A large earthquake is expected to strike Jalisco, Mexico, in the next few years as pressure builds to the breaking point along a major fault in the region.
by C. DeMets, I. Carmichael, T. Melbourne, O. Sanchez, J. Stock, G. Suarez, and K. Hudnut
Note in proof: On October 9, 1995, a magnitude 7.6 earthquake occurred beneath the Jalisco region and caused significant loss of life and property. This earthquake highlights the need for accurate measurements of crustal strain rates in earthquake-prone zones. In the coming months, we plan to measure the amount of displacement that occurred within the Global Positioning System (GPS) network during and after this earthquake.
On June 3, 1932, the largest earthquake recorded in Mexico this century rocked the state of Jalisco, causing widespread casualties and damage. This magnitude 8.2 earthquake and a magnitude 7.8 aftershock ruptured the Rivera subduction zone (Figure 1), a major fault that allows the Rivera plate to descend into the mantle beneath the Jalisco region. There has been little seismic activity at the Rivera subduction zone since 1932, which suggests that the subduction fault is either locked and accumulating strain, or slipping freely. If the subduction fault is locked, a likely possibility, as much strain as was released by the 1932 earthquakes could reaccumulate in as little as 80 years at the average Rivera-North America convergence rate of 3 million years. It could accumulate sooner if the convergence rate has increased significantly.
A long-term project was begun in March 1995 to monitor subduction-related strain accumulation and investigate rift-related deformation above the subducting Rivera and Cocos plates. Initial measurements of the lengths and orientations of 15 sites in western Mexico (Figure 1) were made using GPS satellite receivers. These instruments use a constellation of orbiting satellites to make extremely precise, millimeter-level measurements of site locations.
Fig. 1. Neotectonic features in the Jalisco block and Rivera plate region. Solid circles represent GPS sites occupied in 1995. The yellow diamond shows estimated location of the June 3, 1932, earthquake, and the white short dashed lines show the estimated extent of aftershocks for the June 3, 1932, earthquake and the zone where the June 18, 1932, aftershock caused the most damage. Red regions indicate volcanic centers active in the last few thousand years. The green-shaded area onshore is Lake Chapala. The offshore region shows the vertical gravity gradient illuminated from the southwest (gravity field supplied courtesy of Dave Sandwell). Solid white lines show active plate boundaries and dashed white lines show inactive plate boundaries. The white dots and crosses show magnetic anomaly crossings being used by S. Traylen and C. DeMets to model the post-10 Ma kinematic evolution of the Rivera plate.
The Jalisco region is an attractive setting for a GPS experiment. It offers a rare opportunity to observe how a continent deforms above two subducting slabs. The Cocos plate subducts into the mantle approximately 1.5 to 2 times faster than the Rivera plate. This requires that the two plates move with respect to each other as they subduct beneath the overlying continent. How motion between the downgoing slabs affects deformation of the overlying continent is unclear, but subduction is known to affect faulting and volcanism within the overlying continent profoundly. Much of the faulting and volcanism is focused along two active rift systems within the Jalisco region. They are the north trending Colima graben, which may lie above the subducted Rivera-Cocos plate boundary, and the northwest trending Tepic-Zacoalco rift zone, which extends northwest to the Gulf of California. These rift systems intersect a third southeast trending zone of active faulting, the Chapala rift, and form a continental triple junction near Guadalajara, Mexico's second largest city.
The two plates subducting beneath Jalisco also affect the volcanism of the overlying Jalisco block. The Jalisco region is a volcanologist's playground, containing what may be the world's most diverse volcanic rocks. Much of the volcanism is focused in grabens that indicate active regional extension. The Tepic-Zacoalco fault zone and Colima graben, which separates the Jalisco block from the North American plate, contain both active and dormant andesitic volcanoes, including Volcan Colima, North America's most active volcano. The Chapala-Tula fault zone, part of the TransMexican Volcanic Belt, also contains many active volcanoes. Additional volcanism is also concentrated in several smaller grabens, which form a volcanic front approximately 90 km from the trench. The Rivera slab dips about 45°, reaching depths of 100 km beneath the volcanic front defined by the small grabens near the trench. Speculation about a link between the subducting Rivera and Cocos plates and the onshore history of volcanic activity and plate tectonics of the Jalisco region have led to several types of models for the deformation in the region. In one model the Colima graben represents deformation of the continent above a tear in the subducting slab. Others consider the Colima graben to be an early manifestation of an eastward jump of the East Pacific Rise. Other models invoke differential motion and oblique subduction of the Rivera and Cocos plates beneath the continental margin to explain extension along the Colima graben and other fault zones. One model attributes the volcanism and rifting at the Colima graben to a hotspot beneath the Guadalajara region. GPS measurements are best suited for distinguishing between models that invoke differential motion and oblique subduction of the Rivera and Cocos plates. These models postulate that the dip, configuration, convergence rate, and angle of obliquity of the downgoing slabs are important in controlling deformation and volcanism within the overlying plate. In particular, GPS measurements will be used to test the predictions of two alternative models for present-day Rivera and Cocos plate motion beneath the continental margin. One model uses magnetic anomalies from the Pacific-Rivera rise in conjunction with plate circuit closures to predict differential motion of about 15 mm/yr between the subducting Rivera and Cocos plates. An alternative model predicts that motion between the Rivera and Cocos plates must be slow or nonexistent because there is no clear boundary west of the Middle America trench and because of the seismic characteristics of the two subducting plates.
These models predict different patterns of large-scale strain accumulation within the overlying North American plate. According to the former model, strain should accumulate significantly faster over the subducting Cocos plate than over the Rivera plate, with a discontinuity somewhere in the vicinity of the Colima graben. According to the latter model, strain rates within the North American plate should decrease to the northwest, with no discontinuity near the Colima graben.
The regional strain field is likely to be dominated by subduction-related strain, with a smaller contribution from rift-related fault slip and block rotation. GPS measurements, when combined with seismological and volcanological constraints on the dip and configuration of the subducting plates, should lead to a suite of plausible models that link the onshore deformation to the motions of the slabs.
In the short term, the along-trench variations in the strain rates and maximum shortening directions will be used to find the transition between lithosphere overlying the subducting Rivera and Cocos plates, and to determine whether this transition coincides with the Colima graben and/or with a gravity anomaly attributed to the subducted portion of the Rivera-Cocos boundary. More challenging objectives include measuring vertical uplift rates and extracting information on rift-related deformation from the subduction-related strain. Next year progress will be made on three fronts. A passive seismic array in the Jalisco region will be used to image the main subduction fault and other faults that generate microseisms. In addition, GPS measurements at several sites in the existing network will be repeated, and closely spaced geodetic monuments along a trench-normal profile will be installed and surveyed. Finally, we will explore the possibility of installing one or more continuously operating GPS receivers to establish a mini-North America reference frame. The Jalisco region of western Mexico is an excellent site for studying how continental lithosphere responds to two plates underthrusting at different velocities. Measurements within a newly installed GPS network in this region will determine whether significant strain accumulation accompanies subduction. They will also determine whether strain accumulation above the Cocos and Rivera plates differs significantly, and if so, where the transition between the strain fields is located. GPS measurements over several decades should detect vertical uplift and motion across the many grabens of Jalisco, offering for the first time regional-scale kinematic observations that could help understand the enigmatic volcanic rocks in this region.
Source: Eos, Vol. 76, October 17, 1995, p. 417.
Return to Science and
Society
Return to Starting
Point
