Submitted to Seismological Research Letters
El Salvador 2001: Earthquake disaster and disaster preparedness in a tropical volcanic environment
Cinna Lomnitz and Sergio Rodríguez Elizararrás
Universidad Nacional Autónoma de México
We visited El Salvador two weeks after the disastrous earthquake of 13 January 2001 (Mw=7.6) at the request of the Ambassador of El Salvador in Mexico. Mexico is close to El Salvador in many ways, and one of us had studied an earlier earthquake disaster in this country (Lomnitz and Schultz, 1966). One result is the present report of our fact-finding emergency mission. We thank the government of El Salvador for timely and effective assistance and hospitality. It is a pleasure to acknowledge the valuable comments of José Luis Chávez and Franz Sauter in the course of field work.
The earthquake
The main earthquake occurred on Saturday, 13 January 2001 at 11:33 a.m. local time off the south coast of El Salvador at 12.830N, 88.790W with a focal depth of around 40 km. A second major shock (Mw=6.6) occurred on February 13, 2001. There were over 1,000 deaths, mostly in landslides. The focal mechanism of the first event suggested intraplate normal faulting in the subducting Cocos Plate. The trend of the fault was nearly parallel to the coast. The magnitude was Mw=7.6. There were reports of a minor tsunami. Over a thousand aftershocks occurred in a nearly circular area with a diameter of about 50 km (Fig. 1). The rate of occurrence of aftershocks decayed roughly according to Omori’s Law. The second major shock occurred inland from the first one, near the town of San Vicente: it had a shallow focus and a strike-slip mechanism.
The capital city of San Salvador is located more than 100 km from the epicenter. Accelerations of 0.3 g to 0.6 g were recorded in the city, and substantial seismic damage to housing was feared. We found, however, that the seismic behavior of structures in San Salvador--a rapidly growing city of nearly 2 million--had been excellent. Non-structural did damage did exist but was not much in evidence: we had to look for it.
The Central American University, a private college, had obtained a good set of strong-motion records from their network of nine digital stations, most of them in San Salvador and nearby localities. They recorded peak horizontal ground accelerations of up to 0.89g at the port of La Libertad, and up to 0.59g at inland locations. The frequencies seemed high, perhaps because of the focal depth. We were puzzled that no more substantial structural damage had occurred in spite of the high intensity values.
Major loss of life occurred in a landslide in Las Colinas, a middle-class residential development in the Santa Tecla district of greater San Salvador. The scar of the landslide was relatively shallow and rectangular in shape (Fig. 2). During the strong ground motion, the hillside flowed down with a hissing noise; the homes in the path of the landslide were destroyed and about 400 people were killed. The homes on either side of the landslide were evacuated for fear of further slides, which have not yet materialized at this writing; but further slides were expected in the rainy season.
The debris of the slide had been cleared and the evacuated area was cordoned off and under military control.
Landslides
The Las Colinas slide occurred in pyroclastics of the Cordillera del Bálsamo, a hilly, deeply dissected mountain range of moderate altitude located between the valley of San Salvador and the ocean (Fig. 1). When we visited the slide area we found that the material in this formation was a fine wind-blown volcanic ash of early Neogene age. It had not rained for some months and the roads were extremely dusty, but the slide material was moist. A similar material can be seen on the road from the airport and in fact over large areas of the country and Central America. Erosion in this formation frequently forms deep gullies with near-vertical slopes;fortunately the northwestern slopes of the Balsamo Range toward San Salvador are wooded—a fact which may help prevent some of the more dramatic forms of erosional damage. It was suggested to us that the developers had contributed to the accident by removing the vegetation in the area.
A helicopter overflight was kindly provided by the Salvadorean Air Force. We discovered many small and middle-sized slides in the Bálsamo Range, all in volcanic ash and at an altitude range of 500 to 1,000 meters. The ash layers looked horizontal and fairly homogeneous; their source was presumably in the Ilopango Caldera. We found no evidence of recent volcanism in the Bálsamo Range itself, but the Central Valley was lined with huge stratovolcanoes and large calderas.
A white ash horizon could be traced from slide to slide over many kilometers. We surmise that confined aquifers may be present at different levels in the Bálsamo Range. Large pressure pulses of up to one bar caused by a distant earthquake could help destabilize the steeper hillsides and trigger landslides. A photograph provided by a local engineer, of a fresh scarp in volcanic ash, showed water leaking out of a horizontal aquifer cut by the landslide. Installation of piezometers in selected aquifers should provide a test of this hypothesis.
We surveyed the Santa Ana Volcano from the air because of some reported increase in activity, but we found none. Fumarolic activity was minor. However, changes in water level of the caldera lake at the foothills of the Santa Ana were credibly reported. None of the other volcanoes showed any sign of increased activity.
On foot, we surveyed the ridges overlooking the Las Colinas slope failure and we found some fine cracks or fissures on level ground. We could not see any evidence of relative motion but it seemed possible that further slides could occur in this material in the next rainy season (usually starting in May). Some large aftershocks could still take place at such a late time. If we were homeowners in the undamaged section of Las Colinas we would probably prefer to stay away from the area.
The four-lane Pan American Highway is the main land connection of the country; it traverses El Salvador lengthwise from the Guatemalan border in the north to Honduras and Nicaragua in the south. Two major landslides blocked the highway, one north of San Salvador at Los Chorros, and one to th south just beyond Lake Ilopango (Fig. 1). Los Chorros (“The Jets”) is an area of hot springs, which seems to point to a causal connection with ground water movements. At the time of our visit the southbound lanes of the highway at Los Chorros had been cleared and traffic on this section of the highway had resumed—southbound in the mornings, and northbound in the afternoons.
The landslide south of Lake Ilopango was much larger and the Department of Public Works had not yet succeeded in clearing the highway. Traffic was detoured over country roads to bypass the slide area. As far as we could determine, there were dozens of smaller slides blocking dirt roads in the countryside, especially in the coffee-growing areas of the Bálsamo Range. The event of February 12, 2001 occurred after our departure.
Damage due to seismic waves
In the course of our overflight we visited several localities where especially heavy damage had been reported. The village of Somosagua is located on a ridge in the northern foothills of the Bálsamo Range, reflecting the regional preference for hilltop settlements. This preference may be attributed to fear of floods and landslides though it might also be due to strategic considerations related with the recent civil war. Houses in Somosagua are mostly located along one long street. About 30% of the houses show moderate to severe damage. In two-level brick masonry structures the presence of sparse 20-cm concrete columns was often sufficient to prevent total collapse.
In contrast, in the village of San Agustín (Usulután Department) the prevailing construction type was bahareque and adobe reflecting extreme poverty conditions. About 50% of the housing collapsed. This is in the southern part of the country, relatively nearer to the epicenter; however, the nearby provincial capital of Usulután presented no evidence of damage from the air. All damaged localities which we visited were located well inland, on the top of ridges--suggesting possible geometric amplification due to hilltop location. Damage at coastal locations, e.g. La Libertad, was not significant in spite of the high accelerations.
The flight of populations from the countryside to the cities had accelerated in recent years as a result of insecurity due to the armed conflict. However, when compared to the local earthquakes of 1965 (M6.5) and 1976 (M5.9), the seismic damage in the 2001 earthquakes was less significant. Local earthquakes near San Salvador City are clearly more damaging to structures because of the shorter epicentral distance and the shallower focal depth. On the other hand, offshore subduction events are a potential cause of severe landslides, especially in the Bálsamo Range and in roadcuts. Such accidents were not observed in local earthquakes.
Emergency measures
By the time of our arrival most foreign emergency missions had returned home. Military units of some Latin American nations were still there, providing much-needed assistance in the countryside. We were especially fascinated with the members of a Venezuelan army unit stationed in Comasagua. These fine boys and girls provided real help. Medical, dental, pediatric and obstetric treatment was deeply appreciated by the local people; it was competently and cheerfully provided with no questions asked. No difference was made between earthquake victims and the rural population at large.
The construction equipment of the Ministry of Public Works was thinly stretched over hundreds of earth slumps and seemed inadequate to the task. Most of the heavy machinery was concentrated at the two major slides on the Pan American Highway. The many homeless were not much in evidence; in the countryside they had been housed in temporary huts under the supervision of the armed forces, or with relatives. No homeless people were seen in the streets of San Salvador, presumably because the middle class had sustained the brunt of the damage. There was a palpable desire in the capital to forget the earthquake drama as quickly as possible.
This may be the wrong time to forget. Normal earthquakes in subduction zones rarely have the last word. They are apt to be followed within four or five years by large thrust events or by shallow intraplate events. Even if the destabilization of slopes in the rainy season does not materialize, it would be sensible to interpret the 2001 event as a warning. The guard should not be let down and defenses against disasters should be hardened.
According to some press reports, the developers at Las Colinas had been authorized to proceed in spite of existing zoning regulations designed to prevent residential developments on the slopes. The location was a desirable one because the Santa Tecla area was relatively safe from guerrilla operations. After pacification the pressure from developers subsided as there seems to be plenty of available land in the valley; but there is a definite need for setting up enforceable zoning regulations in order to protect the hillsides from future deforestation and encroachment by developers.
Conclusion
There is a need for a highly professional, politically independent agency to supervise research and evaluation of natural hazards and prevention of earthquake damage in El Salvador and Central America in general. Conditions are sufficiently unique to warrant a special effort. Central America features a combination of volcanism and tropical climate found nowhere else in the world except in parts of Indonesia. Slopes in volcanic ash tend to be concave rather than convex, which produces interesting variations on Mohr’s circle. Fluctuations of permeability in depositional sequences are due to little-understood climatic change. Tropical rainfall patterns must be taken into account in designing remedial drainage of potentially unstable slopes.
The Center for Geotechnical Research (CIG) of the Ministry of Public Works —a public agency in charge of hazard maps, building codes and monitoring of earthquake activity—is apparently under re-organization. The original strong-motion records were not available as they had been sent to Menlo Park for processing; no copies were kept. The German geologists who had been actively mapping the country for almost a century left in 1980. Whatever the upshot, competent personnel at all levels will be sorely needed. We were moved to offer the training facilities of our graduate program in the earth sciences, a successful project which is providing masters’ and doctors’ degrees for more than 400 students at our university. Scholarships are available to qualified foreign applicants.
We are painfully aware that even the best-trained young graduates will be of little use unless the country can provide decent jobs for them, including adequate careers and professional challenge. We recommended a two-pronged approach consisting in graduate-level training in the earth sciences for a hand-picked group of young Salvadorean students, plus the simultaneous creation of an independent research facility endowed with adequate support to lend some permanence and credibility to this important effort. Because of common geographical, cultural and human factors we trust that the study opportunities provided by our university (and perhaps by other universities in Mexico and Latin America) could be decisive in helping overcome the lag in preparedness which we detected. El Salvador is subject to at least ten natural emergencies a year, including volcanic eruptions, hurricanes, floods, and droughts. The 2001 earthquake did not approach the level of severity of some previous earthquakes, yet it wiped out the equivalent of half the annual gross national income. A small investment in preparedness would pay off handsomely.
References
Alvarez G., Salvador (1987). Informe Técnico-Sismológico del Terremoto de San Salvador del 10 de Octubre de 1986, Centro de Investigaciones Geotécnicas, San Salvador, El Salvador.
Centro de Investigaciones Geotécnicas (1983). Boletín Sismológico, San
Salvador, El Salvador.
Centro de Investigaciones Geotécnicas(1985). Informes Sísmicos, San
Salvador, El Salvador.
Geological Survey (1976). The Guatemala Earthquake of February 4th, 1976, A Preliminary Report. Professional Paper 1002. Reston, Va.
Jordan, J.N., and Martínez, M., (1979). Seismic History of El Salvador (unpublished).
Lomnitz, C. , and Schulz, R. (1966). The San Salvador Earthquake of May 3, 1965. Bull.Seism. Soc.Am. 56, 561-575.
Martínez H., M.A. (1978). Cronología Sísmica y Eruptiva de la República de El Salvador a partir de 1520, Centro de Investigaciones Geotécnicas, Ministerio de Obras Públicas, San Salvador.
Meyer-Abich, Helmut, (1952). Terremoto de Jucuapa en El Salvador (América Central), 6-7 de Mayo de 1951, Comunicaciones, No. 2, Instituto Tropical de Investigaciones Científicas, Universidad de El Salvador.
White, R.J. et al. (1987). The San Salvador Earthquake of October 10th, 1986 - Seismological Aspects and other Recent Local Seismicity. Earthquake Spectra, Vol.3.
White, R.A. and Harlow, D. (1993), Destructive upper crustal earthquakes in Central America since 1900" Bull. Seismol. Soc. Am.,83,4.