The seismicity of north-central New Mexico with particular reference to the Cerrillos earthquake of May 28, 1918
— Kenneth H. Olsen

Earthquakes within the State of New Mexico have been known since 1849 when a swarm near Socorro was described in a diary by a U.S. Army surgeon (Hammond, 1966). Prior to this date, the relative absence of written records among the small and widely scattered settlements of the Indian and Spanish cultures means that even fairly substantial earthquakes (i.e., magnitudes perhaps slightly in excess of 6) would have gone unreported. For the period from 1849 until 1960, the record of seismic activity within the state is dependent largely on felt effects which were published principally in the local newspapers of the day. Northrop (1961, 1976) has compiled many of the data from old newspaper files which indicate over 600 felt earthquakes in the state. The majority of these occurred along the Rio Grande valley, the most densely populated region. Since most of the state outside the agricultural lands of the Rio Grande valley was populated very sparsely, this undoubtedly has led to a bias in the reporting of felt shocks and to perhaps an overemphasis of the seismicity of the Rio Grande rift relative to neighboring physiographic provinces. However, these felt data do indicate correctly that the section of the rift from Belen to Socorro had the highest rate of earthquake occurrence, as has been confirmed by modern instrumental studies (Sanford and others, 1979a).

 

Shortly after the turn of the 20th century, various universities, government agencies and other organizations (most notably the Jesuit Seismological Association) began installation of seismograph stations at widely scattered locations throughout North America. The earliest instruments were quite primitive by modern standards, and since they were designed primarily to register long-period waves from very large, distant earthquakes (teleseisms), they had very low sensitivity to strong local and regional shocks. The closest of these stations to New Mexico, at Denver and at Tucson, began operation about 1909. Because of the distance and relative insensitivity of seismograph stations in nearby states, only marginal instrumental data on a few of the larger New Mexico earth-quakes are available for the period 1910-1960.

 

The first permanent seismograph station in New Mexico was installed at Socorro (SNM) in 1960 near the campus of New Mexico Institute of Mining and Technology (NMIMT). This high-gain, short-period instrument permitted the initiation of studies of the seismic activity in the vicinity of Socorro (Sanford, 1963). Various organizations began operating high-quality, modern stations in 1961-1962 at Albuquerque (ALQ) and Las Cruces (LCN), New Mexico, and at Tucson (TUC) and Payson (TFO), Arizona. Readings from these stations combined with the SNM data then permitted the location of many earthquake epicenters throughout the state. This distribution and density of seismograph stations achieved essentially complete coverage of all earthquakes within New Mexico stronger than the detection threshold of Richter magnitude 2.4 (San-ford and others, 1976a,b). The magnitude scale used in this work was an adaptation of the "local" (MO scale originally developed by Richter (1958) for use in southern California. Detailed mapping of even smaller shocks for the north-central part of the state was initiated when the Los Alamos Scientific Laboratory (LASL) installed a network of continuously recording high-gain stations beginning in 1973. A similar network in the vicinity of Albuquerque begun in 1976 by the USGS Albuquerque Seismological Laboratory (ASL) has permitted the extension of these detailed studies for the entire section of the Rio Grande rift from the Colorado-New Mexico border to Socorro.

 

The instrumental data on earthquake epicenters throughout the state since 1962 have been summarized by Sanford and others (1976a, 1976b, 1979b), who showed that the greatest concentrations of continuing activity are centered near Socorro and northwest of Espanola. Reviews of the details of seismic activity, specifically for the Rio Grande rift, including the historic felt reports as well as instrumental data, are given by Sanford and others (1972, 1979a). Sanford and coworkers also have conducted specialized and detailed studies of the microearthquake activity in the vicinity of Socorro and the association of this seismicity with a probable magma body at midcrustal depths (Rinehart and others, 1979; Sanford and others, 1977).

 

Interest in the assessment of possible geologic and seismic hazards in the north-central section of New Mexico has increased rapidly in recent years. Not only is this region the site of a developing major urban area including the state's two largest cities of Albuquerque and Santa Fe, but the area contains, or is in close proximity to, significant potential resources of geothermal energy and ground water in addition to known major production areas of coal, uranium, oil and gas.
One of the principal goals in the installation of the LASL seismic network north of Albuquerque was to obtain better information on the long-term seismic risk to this section of the state and to correlate the seismic patterns with the contemporary tectonic activity. Because of the history of development of seismographic stations, a less detailed picture of seismicity over a shorter time period is available for the region between Albuquerque and the Colorado-New Mexico border than for the Albuquerque-Socorro segment. This is also the situation for the historic felt earthquake data, a consequence of both a lower population density and probably a lower rate of earthquake occurrence in the 1849-1979 time period. The evaluation of earthquake risk in a particular geographic area relies on three complementary techniques: (1) the determination of the centers and rates of present day activity by instrumental means, (2) a compilation of earthquake information over the somewhat longer periods covered by historic records, and (3) an assessment of probable sizes and recurrence intervals of larger earthquakes over geologic time scales by geological techniques involving measurements and dating of offsets on fault scarps (Allen, 1975). Because of the highly episodic nature of earthquake activity, all three techniques are necessary in order to arrive at a valid estimate of the earthquake potential of a region. The instrumental studies are capable of defining patterns of usually small magnitude and more frequent events, but data are usually available for only a few years; the historic information is a valuable adjunct since it is often available for tens to hundreds of years, and serves to define better the occurrence rate of the infrequent but higher magnitude shocks which have most of the potential for creating damage. For north-central New Mexico, these three types of earthquake risk studies have been undertaken only during the past few years; summaries of some of the early results, mainly in the vicinity of Los Alamos, are contained in reports by Budding and Purtymun (1976) (geological tech-niques), Newton and others (1976) (all techniques), Sanford (1976) (mainly historic felt reports), and Sanford and others (1979a) (instrumental results).

 

The objective of this paper is to examine the relationship between the instrumental studies of microearthquakes carried out by LASL since 1973 and the Cerrillos earthquake of 28 May 1918, which may have been the largest earthquake in the area that is known from the historical record. In addition to its importance in helping define the seismicity of north-central New Mexico, the Cerrillos shock is of interest to the student of New Mexico earthquakes for several reasons: (1) it usually is listed among the five largest known historic earthquakes in the state but is somewhat isolated from the Socorro region where most of the other large shocks occurred; (2) to my knowledge, it is the only New Mexico earthquake where major ground breakage is alleged to have occurred; and (3) its size and location suggest its potential importance in understanding the tectonics of the Rio Grande rift (Sanford and others, 1979a).

 

Many of the facts regarding the Cerrillos earthquake have been known to investigators of New Mexico earthquakes for a number of years, but the details are contained in widely scattered publications. In addition, I know of no previous attempts to locate possible eyewitnesses among still living individuals or to evaluate potential data that may exist on the recordings of the few early seismograph stations of the western United States. For these reasons, I believe it is of some value to review here the circumstances of the Cerrillos earthquake and try to relate it to modern work on the seismicity of the region.