boundary is a megathrust, the same class of fault that produced the 1960 Valdivia and 2011 Tōhoku earthquakes. Stress builds along it for decades or centuries as the plates lock, then releases catastrophically when the lock fails. In 1906 the release was enormous. Geologists today consider the 1906 event the benchmark rupture for the Ecuador-Colombia segment - the single large event against which all subsequent smaller ones are measured.
What killed people was not the shaking. It was the water. Along the coastal strip from Río Verde in Ecuador to Micay in Colombia, the tsunami came ashore with little warning. Fishing villages built on low ground at river mouths were particularly exposed: the wave found its way up estuaries and flooded communities that had no concept of what a tsunami was or how to respond. Estimates of the death toll range from about 500 to as many as 1,500, almost all from drowning and the collapse of flooded structures. These were overwhelmingly fishermen, farming families, and townspeople of small coastal communities, most of whose names appear in no official record. The uncertainty in the count is itself a kind of evidence - evidence of how thinly these places were administered at the turn of the twentieth century, and how much of the damage fell on people whose lives were never fully documented to begin with.
The 1906 tsunami did not stop at the Ecuadorian coast. Seismographs and tide gauges around the Pacific recorded the event. Wave heights were reported in Hawaii, Japan, and California, with runups modest enough in distant places to cause only minor local flooding but scientifically extraordinary for what they demonstrated: a single rupture off one stretch of South American coast had just sent energy rippling across a basin spanning a third of the planet. For tsunami scientists, the 1906 Ecuador-Colombia event became one of the first well-documented trans-Pacific tsunamis of the instrumental era. The waveforms recorded on those old drum gauges, later digitized, are still used to reconstruct the rupture geometry - showing a long, shallow fault slip concentrated off Esmeraldas and extending several hundred kilometres north into Colombian waters.
The 1906 break did not release all the stress on that boundary; it just set the clock running again. Over the next century, the same segment ruptured in pieces. A magnitude 7.8 earthquake in 1942 broke part of it off Manabí Province. A 7.6 in 1958 broke another part near Esmeraldas. A 7.7 off Tumaco in 1979 took a third section. Together, those three later quakes re-ruptured roughly the same reach of fault that had moved in 1906 - but released only a fraction of the energy. This pattern taught seismologists something important: large megathrust segments have structural barriers between sub-segments, but the barriers are not permanent. A very large earthquake can jump them. A smaller one stops at them. On April 16, 2016, a magnitude 7.8 earthquake struck within the same 1906 zone near Pedernales, killing at least 676 people. It was essentially a re-rupture of the 1942 section, roughly on the 74-year recurrence interval seismologists had projected.
From our side of the twentieth century, the 1906 earthquake looks like a bookend - the great event that opened a century of smaller but still devastating quakes along a single plate boundary. For the residents of coastal Ecuador and Colombia in early 1906, it was a disaster, full stop. A Wednesday morning that became a flood. A fishing village that was there at dawn and was not there by evening. The scientific understanding that has since built up around their deaths - rupture geometry, tsunami propagation models, recurrence intervals - is partly a gift those people unknowingly gave the rest of us. The coast today is dotted with tsunami evacuation signs, and the schoolchildren of Esmeraldas, Tumaco, and Pedernales drill for what to do when the ground starts shaking. That preparedness has its roots, eventually, in 1906.
The 1906 epicenter lies offshore at approximately 1.77°N, 80.00°W, off the coast of Ecuador's Manabí and Esmeraldas provinces. The nearest major airport is Manta's Eloy Alfaro International (SEMT / MEC); Quito's Mariscal Sucre (SEQM / UIO) lies about 170 km east-northeast, over the Andes. The plate boundary runs along the seafloor from south of Guayaquil northward to southern Colombia. Tropical maritime weather prevails; rainy season runs roughly January to April. The coastal topography rising to the Andes is strikingly visible at cruising altitude, with the high volcanoes of the Ecuadorian Avenue of the Volcanoes inland.