The 1985 Mexico City Earthquake: A Case Study in Microtremor Site Response

On September 19, 1985, a magnitude 8.1 earthquake occurred off the Michoacán coast of Mexico, within the subduction zone where the Cocos Plate slides beneath the North American Plate. While the epicenter was located over 350 kilometers from Mexico City, the capital experienced catastrophic destruction, leading to thousands of fatalities and the collapse of hundreds of reinforced concrete structures. This event remains a seminal case study in seismology and geotechnical engineering due to the extreme amplification of seismic waves within the Valley of Mexico’s soft lacustrine deposits.

The disproportionate damage observed in Mexico City provided empirical evidence for the significance of site-response analysis. Specifically, the earthquake validated theories regarding how sedimentary basins can trap and amplify surface waves, particularly Rayleigh and Love waves, through resonance. The event also served as a real-world laboratory for the Nakamura technique, or the Horizontal-to-Vertical (H/V) spectral ratio method, which has since become a standard tool for estimating the fundamental frequency of soil deposits and characterizing subsurface lithology.

What happened

• Seismic Event:A subduction-related thrust earthquake (M8.1) followed by a significant aftershock (M7.5) the next day.
• Propagation:Seismic energy traveled 350 km, with high-frequency waves attenuating over distance while long-period waves (approximately 2 seconds) were sustained.
• Site Amplification:The soft clay beds of the former Lake Texcoco amplified ground motions by factors of 10 to 50 compared to nearby rock sites.
• Structural Resonance:Buildings between 6 and 15 stories suffered the most damage because their natural periods matched the 2-second resonance of the lake-bed soil.
• Data Capture:The Guerrero Accelerograph Network (GAN) provided high-quality spectral data that allowed researchers to compare rock-site motion with soft-soil motion.

Background

Mexico City is built upon a complex geological environment known as the Valley of Mexico. Historically, the valley was occupied by a system of lakes, the largest being Lake Texcoco. Over centuries, these lakes were drained, leaving behind thick layers of highly plastic, saturated lacustrine clays. These clays possess unique geotechnical properties, including exceptionally high water content—often exceeding 400%—and very low shear wave velocities, typically ranging from 40 to 90 meters per second.

The stratigraphic profile of the city consists of three main zones: the Foothill Zone (firm soil and volcanic rock), the Transition Zone (interbedded clay and sand), and the Lake Zone (deep deposits of soft clay). Before 1985, the extent to which these clays could amplify distant seismic energy was underestimated. The interaction between the incident seismic waves and the basin’s geometry creates a condition where surface waves are trapped, bouncing between the hard volcanic basement and the surface, leading to prolonged durations of shaking and constructive interference.

The Validation of the Nakamura Technique

The 1985 earthquake and its subsequent analysis brought the Nakamura technique (H/V ratio) to the forefront of seismic microzonation. Developed by Yutaka Nakamura, this method involves taking the ratio of the horizontal Fourier spectrum to the vertical Fourier spectrum of ambient noise or microtremors. The technique is based on the assumption that the vertical component of ground motion is relatively unaffected by the soft surface layers, while the horizontal components are heavily amplified by the resonance of Rayleigh waves within the sedimentary column.

In the aftermath of the 1985 event, researchers applied the H/V ratio to microtremor data across various zones of Mexico City. They found that the peaks in the H/V spectral ratios coincided precisely with the periods of maximum amplification observed during the earthquake. This correlation demonstrated that microtremor measurements could effectively predict the fundamental period of a site without the need for expensive deep boreholes or large-scale controlled source experiments. For the Surface Wave Hub, this validation remains a cornerstone of empirical site characterization, illustrating how microtremors—once considered mere "noise"—contain vital information regarding the elastic moduli and thickness of subsurface stratigraphy.

Sedimentary Basin Effects and Wave Attenuation

The Texcoco lake bed acts as a resonant cavity. When seismic energy enters the basin, the high-impedance contrast between the volcanic bedrock and the soft clay causes total internal reflection. This traps the energy within the upper layers. The 1985 data revealed that while the clay layers are highly compressible, they exhibit remarkably low material damping for small to moderate strains. This lack of attenuation allowed the seismic waves to oscillate for several minutes, a duration much longer than the initial rupture at the fault source.

The spectral analysis of these waves showed a dominant period of approximately 2 seconds in the Lake Zone. This 2-second period proved fatal for mid-rise buildings. In structural engineering, a common rule of thumb is that a building's natural period is approximately 0.1 seconds per story. Thus, 20-story buildings (2.0s) or buildings with slightly reduced stiffness due to poor construction (10-15 stories) were in perfect resonance with the ground, leading to catastrophic structural failure. This resonance effect was absent in the Foothill Zone, where the bedrock's fundamental period is much shorter.

The Guerrero Accelerograph Network (GAN) Findings

The Guerrero Accelerograph Network played a critical role in the scientific understanding of the 1985 disaster. Established shortly before the event as a joint project between the University of California, San Diego, and the National Autonomous University of Mexico (UNAM), the GAN provided a longitudinal record of the seismic energy as it moved from the subduction zone toward the interior.

Spectral analysis of the GAN records revealed that the earthquake’s source spectrum was not unusually rich in 2-second energy. Instead, the data showed that the 2-second peak was a site-specific phenomenon generated by the local geology of Mexico City. By comparing the accelerograms from the station at Tacubaya (firm soil) to those from the SCT (Secretaría de Comunicaciones y Transportes) station in the Lake Zone, engineers quantified the amplification. The SCT station recorded peak ground accelerations (PGA) significantly higher than predicted, with a spectral acceleration that exceeded the design codes of the time by a factor of five. This discrepancy forced a total re-evaluation of urban seismic zoning and building regulations globally.

Inversion Algorithms and Subsurface Imaging

The data from 1985 drove the development of more sophisticated inversion algorithms used to infer material properties from observed wave velocities. By analyzing the dispersion curves—where wave velocity varies with frequency—researchers could map the depth and density of the clay layers with high precision. Inversion techniques allow for the calculation of the shear modulus (G) and density (ρ) from the observed Rayleigh wave phase velocities.

These algorithms have since been refined for use in non-destructive testing (NDT) of infrastructure. By inducing surface waves and measuring their dispersion, engineers can detect voids, buried utilities, or structural weaknesses in foundations. The 1985 case study remains the primary reference for how these inversion models must account for extreme impedance contrasts in heterogeneous media.

What sources disagree on

While the role of site amplification is universally accepted, there remains debate regarding the exact contribution of different wave types to the total damage. Some researchers argue that the primary cause of damage was the resonance of vertically propagating S-waves (shear waves). Others, citing the long duration of the shaking, emphasize the role of laterally propagating surface waves (Rayleigh and Love waves) generated at the edges of the basin. The relative importance of these two mechanisms—1D resonance versus 2D/3D basin effects—remains a subject of complex numerical modeling in modern seismology.

Additionally, there is ongoing discussion regarding the linearity of the soil response. Some studies suggest that the Texcoco clays behaved almost linearly despite the large strains, which is atypical for most soil types. Others argue that non-linear effects, such as soil softening and increased damping, were present but were masked by the overwhelming resonance effect. This distinction is important for modern engineers designing infrastructure to withstand even larger