Hidden Fault Network Reveals How Ecuador and Colombia Share the Strain

Deep in the high Andes of northern Ecuador, where glacial valleys cut across the flanks of active volcanoes and the land quietly shifts a few millimeters each year, scientists have uncovered a hidden network of faults that is rewriting how we understand earthquake hazard along the Ecuador–Colombia border. A new study published in the journal Solid Earth combines satellite radar interferometry, high-resolution terrain models, field mapping, and radiocarbon and cosmogenic dating to show that the northern boundary of the Quito-Latacunga microblock is not a single, clean tectonic line. Instead, it is a broad zone of distributed deformation, up to roughly 70 kilometers wide, in which several parallel right-lateral strike-slip faults share the burden of accommodating tectonic strain. The finding matters because each of these structures appears capable of producing moderate to large earthquakes, potentially up to magnitude 7, in one of the most densely populated parts of the Ecuadorian and Colombian Andes.

The tectonic backdrop is dramatic. Off the Pacific coast, the Nazca Plate slides obliquely beneath the South American Plate, and the collision of the buoyant Carnegie Ridge with the margin pushes a sliver of crust, the Northern Andean Sliver, northeastward at 5.8 to 9.5 millimeters per year relative to stable South America. For decades, geoscientists assumed that most of this sliver motion was absorbed along its great eastern boundary faults, such as the Chingual-Cosanga-Pallatanga-Puná system in Ecuador and the Cayambe-Afiladores-Sibundoy system in Colombia. But more recent GPS-based block models have revealed a more fragmented picture: the sliver itself is broken into several independently moving microblocks, each bounded by faults that must absorb their share of the deformation. One of these, the Quito-Latacunga microblock, sits in the Inter-Andean valley, home to Quito and hundreds of thousands of people living near active faults and volcanoes.

Geodetic models predicted about 3 millimeters per year of right-lateral strain along the microblock’s northern edge, near the border between Ecuador and Colombia. The problem was that no one knew which faults actually carried that motion. GPS stations in the region are spaced tens of kilometers apart, far too sparse to resolve whether the strain is concentrated on one through-going structure or smeared across many. The distinction is critical for hazard assessment. A single, long, mature fault with a high slip rate could produce very large earthquakes at relatively short intervals, whereas a family of smaller, segmented faults would limit rupture size but spread the hazard across a wider area, complicating everything from building codes to emergency planning.

The catalyst for the new work came on 25 July 2022, when a magnitude 5.6 earthquake struck about 10 kilometers northeast of El Angel and roughly 15 kilometers southeast of the Chiles-Cerro Negro volcanic complex. The earthquake was remarkably shallow, with a hypocentral depth of less than 5 kilometers, and its focal mechanism indicated almost purely strike-slip motion. Using interferometric synthetic aperture radar, or InSAR, from both the European Sentinel-1 C-band satellites and the Japanese ALOS-2 L-band sensor, the research team mapped the coseismic deformation in unprecedented detail. The interferograms revealed a surface rupture oriented roughly east-northeast, with line-of-sight displacements of 5 to 13 centimeters. In the eastern section, a discrete rupture trace extended over 5 kilometers, while to the west the deformation splintered into en-echelon strands and diffuse off-fault deformation, a signature of a fault zone that is young, immature, and geometrically complex.

Field reconnaissance three months after the earthquake confirmed the InSAR mapping in a limited way: subvertical cracks a few centimeters wide were found on the walls of an irrigation channel precisely where the radar data placed the rupture. Dense vegetation in the El Angel ecological reserve and landslide-blocked roads made the terrain difficult to survey, and the thick, organic-rich Andisol soils, which can deform ductilely, likely absorbed much of the shallow slip. But the earthquake had done its job as a scientific beacon, pointing the team toward a previously underappreciated fault, now named the 25 July fault, that cuts across glacial moraines near the microblock’s northern edge.

To place this rupture in context, the researchers turned to remote sensing at multiple scales. They used 4- and 5-meter digital terrain models from lidar and GeoSAR surveys, satellite imagery, and, crucially, new high-resolution terrain models built from Pleiades tri-stereo satellite images processed with the NASA Ames Stereo Pipeline. These data revealed a swarm of southwest-to-northeast-trending lineaments within a zone roughly 70 kilometers wide, oblique to the better-known north-trending thrust faults of the region. Two structures stood out: the Reservoir fault and the Polylepis fault, both north of and subparallel to the 2022 rupture. Along the Reservoir fault, the Pleiades models showed fault scarps up to 3 meters high, sag ponds where streams had been dammed, landslides, and, most tellingly, glacial lateral moraines and stream channels offset right-laterally by measured amounts ranging from about 5 to 37 meters.

Dating those offset landforms required two independent geochronological tools. The team sampled large, fresh andesitic and basaltic boulders embedded in the moraine crests and measured cosmogenic helium-3 in pyroxene grains, which records how long the boulders have been exposed to cosmic rays since glaciers deposited them. The exposure ages clustered between roughly 11.9 and 20.4 thousand years, consistent with deglaciation at the end of the last glacial maximum in the tropical Andes. Dividing the measured offsets by these ages yields slip rates on the Reservoir fault between about 0.8 and 6.1 millimeters per year, depending on which offset and age combinations are used. Notably, the higher end of this geologic range exceeds the 3 millimeters per year predicted by the geodetic block model, an intriguing discrepancy that the authors attribute, at least in part, to the influence of the nearby volcano.

Roadcut exposures along the Reservoir fault provided a paleoseismic record. A vertical fault zone cuts through post-glacial sediments and volcanic soils, with three stacked colluvial wedges marking three separate surface-rupturing earthquakes. Radiocarbon dating of organic-rich Páramo soils, analyzed with Bayesian modeling in OxCal, brackets the events at about 7,030, 4,890, and 4,320 years before present, implying an average recurrence interval of roughly 2,300 years for large ruptures. Scaling relations linking rupture length and displacement to magnitude suggest these prehistoric earthquakes were between magnitude 6.3 and 7.0. Significantly, the Reservoir, Polylepis, and 25 July faults all fall within the proposed epicentral area of the destructive 15 August 1868 earthquake, estimated at magnitude 6.4 to 6.8, which damaged the town of El Angel. The fresh-looking surface expression and recent landslides along the Polylepis fault lead the team to speculate that it may have hosted that event, though dedicated paleoseismic trenching will be needed to confirm it.

Perhaps the most provocative element of the study is the proposed link between faulting and volcanism. The Chiles-Cerro Negro volcanic complex has been inflating since 2014, with InSAR documenting uplift rates of up to 2.9 centimeters per year and ground spreading away from a north-south axis through the volcano. The researchers argue that this inflation could mechanically load the adjacent faults, inducing right-lateral shear on the 25 July fault and potentially triggering earthquakes through elevated pore fluid pressures, a mechanism previously invoked for a 2014 magnitude 5.6 event south of the volcano. If volcanic unrest periodically accelerates slip, the geologic slip rates over the Holocene could legitimately exceed the decadal geodetic averages, resolving the apparent mismatch between the two datasets. The same logic may explain why ancient terrane-bounding suture zones near Chiles-Cerro Negro are not being reactivated, while in southern Colombia, near the Galeras volcano, the Romeral shear zone is being reactivated as active right-lateral reverse faults such as the Buesaco and Aranda faults, which slip at 0.7 to 2.6 millimeters per year.

Taken together, the results portray the northern boundary of the Quito-Latacunga microblock as a crustal-scale distributed shear zone rather than a discrete plate-boundary-style fault. No sharp velocity gradient appears in the InSAR data, seismicity is diffuse rather than aligned, and the mapped faults are discontinuous, segmented, and spread over tens of kilometers. For the communities of El Angel, Tulcán, Ibarra, Pasto, and Túquerres, the message is sobering but actionable: the hazard is real, distributed, and comes from multiple sources, each capable of magnitude 7 events, and possibly modulated by the breathing of the volcanoes above them. The study’s open dataset, including the Pleiades terrain models, offset measurements, and dating results, now provides a foundation for the denser GPS networks, systematic paleoseismic trenching, and refined hazard models that this tectonically restless corner of the northern Andes urgently needs.

Subject of Research: Distributed right-lateral faulting and earthquake hazard at the northern boundary of the Quito-Latacunga microblock in Ecuador and Colombia

Article Title: Distributed right-lateral strain at the northern boundary of the Quito-Latacunga microblock

Article References: Harrichhausen, N., Marconato, L., Audin, L., Lacan, P., Baize, S., Jomard, H., Alvarado, A., Hollingsworth, J., Blard, P.-H., Mothes, P. A., Rolandone, F., & Ortiz Martin, I. D. (2026). Distributed right-lateral strain at the northern boundary of the Quito-Latacunga microblock. Solid Earth, 17(5), 763-787. https://doi.org/10.5194/se-17-763-2026

Image Credits: AI Generated

DOI: 10.5194/se-17-763-2026

Keywords: Ecuador, Colombia, Quito-Latacunga microblock, strike-slip faulting, InSAR, paleoseismology, cosmogenic dating, Chiles-Cerro Negro, Northern Andean Sliver, seismic hazard, tectonic blocks, volcano-tectonic interaction