Argentina's Mendoza Region: Earthquake Risk in Wine Country

In the early morning hours of March 20, 1861, an earthquake estimated at approximately M7.0–7.2 destroyed the city of Mendoza so thoroughly that survivors and contemporary observers described finding virtually no building left standing. The death toll reached approximately 6,000–11,000 people — a number that represented a catastrophic fraction of the city's entire population of roughly 12,000. Almost every structure in the urban core collapsed. Fires broke out from overturned cooking fires and spread through the rubble of collapsed adobe buildings. The few survivors who escaped the city described Mendoza as a ruin so complete that rebuilding on the same site was debated for years before the decision was made to reconstruct the city on a new grid oriented to allow light and air into the streets — a layout specifically designed with tree-lined acequia canals, wide boulevards, and building setbacks that would also provide fire breaks and open space for survivors if the earthquake recurred.

The reconstructed Mendoza grew over the following century and a half into Argentina's fourth-largest city — the heart of South America's premier wine region, the gateway to Aconcagua and the highest peaks of the Andes, and an increasingly important hub for tourism, mining, and the trans-Andean economy. Today the province of Mendoza has a population of approximately 2 million people and the city of Mendoza itself around 1.1 million in the greater metropolitan area. It sits, as it did in 1861, at the foot of the Andes in one of the most seismically active regions of South America — directly above the zone where the Nazca plate subducts beneath the South American continent, within 50 kilometers of several active thrust fault systems, and on an alluvial plain underlain by sediments that amplify earthquake shaking and are susceptible to liquefaction. The 1861 disaster is not ancient history. It is a calibration event for a hazard that has not diminished and that the city's modern building stock is, in significant measure, not well equipped to survive.

The Tectonic Setting: Flat-Slab Subduction and Its Consequences

The seismicity of the Mendoza region is shaped by one of the most unusual subduction geometries on Earth. Beneath most of South America, the Nazca plate subducts at a moderate angle — 25–45° — producing the well-developed volcanic arc of the Andes and a relatively conventional distribution of seismicity. But in a zone between approximately 28°S and 33°S latitude — precisely the latitude of the Mendoza-San Juan region — the subduction angle flattens dramatically. The Nazca plate here descends at angles of less than 10° before flattening out nearly horizontally beneath the South American continent, traveling hundreds of kilometers inland at depths of only 90–120 km before eventually resuming its descent into the mantle.

This flat-slab subduction geometry — caused by the subduction of the buoyant Juan Fernández Ridge on the Nazca plate, which resists sinking into the mantle and keeps the slab shallow — has profound consequences for the seismicity and structure of the overlying crust. Where the slab is flat, the mantle wedge between the slab and the overlying continent is essentially absent — there is no hot, fluid-saturated mantle available to generate arc magmas. This is why the Central Andes in the Mendoza-San Juan region lack the active volcanic chain that characterizes the Andes of Colombia, Ecuador, and the southern Andes of Chile. The Aconcagua massif is a long-extinct volcanic edifice; the nearest active volcanoes are hundreds of kilometers to the north and south, outside the flat-slab zone.

The Wadati-Benioff Zone in the Cuyo Region

Despite the flat geometry, the Nazca plate beneath Mendoza does eventually descend into the mantle — and as it does, it generates a distinctive Wadati-Benioff zone of intermediate-depth earthquakes at depths of 80–200 km beneath the Argentine interior. These intermediate-depth events, occurring in the subducting slab well east of the main Andean chain, affect a wide geographic area and generate shaking that is felt across multiple Argentine provinces simultaneously. Some of the most widely felt earthquakes in Argentina's history — events that registered in Buenos Aires and were felt throughout the pampas — have been intermediate-depth events from the Nazca slab beneath Mendoza and San Juan.

The hazard from intermediate-depth events to Mendoza and San Juan is compounded by depth-distance geometry: a deep slab earthquake at 150 km depth can be directly beneath both cities simultaneously, generating ground motions at both locations that approach or exceed the design-level shaking for local building codes. This is qualitatively different from the geometry of shallow crustal earthquakes, where distance from the source provides meaningful attenuation — at 150 km depth, both cities at 50 km horizontal distance from each other experience similar shaking from the same event.

The Argentine Precordillera: Crustal Faults Pressing Against the Andes

In addition to the subduction interface and the Wadati-Benioff seismicity, the Mendoza-San Juan region hosts an active thin-skinned thrust belt in the Precordillera — the low mountain ranges immediately east of the main Andean chain. These thrust faults are driven by the ongoing shortening of the Andean foreland above the flat-subducting Nazca plate, which transmits compressional stress far into the interior of the continent. The Precordillera faults are responsible for some of the most damaging historical earthquakes in Argentina — including the 1944 San Juan event — and represent the most directly hazardous seismic sources for the densely populated Cuyo region.

The Main Thrust Systems

The Precordillera thrust belt consists of a series of east-vergent thrust faults that have progressively built the Precordillera ranges by stacking thrust sheets of Paleozoic sedimentary rocks eastward over the undisturbed foreland. The most active frontal thrust faults — those closest to the cities of Mendoza and San Juan — include the Eastern Precordillera thrusts, the San Ramón fault directly beneath the eastern flank of Mendoza, and the Las Bermudas-La Laja fault system that ruptured in the 1977 Caucete earthquake north of San Juan.

The San Ramón fault deserves particular attention because of its direct exposure to Mendoza's urban population. This east-dipping reverse fault runs along the Piedemonte — the alluvial fan apron at the base of the Andes that forms the eastern foot of the mountains and constitutes the primary residential expansion zone of modern Mendoza. Paleoseismic studies indicate that the San Ramón fault has produced surface-rupturing earthquakes in the Holocene, with the most recent large event estimated at approximately 900–2,000 years before present based on dating of offset alluvial terraces. Its slip rate is estimated at 0.5–2 mm/year, and its maximum magnitude capability is estimated at M6.5–7.0 based on its mapped surface length of 70–100 km and estimated seismogenic depth. Most critically, the fault trace runs directly beneath or adjacent to several high-density residential neighborhoods and new urban developments that have expanded eastward from the old city center over the past three decades — developments that were largely built on alluvial fan sediments without systematic seismic fault assessment.

The 1861 Mendoza Earthquake: Understanding the Event

The 1861 earthquake remains the deadliest natural disaster in Argentine history — a record it holds despite 160 years of subsequent urbanization and industrialization. Its estimated magnitude of M7.0–7.2 places it firmly in the range of damaging crustal earthquakes, though pinning down the exact magnitude requires working backward from intensity observations collected in the weeks and months after the event, in the absence of any instrumental seismographic data (Richter didn't develop the magnitude scale until 73 years later).

The likely source of the 1861 earthquake is debated. Several researchers have argued for the San Ramón fault as the source — its geometry, proximity to the felt intensity maximum, and the pattern of building damage in the city are consistent with a rupture on this structure. Others have pointed to unnamed thrust faults in the Precordillera front west of the city, or to a deep (30–40 km) crustal event on a thrust structure beneath the Piedemonte. The absence of surface rupture clearly attributable to 1861 — either because surface rupture did not occur or because 163 years of subsequent development have obscured the scarp — means the source cannot be definitively assigned, frustrating paleoseismic recurrence estimates that would require a known source fault.

The 1944 San Juan Earthquake: Argentina's Deadliest Modern Disaster

If the 1861 Mendoza earthquake defined the region's seismic history, the January 15, 1944 San Juan earthquake defined its modern seismic consciousness — and had political consequences that extended far beyond the question of earthquake preparedness. The M7.4 earthquake struck at 8:52 PM local time, when most residents were at home in the evening. San Juan at the time had a population of approximately 100,000 people, housed predominantly in adobe and unreinforced masonry buildings. The earthquake killed approximately 10,000 people — 10% of the city's entire population — in what was the deadliest natural disaster in the history of modern Argentina. Approximately 12,000 more were injured, and 75% of the city's buildings were destroyed.

The earthquake's political resonance was immense. The national fundraising campaign for San Juan reconstruction — held in Buenos Aires — was the occasion on which Juan Perón, then an obscure army colonel, first appeared on the national political stage by organizing the relief effort and meeting the actress Eva Duarte, who would become Eva Perón. The earthquake and the subsequent reconstruction of San Juan thus became an unlikely catalyst for one of the most significant political transformations in Argentine history, a connection that Argentine political historians have noted as a striking example of how major disasters can reshape political trajectories.

Why 10,000 People Died in an M7.4

The 1944 San Juan earthquake's death toll was catastrophic relative to its magnitude for reasons that are directly analogous to the 1999 Armenia earthquake in Colombia: the combination of shallow depth (approximately 15 km), near-field epicenter (the city of San Juan was essentially on top of the rupture), vulnerable building stock (predominantly unreinforced adobe and brick masonry), and dense evening occupancy (people at home in buildings built to resist nothing). The pattern of collapses in San Juan — documented photographically in striking detail by contemporary observers — shows the complete pancake collapse of unreinforced masonry construction under lateral ground forces, with survival restricted almost entirely to people who happened to be in or near open areas when the shaking started.

San Juan was rebuilt — again — and this time with a more explicit attempt to apply improved seismic design to the reconstruction. Argentina's first national seismic building code was developed in direct response to the 1944 earthquake, establishing rudimentary requirements for seismic design of new construction in the Cuyo region. Those requirements were progressively strengthened through successive code revisions — the CIRSOC 103 and INPRES-CIRSOC 103 standards — that form the basis of Argentina's current seismic design framework.

The Modern Seismic Hazard: Where the Risk Concentrates

One hundred and sixty-five years after the destruction of Mendoza and 82 years after the destruction of San Juan, the Cuyo region faces a seismic hazard that its historical record describes with unusual precision but whose consequences for the modern built environment remain incompletely characterized and incompletely addressed.

The Alluvial Fan Amplification Problem

Both Mendoza and San Juan are built primarily on alluvial fans — the broad, gently sloping aprons of unconsolidated gravel, sand, and silt deposited by Andean rivers as they emerge from the mountains onto the flatter piedmont zone. These alluvial fan sediments have S-wave velocities of 200–500 m/s in the shallowest layers — significantly higher than the very soft sediments of Mexico City or Bogotá, but low enough relative to bedrock (typically 800–2,000 m/s in the Precordillera rocks at depth) to generate amplification factors of 2–5 at frequencies relevant to most buildings. The specific amplification profile varies across the cities depending on the depth and grain size of the alluvial material at each location, with the finest, most water-saturated deposits in the older downtown areas of both cities tending to produce higher amplification than the coarser gravels of newer peripheral developments closer to the Andean front.

INPRES (Instituto Nacional de Prevención Sísmica), the Argentine national seismic hazard institution based in San Juan, has produced microzonation studies for both cities that characterize the site amplification factors and soil classifications needed for building design. These studies form the basis of the site-specific design requirements in the INPRES-CIRSOC 103 code — the technical foundation for seismic design in Argentina. The challenge, as always, is that code-compliant new construction represents only a fraction of the total building stock, and the legacy of pre-code or inadequately-code-compliant construction is extensive throughout both cities.

Adobe and Unreinforced Masonry: The Persistent Vulnerability

Despite 80 years of post-1944 earthquake awareness in Argentina's Cuyo region and the development of progressively stronger seismic codes, unreinforced adobe and brick masonry remain widespread in the existing building stock of Mendoza and San Juan, particularly in the older neighborhoods of the city centers and in the peripheral lower-income communities. Adobe construction — which was the dominant residential building type in 1861 and 1944 — is still present in significant numbers in rural areas and lower-income urban neighborhoods, and unreinforced brick masonry is ubiquitous throughout the commercial and residential building stock of neighborhoods built between 1920 and 1980.

The INPRES vulnerability assessment for San Juan, conducted after the 2010 Chile M8.8 earthquake provided a regional calibration event, found that a significant fraction of the city's residential building stock remained in the highest vulnerability category — capable of severe damage or collapse at the ground motion levels expected from a moderate local earthquake. For Mendoza, comparable assessments have identified the historic core and the older residential neighborhoods of the eastern urban expansion as having concentrated adobe and unreinforced masonry exposure. The wine industry facilities — the large bodega buildings that process and store the region's production — are a specific category of infrastructure concern, as many are constructed in older masonry styles that reflect the heritage of 19th and early 20th century construction rather than modern seismic design.

The 2010 Chile M8.8: A Regional Calibration Event

On February 27, 2010, the M8.8 Maule earthquake struck the Chilean coast approximately 1,000 km northwest of Mendoza — one of the largest earthquakes in South American history and the sixth largest ever recorded globally. The earthquake was felt with MMI IV–V intensity across most of Argentina west of the Andes, including strong shaking in Mendoza and San Juan that lasted more than two minutes. In Mendoza, some adobe structures and older masonry buildings sustained damage, several roads cracked, and residents experienced the kind of prolonged, low-frequency shaking characteristic of a large, distant megathrust event. The earthquake served as a direct calibration of building performance across the region for shaking well below the design-level earthquake — providing INPRES and Argentine emergency management with an inventory of structures that showed cracks or partial damage at relatively low ground motion levels and therefore would likely fail completely in a near-field design-level event.

The 2010 Maule earthquake also provides context for understanding the Chile-Argentina subduction interface hazard — the threat from a great Chilean megathrust earthquake that, while occurring on the Chilean side of the Andes, transmits long-period seismic energy through the Andean crust to Argentine cities at 200–400 km distance. The 1985 Chile M8.0 and the 2010 M8.8 both caused observable (if not catastrophic) effects in Mendoza and San Juan — and the potential for a larger event on the Atacama section of the subduction interface to produce more severe shaking at these distances is a component of the regional hazard assessment that extends beyond the local fault sources.

The Wine Industry: A Specific Economic Exposure

Mendoza produces approximately 70% of Argentina's wine — roughly 10–12 million hectoliters per year — and the Mendoza and San Juan wine regions together account for nearly all of Argentina's substantial export production. This concentration creates a specific economic vulnerability that goes beyond the general building stock and infrastructure exposure: the wine industry's physical capital — vineyards, irrigation systems, wineries, barrel storage, and bulk storage tanks — is extensively distributed across exactly the terrain most affected by the seismic hazard.

Vineyard infrastructure is relatively resilient to earthquake shaking — the vines themselves survive ground motion that would destroy masonry structures, and surface irrigation systems can be repaired relatively quickly after moderate events. But the processing and storage facilities where grapes are converted to wine — the bodegas — range from small family operations in century-old masonry buildings to large modern facilities that may or may not have been designed to seismic standards. The concrete tanks used for bulk wine storage are particularly vulnerable to cracking and collapse in strong shaking, with the loss of stored wine representing both direct economic loss and environmental contamination of the irrigation water supply if spilled wine enters the acequia system.

The Mendoza acequia system itself — the network of irrigation canals that distributes Andean snowmelt water throughout the urban and agricultural landscape and without which the desert environment of Mendoza could not support either agriculture or a million-person city — is highly vulnerable to earthquake damage. The concrete-lined canals and their distribution infrastructure cross multiple points where ground displacement from fault rupture or liquefaction of the alluvial substrate could disrupt water delivery for the weeks to months required for repair. Disruption of irrigation during the critical February–March harvest season — which coincides with the statistical likelihood of earthquake occurrence at any given time — could devastate a single vintage without causing any other infrastructure damage at all.

INPRES and Argentina's Seismic Monitoring

Argentina's national seismic monitoring and hazard assessment institution — INPRES (Instituto Nacional de Prevención Sísmica), based in San Juan — operates one of the more sophisticated national earthquake monitoring programs in South America. The INPRES national seismograph network covers Argentina with approximately 100 permanent stations, achieving a detection threshold of approximately M2.0–2.5 across the Cuyo region and providing near-real-time earthquake location and magnitude determination for the most seismically active parts of the country.

INPRES has produced Argentina's national seismic hazard maps — the basis for the zonation provisions in the INPRES-CIRSOC 103 building code — and maintains an ongoing research program on paleoseismology, active fault mapping, ground motion prediction, and site response characterization. The institution's San Juan location puts it in the heart of the highest-hazard zone — a practical advantage for fieldwork and local expertise that has made INPRES one of the most technically sophisticated earthquake hazard agencies in Latin America.

The Broader Andean Context: Chile and the Argentine Seismic Belt

Mendoza's seismic hazard cannot be understood in isolation from the broader Chilean-Argentine seismic system that affects the entire southern Andean region. Chile — Argentina's immediate neighbor across the Andes — is one of the most seismically active countries on Earth, hosting the 1960 M9.5 Valdivia earthquake (the largest ever recorded globally), the 2010 M8.8 Maule earthquake, and dozens of M8+ megathrust events in the instrumental record. These Chilean subduction events transmit long-period energy through the Andes to Argentine cities and provide a continuous calibration of Argentine building performance at sub-design-level ground motions.

The portion of the Chile-Argentina subduction interface most directly threatening to Mendoza is the Copiapo-Aconcagua section — the Atacama segment — which extends from approximately 27°S to 33°S, encompassing precisely the flat-slab zone above which Mendoza and San Juan sit. The last major rupture on this segment was the 1922 M8.5 Atacama earthquake — 104 years ago. GPS measurements indicate this segment is currently locked and accumulating elastic strain, with a slip deficit consistent with an M8.5–8.8 event when it eventually ruptures. Such an event would generate significant long-period shaking in Mendoza at a distance of approximately 200–300 km — well within the range to damage vulnerable building stock.

Conclusion: The City That Was Destroyed and Rebuilt Twice

Mendoza has a more intimate relationship with earthquake destruction than almost any city of its size in the world. It was destroyed completely in 1861 and rebuilt from the ruins. San Juan — 170 km to the north and seismically inseparable from Mendoza's hazard context — was destroyed in 1944 with 10,000 dead and rebuilt from its ruins. Both cities carry the physical memory of those events in their street plans, their institutional culture of seismic awareness, and the presence of INPRES as a permanent, well-resourced national hazard institution on their doorstep. In this sense, the Cuyo region of Argentina is better institutionally prepared for earthquake hazard than many regions globally — better than the New Madrid zone, better than the Charleston area, arguably better than the Wasatch Front.

But institutional awareness and technical competence at the regulatory and research level cannot retrofit the tens of thousands of adobe and unreinforced masonry buildings that constitute the most vulnerable portion of the residential building stock in both cities, cannot replace the irrigation headworks that cross the San Ramón fault trace, and cannot prevent the alluvial fan amplification that will increase shaking across the urban zone in the next major earthquake. The history of 1861 and 1944 is clear: when the earthquake comes, the buildings that fail are the unreinforced ones, the water systems that fail are the ones that cross active fault traces, and the people who die are overwhelmingly those who live in the most vulnerable structures in the highest-hazard zones. The seismic science is not the limiting factor. The pace of vulnerability reduction — in a region that knows, better than most, exactly what is at stake — is.