A Detailed Microseismicity Study and Current ... - GeoScienceWorld

Bulletin of the Seismological Society of America, 90, 5, pp. 1133–1142, October 2000

A Detailed Microseismicity Study and Current Stress Regime in the Peninsular Ranges of Northern Baja California, Mexico: The Ojos Negros Region by J. Frez, J. J. Gonza´lez, J. G. Acosta, F. A. Nava, I. Meńdez, J. Carlos, R. E. Garcıá-Arthur, and M. Alvarez

Abstract The NW-trending San Miguel fault system is one of the most important seismogenic systems in northern Baja California, and the Ojos Negros region, comprising the Ojos Negros valley and bordering areas, is one of its most active regions. Within this region are found most of the mapped faults of the system: Ojos Negros, Tres Hermanos, most of San Miguel, and portions of the Vallecitos fault, which makes this a very important region from the points of view of intraplate tectonics and regional seismic hazard. A detailed microseismicity (0.2 ⱕ M ⱕ 4.0) survey of the Ojos Negros region, carried out in 1997 (one month, 13 Reftek stations recording at 200 samples/sec, plus two permanent RESNOM stations and other less sensitive instruments), yielded important results about the fault activity and the stress regime in the region. Our results are based on 278 hypocenters and 50 focal mechanisms selected from almost 2500 earthquakes recorded at a minimum of four stations. The selected database is comprised of good quality local events, for which the hypocentral depth can be reliably estimated. Locations and focal mechanisms were obtained using an improved velocity model (Sierra97) for this part of the Peninsular Ranges. The hypocenters tend to cluster in space and time, with cluster interepicenter separations of the order of the location error (Ⳳ1 km). The Ojos Negros valley (as defined by its sedimentary soil) is roughly covered by epicenters. The Tres Hermanos fault shows no significant seismicity, and the few earthquakes near (although not very close to) its southern third seem to be associated with seismicity that extends into the valley. Seismicity associated with the Ojos Negros fault consists almost exclusively of one large cluster. The San Miguel fault, the most active fault in the area, has epicenters within a 6–8 km wide band centered along its mapped trace. Most focal mechanisms are strike-slip with a minor normal component, while others are dominantly normal. The resulting pattern for the valley indicates a regional extensional regime with the average T axis in the ENE-WSW direction, and P axes distributed along an N-S strip with a slight inclination and concentrated near the poles. Introduction Northern Baja California and southern California constitute part of the active boundary between the North American and Pacific Plates (Fig. 1) (for details see e.g., Sua´rezVidal et al., 1991). All large earthquakes with ML ⱖ 6.0, and most of the microseismic activity in northern Baja California, occur along two fault systems (e.g., Frez and Gonza´lez, 1991; Frez and Frıás, 1998): the southern extension of the San Andreas fault system (which includes the Imperial and Cerro Prieto faults, located along the Salton Trough)

and the San Miguel fault system, which crosses the Peninsular Ranges in a roughly NW-SE direction. The first system is the main western boundary for interplate transcurrent motion, and accounts for about 85% (Gonza´lez et al., 2000) of the total 49 to 50 mm/yr estimated for the relative motion between the North American and Pacific Plates (DeMets et al., 1987; Shen et al., 1997; Gonza´lez et al., 2000); this slip generates most of the destructive earthquakes in Baja California. The Vallecitos–San Miguel fault system takes a large

1133

1134

J. Frez, J. J. Gonza´lez, J. G. Acosta, F. A. Nava, I. Meńdez, J. Carlos, R. E. Garcıá-Arthur, and M. Alvarez

Figure 1.

Regional tectonism, main faults (E.), and seismic zones (S.Z.) in northern Baja California and southern California, adapted from Frez and Gonza´lez (1991). Hatchured areas are assumed spreading centers (Lomnitz et al., 1970; Elders et al., 1972). The study area, covering about 25 ⳯ 50 km2 and comprising most of the San Miguel fault system, is shaded gray and the rectangle indicates the area covered in subsequent figures.

part of the remaining interplate motion: although paleoseismicity measurements indicate displacements of about 0.05 to 0.55 mm/yr (Hirabayashi et al., 1996), GPS measurements indicate current rates of about 5 Ⳳ 2 mm/y (Gonza´lez et al., 2000). The San Miguel fault system, comprised of the San Miguel, Vallecitos, Tres Hermanos, and Ojos Negros faults (Figure 1) features a microearthquake activity level (Fig. 2) comparable to that of the San Andreas system. The largest instrumentally recorded earthquakes in the San Miguel system are: one ML ⳱ 5.8 in 1963 at the junction of the Vallecitos and San Miguel faults (Leeds, 1979), six ML ⬎ 6.0 that occurred in 1954 and 1956 at the SE section of the San Miguel fault (Shor and Roberts, 1958; Leeds, 1979), and one M ⳱ 5.7 earthquake in 1949 in the Vallecitos fault. Currently, regional seismicity is monitored by two regional networks: the Southern California Seismic Network (SCSN) and the Red Sismica del Noroeste de Me´xico (RESNOM). The SCSN has published the Southern California Cat-

alog (SCC) since 1932 (Nordquist, 1964), and has experienced large increases in number of stations and coverage (e.g. Norris et al., 1986). The SCC is complete for magnitudes 2.1 to 3.5 (varying with distance to the SCSN) and above in the study area (Frez and Gonza´lez, 1991). The RESNOM catalog, Cata´logo del Noroeste de Me´xico (CNOM), achieved regional coverage since 1982 (Medina and Duarte, 1979; Hinojosa, 1990) and is complete for magnitudes ⱖ2.1 since 1987. Prior to 1973, the epicentral determinations in the SCC for the study area may have significant errors, as reflected in Leeds (1979) and Doser (1994) who relocated the epicenters of several important earthquakes. In particular, the 1954 ML ⬎ 6.0 earthquakes, originally listed by the SCC near the Agua Blanca fault, were relocated by Leeds (1979) to SE of the San Miguel fault. Although the corrected epicenters have been used in various studies (e.g., Frez and Gonza´lez, 1991; Doser, 1994; Frez and Frıás, 1998), they have not been incorporated into the SCC, so that they are not always taken

A Detailed Microseismicity Study and Current Stress Regime in the Peninsular Ranges of Northern Baja California

1135

Figure 2.

Microseismicity from the SCC catalog for the study area and neighborhood, which shows the diffuse seismicity within the Ojos Negros valley, and the conspicuous gap between it and the seismicity concentration at the SW tip of a large SW-NE trending distribution, which does not correspond to any of the mapped faults in the region.

into account in articles dealing with earthquakes in this region (e.g., Brehm and Braile, 1999). There is abundant seismicity associated with the area located between the San Miguel, Ojos Negros, and Tres Hermanos faults (Figs. 1 and 2), which includes the Ojos Negros valley and bordering regions and which we will henceforth call the Ojos Negros region. Reyes et al. (1975) observed the relatively large number of 117 events over 23 hours at one station in the Ojos Negros valley. Soares (1981) also found high seismicity for the northern part of this valley, most of it in clusters, with depths between 1 and 24 km. Although there have been several studies on earthquakes associated with the San Miguel fault system (Johnson et al., 1976; Reyes et al., 1975; Brune et al., 1979; Nava and Brune, 1982, 1983; Soares, 1981; Gonza´lez, 1987; Rebollar y Reichle, 1987), none of them has precise hypocentral and fault-plane-solution determinations, due to limitations in both the number and the distribution of seismograph stations, with large gaps in distance and azimuth coverage. Since detailed knowledge of the seismic processes go-

ing on along the San Miguel fault system is important from the points of view of intraplate tectonics and of regional seismic hazard, we felt that detailed microseismicity studies of the area would be valuable for identifying currently active faults and other seismogenic features, and for learning about the stress regimes. Here, we use precise determination of hypocenters and fault-plane solutions in the Ojos Negros region,and we hope to continue complementing or supplementing our knowledge about this important area in future detailed studies.

Seismograph Network and Data An array of 19 stations (Fig. 3) operated from 20 May to 23 June 1997. Thirteen stations consisted each of a Reftek 72A-07 digital seismograph with a 111A GPS and a MARK L-22 three-component, short-period seismometer, with triggered event recording using 200 samples/s and 32 bit words. These instruments were kindly loaned to our project by IRIS/ PASSCAL. The portable station data were complemented

1136


Figure 3.

Epicenters of the local earthquakes shown as rhombi sized proportionally to magnitude, with lines indicating estimated location errors. The portable Reftek seismograph stations are indicated by triangles, the RESNOM stations by squares, and the circles indicate complementary stations that operated over short times with relatively low amplification. Shaded areas and labels indicate the seismicity groupings discussed in the text.

with data from two RESNOM stations, PGX and RDX (Fig. 3). Since these two stations have lower sensitivities than the Reftek seismographs, they contributed data for only ⬃10% of the recorded earthquakes. Two other types of instruments operated over short time periods. During the first two weeks, five Sprengnether MEQ800 smoked paper seismographs with Ranger SS1 vertical seismometers at stations ELCA, ELOA, MARI, CDEM, and

SJDI, were used to establish seismic activity and noise levels;

all except the last one were later substituted by Reftek instruments. Also used in about 10% of the hypocentral and focal mechanism determinations were data from two Kinemetrics SSR-1 and two Terra-Tech DCS-302 seismographs at stations PARM, TULA, LHER and DUKE, belonging to an array gathering data for a parallel study on soil response. During the campaign, thousands of events were re-

1137


corded, about 2500 of them by at least four stations. The seismograms were visually inspected, one by one, and were initially classified as “local” (within the network or very close to it) or “regional” (⬃20 km or more from the network) based on their S-P times. More than 50 regional earthquakes (those with the largest magnitudes in the regional catalogs) close to the network’s periphery were located; a large number of these events were associated with the Sierra Juarez fault which had a considerable microseismic activity during 1997. This activity will be the subject of a separate publication. For local events, P and S arrival times were then carefully picked for earthquakes recorded at a minimum six stations. P-wave arrival times have a precision close to 0.03 sec, determined by comparing picks from various readers; this precision improves to 0.02 sec for impulsive arrivals in records with high signal-to-noise ratio. Dubious P arrivals, surpassing ⬃0.5-sec uncertainty, were discarded. Precision for S-wave arrivals is more variable (⬃0.05 sec), even though the three components were always used for arrival picking; particularly uncertain are S arrivals emerging from large amplitude secondary P arrivals that appear in about 10% of the seismograms. Programs HYPO71 (Lee and Lahr, 1975) and FPFIT (Reasenberg and Oppenheimer, 1985) were used to determine hypocentral locations and fault-plane solutions, respectively. For hypocentral locations within the Peninsular Ranges RESNOM routinely uses the Nava and Brune (1982) velocity structure. While appropriate for regional locations, this model does not resolve the shallow velocities needed for very local determinations (their fit of arrival times for the first layer was forced to pass through the origin). To the Nava and Brune (1982) arrival times, we incorporated the shallow velocity Vp ⳱ 4.6 km/sec found by Camacho (1989) for the San Miguel fault region, and recalculated the model using an unrestricted least-squares fit for the earlier arrivals. A Wadati D(S-P) vs. DP analysis (Fig. 4) yielded Vp /Vs ⳱ 1.709 Ⳳ 0.007, locally confirming the Nava and Brune (1982) estimate of Vp /Vs 艑 Z3. The resulting Sierra97 model used for locations in this article is shown in Table 1. Use of other velocity models that have been applied for this region: Nava and Brune (1982), Hadley and Kanamori (1979), and an “extreme” model featuring slower velocities and a velocity gradient for the shallower 5 km, to represent compaction effects (e.g., Kern, 1982), did not significantly change our results described subsequently. Given adequate coverage in distance (at least one station at an epicentral distance smaller than the source depth) and azimuth (gap ⬍180⬚), hypocenters obtained using the different velocity models change by less than 1 km. Hypocenters were determined for events recorded at a minimum of six Reftek stations so that most hypocenter determinations are based on more than 10 P and S arrivals. Given the dense station coverage of the area, we are of the opinion that this selection does not bias the spatial distribution. Two-hundred seventy-eight hypocenters were relia-

Wadati diagram of D(S ⳮ P) vs. DP, for HYPO71 A-quality earthquakes.

Figure 4.

Table 1 The Sierra97 Seismic Velocity Structure VP (km/sec)

VS (km/sec)

Depth (km)

4.60 5.75 6.57 6.95 8.02

2.66 3.32 3.79 4.01 4.63

0.00 0.50 5.23 19.88 42.02

bly determined, with HYPO71 qualities A (9%), B (55%), and C (36%), and with one or more stations at epicentral distances smaller than the focal depth, excepting a few shallow earthquakes with depths less than 5 km. The formal errors given by HYPO71 are smaller than 1.0 km for 90% of the cases and never exceed 1.50 km; 80% of the locations have rms errors less than 0.07 sec. In order to check the reliability of the few locations made with less than complete coverage for events occurred toward the end of the experiment, 14 events with good coverage were relocated using only part of the data with azimuthal gaps greater than 180⬚. For epicenters, the smallest and largest variations were 0.04 km and 1.63 km, respectively, with a 0.96 average: except for two events outside the array, depths varied (absolutely) by less than 1.7 km, with a 0.69 km average.

Magnitudes Local magnitudes were computed for all located events using the empirical formula determined by Vidal and Munguıá (1999) for the Peninsular Ranges of northern Baja California:

冢100冣

M ⳱ log A Ⳮ 1.1319 log

r

Ⳮ 0.0017 (r ⳮ 100) Ⳮ 6.4472,

1138


where r is the hypocentral distance in km, and A is the maximum amplitude, in mm, of an equivalent standard WoodAnderson (W-A) seismogram (Richter, 1958) having an 0.8 sec free period, 0.8 critical damping and unit amplification; the 2800 amplification of the standard W-A is included in the last term of the formula. To obtain the equivalent W-A seismogram, the digital records from our portable stations were corrected for base-line shift and parabolic trend, deconvolved of the Reftek response, and convolved with the W-A response. Magnitudes were computed for each of the two horizontal (N and E) components from stations EWAG and LOSC, both sited on granitic terrain, to obtain a maximum of four magnitude estimates for each earthquake; the average from these estimates gave the magnitude assigned to each earthquake. Our magnitude determinations were checked against the 14 magnitudes determined by the SCC|RESNOM catalogs, of which the four largest were: 20 July, M ⳱ 3.7|3.8 and M ⳱ 2.6|2.6; 7 June, M ⳱ 2.9|2.5, and 21 May, M ⳱ 2.6|2.8. Our local magnitudes showed good agreement with the catalog magnitudes, considering the differences in coverage among the different networks, and there is no apparent relation between magnitude and position. The Gutenberg–Richter distribution of our data (not shown) indicates that our sampling is approximately homogeneous for M ⱖ 0.8.

Hypocentral Distribution Figure 3 shows the epicenters of the local earthquakes, and Figure 5 shows two stereo-pairs on which the spatial distribution of local earthquake hypocenters may be clearly appreciated in relation to the fault traces that are plotted on the surface. The epicenters determined in this study have much smaller uncertainties than those determined by regional networks, and their spatial distribution agrees quite well with the concentrations and gaps featured in the regional microseismicity (Fig. 2). It can be seen in Figures 3 and 5 that hypocenters are not distributed uniformly over the study region or along seismogenic features, but present a high degree of clustering. Activity in the five main clusters (Table 2) occurs in spurts lasting a few hours with days of quiescence between them. Clusters can also be appreciated in Soares’ (1981) study of the northern part of the Ojos Negros valley, but they occur at different sites from the clusters in the present study, which suggests that although seismicity may concentrate in clusters, these are not permanent features. Three spatial groups (shaded areas in Figure 3) can be distinguished. The first group, SMN, corresponds to seismicity within a band about 6 to 8 km wide along the northwestern segment of the San Miguel fault trace (in the characteristic NW-SE direction of the main regional tectonic fault systems indicated in Figure 1). This seismicity is particularly dense for the northernmost third, with maximum depths ranging from 13 km in the NW to about 18 km near

Figure 5.

Stereo pairs showing the 3D distribution of hypocenters (circles) within the valley. The spatial effect can be seen, without a stereoscope, by holding the figure about 40 cm away from the eyes and focusing at infinity until the 3D image appears in the middle. The top pair is a view from directly above (azimuth 180⬚, elevation 90⬚), the bottom one is a view from azimuth 315⬚ and elevation 45⬚; the coordinates origin is at 31.6⬚N, 116.4⬚W, and zero depth; N, E, and Z axes are 50, 50, and 30 km long, respectively, and distance between ticks is 10 km.

Table 2 Main Hypocentral Clusters Cluster

Latitude

Longitude

Events

Depth (km)

Duration (days)

CL1 CL2 CL3 CL4 CL5

31.92⬚ 31.94⬚ 31.92⬚ 31.86⬚ 31.80⬚

116.30⬚ 116.23⬚ 116.18⬚ 116.11⬚ 116.18⬚

45 23 14 14 8

11–17; 5–7 13.5–17.5 15–18; 3.5–4.5 17–18 6.5–7.5

15 23 30 27 2

the center; a cluster below a mapped break of the fault trace and a couple of events close to it are quite shallow (⬍5 km), but over the rest of the fault minimum depths stay around 8 to 9 km. The second group, labeled CEV in Figure 3, corresponds to events scattered over the central and eastern parts of the region, between longitudes 116.05⬚W and 116.25⬚W. Most of the seismicity appears to deepen to the east and northeast of the southern part of the Ojos Negros fault and


the southeastern part of the Tres Hermanos fault which borders the valley, but is not, apparently, associated with these faults. A few earthquakes located NE of the point where the Ojos Negros and the Tres Hermanos faults meet, appear to deepen to the NE and roughly suggest a concave surface, while shallow seismicity near the center of the valley (the most numerous) also deepens to the NE to form a band or wedge dipping at around 20⬚ with deep events reaching the middle third of the San Miguel Fault. The third group is constituted by the CL1 seismic cluster (discussed in detail following), which comprises 45 located earthquakes, most of them at depths between 12 and 14 km. The entire seismicity associated with the Ojos Negros fault consists of the CL1 seismic cluster, and one other 13 km deep event. We found no activity above our detection threshold along the rest of the fault or along its inferred SE continuation down to the Tres Hermanos fault. Within the location uncertainties, hypocenters range from 3 to 24 km, with most (99%) events located between 3 and 20 km. The overall source depth distribution has a global maximum that begins at 12 km, peaks at 13 km, and extends to 17 km; there is a local maximum between 3 and 5 km. In between the maxima, depth is distributed approximately homogeneously with some eight earthquakes/km.

1139

three of the four fault plane solutions determined for events within the valley and away from its edges are normal, and the other one is strike-slip. Figure 7 shows the position of the preferred P and T axes for all fault plane solutions. The P axes essentially occupy an N-S band, while the T axes cover only the E and W “polar” sections. The overall distribution of the P and T axis azimuths ␾ are shown in Figure 8 (top); the bidirectional rose histograms are “distributed” by using narrow azimuth classes D␾ ⳱ 1⬚, and convolving each observation with a normal curve with r, to partially account for the low number of data and for possible errors in the azimuth determinations (Nava, 1998, 2000). The three largest maxima in the P distribution correspond to azimuths 176.5⬚, 159⬚, and 11.5⬚ (and/or their supplementary angles); peaks at other angles correspond mainly to axes near the vertical part of the aforementioned band. T-axis azimuths show two preferential azimuths at 65.5⬚ and 90⬚, and a definite gap for ⳮ55⬚ ⱕ ␾ ⱕ 40⬚. Figure 8 (bottom) shows the dispersed distribution of the P and T axis inclinations ␣, for D␣ ⳱ 1⬚ and r ⳱ 2D␣; P inclinations have maxima for 76.5⬚, 22.5⬚, and 88.5⬚, but are essentially distributed over the whole 0⬚–90⬚ range, while T inclinations show two close maxima at 75.5⬚ and 82.5⬚ and a wide gap for 0⬚ ⱕ ␣ ⱕ 60⬚.

Fault-Plane Solutions After thoroughly checking the polarity on each seismogram, each fault-plane solution was determined using a minimum of 10 P-wave first-arrival polarity readings; most determinations used between 11 and 13 readings. All polarity picks were twice made independently on screen. Paper hardcopies at several amplifications were used to check the polarities and the qualitative agreement of relative amplitudes of P and S first arrivals on three components with the requirements of a theoretical double-couple radiation pattern, both near the P and T nodes (maximum P/S ratio) and near the P nodal planes (minimum P/S ratio). Fault-plane solutions were considered adequate when the solution, determined using the FPFIT program (Reasenberg and Oppenheimer, 1985), was unique, and the uncertainty, quantified as the relative area covered by the possible variations of the P and T axes given by FPFIT (Nava and Frez, 1999), was generally less than 16% of the focal sphere. Ambiguous solutions, usually caused by lack of stations very close to the epicenter, were discarded. We were able to determine adequate fault-plane solutions, shown in Figure 6, for 50 of the located events (21 of them from the CL1 cluster). Most of the focal mechanisms are predominantly strikeslip with a normal component, and there is only one reverse mechanism (Fig. 6). Strike slip predominates for earthquakes associated with most of the northern San Miguel fault, the NW tip of the Ojos Negros valley and the CL1 cluster, normal mechanisms predominate for earthquakes located near the discontinuity in the San Miguel fault around 31.95⬚N;

Activity in a Seismic Cluster The CL1 seismic episode occurred between days 970605 and 970620, with earthquakes concentrated within 31.917⬚ Ⳳ 0.017⬚N and 116.300⬚ Ⳳ 0.017⬚W, from which we determined 45 hypocenters, most of them at 13 Ⳳ 0.6 km depth, and 21 fault plane solutions. Focal depth remained constant between 12 and 14 km (essentially the same depth, considering the uncertainty in depth determination) for the 34 located events that occurred during 5 June to 11 June. After three days with no activity, on 14 June there were six earthquakes at depths between 15 and 17 km; between 15 June and 16 June focal depths came back to the “normal” 13 km for three events. Finally, the episode ended with two events that occurred June 19 and 20 with identical epicenters and depths of 4.4 and 4.9 km, respectively, the last event being the largest one with magnitude 2.1|1.9 (all other events in CL1 have magnitudes between 1 and 2). Instruments remained operating for four more days but recorded no further activity from this cluster. Most CL1 events are predominantly strike-slip, with a minor normal component, and their T axes point mainly EW. The largest event has a normal mechanism; one of the nodal planes is almost horizontal while the vertical one coincides with the general orientation of the Ojos Negros fault and the corresponding slip would be downwards toward the valley. It might be argued that the relatively large amount of data (depths and fault plane solutions) from CL1 could be biasing the overall regional results, but regional results do not change significantly if the CL1 data are removed.

1140


Figure 6.

Map showing fault plane solutions (upper hemisphere projection) connected by a dashed line to their corresponding epicenter. Black indicates compression and white dilation. The number below each focal sphere is the focal depth in kilometers.

Conclusions and Discussion Based on the results presented previously and considering the hitherto unparalleled (for this region) quality of our data and determinations, we conclude: 1. The main faults associated with the Ojos Negros region appear to be acting in quite different ways. The San Miguel fault is associated with dense, predominantly strikeslip, seismicity in the SMN region (Fig. 3), in agreement with field evidence that this fault is vertical with strikeslip motion (Shor and Roberts, 1958; Hirabayashi et al., 1996); this activity is continued to the SE by some 7 km of mainly normal mechanisms, followed by deeper seis-

micity (no shallow earthquakes), again predominantly strike-slip, below its middle third. The Ojos Negros fault, which appears to have normal slip along a plane dipping some 50⬚–60⬚E (F. Sua´rez, personal communication), is quiet, except for the predominantly strike-slip CL1 cluster and one isolated strike-slip earthquake some 5 km to the SE of the cluster. The Tres Hermanos fault has no earthquakes along its northern part, but along the S-SE part that borders the valley, but the seismicity that occurs to the NE, deepening in that direction, might be associated with this fault. Unfortunately, it was not possible to determine whether this activity corresponds to shallow-plane listric slip along the


Figure 7.

P and T axes distribution on the upper hemisphere of an equal area projection of the focal sphere.

1141

The hypocentral distribution suggests an aseismic lower crust below 20 km. 3. There is seismicity within the region that does not appear to be related to any of the main surface faults in the area (although part of it might be associated with the aforementioned possible listric faulting along the Tres Hermanos fault). This seismicity appears to deepen to the NNE and constitute a band with this orientation that goes all the way to the San Miguel fault. Of four fault plane solutions for events within this band, three are normal and one is strike-slip with a NW or NE fault plane; again, we lack enough solutions to determine the nature of the tectonic process causing this seismicity. 4. The P axes cover a middle, roughly NS, band of the focal sphere (Fig. 7), which implies a narrow range for azimuths, and a very broad one for inclinations (Fig. 8). The T axes occur only in the E and W “polar” regions of the focal sphere (Fig. 7), predominantly in two narrow azimuthal ranges and steep inclinations (Fig. 8). This behavior is consistent with tensional stress acting in a roughly EW or ENE-WSW mainly horizontal direction, while pressure appears to be perpendicular to the tension but, for the small magnitude earthquakes considered here, without any other preferential orientation, and indicates that the main tectonic stress acting over this area is tensional. E-W tension, together with N-S compression, checks with the predominant NW-SE regional orientation of transform faults and spreading centers related to the NW transcurrent motion along the San Andreas–Gulf of California fault system.

Acknowledgments

Figure 8. Histograms showing (top) the u azimuthal distribution, and (bottom) the ␣ inclination distribution of the P and T axes.

Tres Hermanos fault, as part of the tectonic fracturing within the valley (discussed below). Field observations indicate this fault is vertical with strike-slip motion (F. Sua´rez, personal communication). 2. The seismogenic zone in the Ojos Negros region is found between 3 and 20 km, with maximum activity between 12 and 17 km, and a secondary maximum between 3 and 5 km. The hypocentral depth range, larger than those usually reported for a strictly continental crust for the San Andreas fault system, is similar to that beneath the Transverse Ranges (Hill et al., 1990; Bryant and Jones, 1992).

Our recognition to Flor Carrillo, Francisco Farfań, Luis Orozco, and Oscar Ga´lvez for their participation in the field work, to Jorge Jasso for reading times, to Jose´ Manuel Romo (at the time, Director of the Earth Sciences Division, CICESE) for institutional support, to Hubert Fabriol for the shared utilization of the instruments. Many thanks to Francisco Sua´rez and Luis Delgado for valuable geological comments. Our thanks to IRIS/ PASSCAL for the loan of the Reftek instruments and ancillary equipment, and for technical assistance. We are grateful to SCSN (USGS-CalTech) and RESNOM for use of their data. This project was partially supported by CONACYT (25-309-T) and CICESE. We are grateful to two anonymous referees for pertinent, constructive, and useful comments.

References Brehm, D. J., and C. W. Braile (1999). Intermediate-term earthquake prediction using the modified time-to-failure method in southern California, Bull. Seism. Soc. Am. 89, 275–293. Brune, J. N., R. S. Simons, C. Rebollar, and A. Reyes (1979). Seismicity and faulting in northern Baja California, in Earthquakes and Other Perils, San Diego Region, P. L. Abbott and W. J. Elliott (Editors), San Diego Association of Geologists, San Diego, California, 83–100. Bryant, A., and L. Jones (1992). Anomalous deep crustal earthquakes in the Ventura Basin, southern California, J. Geophys. Res. 97, 437– 447.

1142


Camacho, J. S. (1989). Estudio de la estructura somera (0–250 m) en un sector de la falla San Miguel, M.Sc. Thesis, Centro de Investigacioń Cientifica y de Educacioń Superior de Ensenada, Ensenada, B.C., Me´xico, 75 pp. DeMets, C., R. G. Gordon, S. Stein, and D. F. Argus (1987). A revised estimate of Pacific-North America motion and implications for western North America plate boundary zone tectonics, Geophys. Res. Lett. 14, 911–914. Doser, D. (1994). Contrasts between source parameters of M ⱖ 5.5 earthquakes in northern Baja California and Southern California. Geophys. J. Int. 116, 605–617. Elders, W. A., R. W. Rex, T. Meidav, P. T. Robinson, and S. Biehler (1972). Crustal spreading in southern California, Science 178, 16–23. Frez, J., and J. J. Gonza´lez (1991). Crustal structure and seismotectonic of northern Baja California, in The Gulf and Peninsular Province of the Californias, B. Simoneit and J. P. Dauphin (Editors), American Association of Petroleum Geologists Memoir 47, Tulsa, Oklahoma, EEUU, 261–283. Frez, J., and V. M. Frıás-Camacho (1998). Mapas anuales de sismicidad para la regioń fronteriza de ambas California, GEOS 18, 112–130. Gonza´lez, J. J., R. Bennett, R. King, S. McClusky, and R. Reilinger (2000). GPS constraints on present-day plate boundary deformation in southern California, USA, and northern Baja California, Mexico. Geophys. Res. Lett., submitted. Gonza´lez, M. (1987). Estudio detallado del sismo de Pino Solo, Baja California, Me´xico, del 8 de mayo de 1985, M.Sc. Thesis, CICESE, Ensenada, Me´xico, 106 pp. Hadley, D. R., and H. Kanamori (1979). Regional S-wave structure for southern California from the analysis of teleseismic Rayleigh waves, Geophys. J. Roy. Astr. Soc. 58, 655–666. Hill, D. P., J. Eaton, and L. Jones (1990). Seismicity of the San Andreas fault system: 1980–1986, in The San Andreas Fault System, California, R. Wallace (Editor), U.S. Geol. Surv. Paper 1515, 1015–1151. Hinojosa, A. (1990). Procesado de archivos sı´smicos RESNOR en estaciones de trabajo SUN, Reporte Tećnico Interno, CIST9001, CICESE, Ensenada, B.C., Me´xico, 25 pp. Hirabayashi, K., T. K. Rockwell, S. G. Wesnousky, M. W. Stirling, and F. Suarez-Vidal (1996). A neotectonic study of the San MiguelVallecitos Fault, Baja California, Mexico, Bull. Seism. Soc. Am. 86, 1770–1783. Johnson, T. L., J. Madrid, and T. Koczynski (1976). A study of microseismicity in northern Baja California, Mexico, Bull. Seism. Soc. Am. 66, 1921–1929. Kern, H. (1982). P- and S-wave velocities in crustal and mantle rocks under the simultaneous action of high confining pressure and high temperature and the effect of the rock microstructure, in High-Pressure Researches in Geoscience, W. Schreyer (Editor), E. Schweitzerbart’sche Verlagsbuchhandlung, Stutgart, pp. 15–45. Lee, W. H., and J. C. Lahr (1975). HYPO71 (revised): a computer program for determining hypocenter, magnitude, and first motion pattern of local earthquakes, U.S. Geol. Surv. Open-File Rept. 75–311, 114 pp. Leeds, A. L. (1979). Relocation of mh ⬎ 5.0 Northern Baja California earthquakes using S-P times, M. S. thesis, University of California San Diego, La Jolla, California, 101 pp. Lomnitz, C., F. Mooser, C. R. Allen, J. N. Brune, and W. Thatcher (1970). Seismicity and tectonics of the northern Gulf of California region, Mexico: preliminary results, Geof. Int. 10, 37–48. Medina, M., and C. Duarte (1979). Red sismolo´gica de telemedicioń digital del campo geote´rmico de Cerro Prieto, Mexicali, Second Symposium on the Cerro Prieto Geothermal Field proceedings, Comisioń Federal de Electricidad, Mexicali, Me´xico, 355–368. Nava, F. A., and J. N. Brune (1982). An earthquake-explosion reversed

refraction line in the Peninsular Ranges of Southern California and Baja California Norte, Bull. Seism. Soc. Am. 72, 1195–1206. Nava, F. A., and J. N. Brune (1983). Source mechanism and surface wave excitation for two earthquakes in northern Baja California, Geophys. J. Roy. Astr. Soc. 73, 739–763. Nava, F. A. (1998). Histograms repartidos para el ana´lisis de muestras estadı´sticas escasas, GEOS 18, 174–179. Nava, F. A. (2000). Histogramas repartidos para el ana´lisis de muestras estadisticas escasas II: Rosas y Mapas, GEOS, submitted. Nava, F. A., and J. Frez (1999). Cuantificacioń de la incertidumbre en las soluciones de plano de falla sismica obtenidas mediante el programa FPFIT, GEOS 19, 175–178. Nordquist, J. M. (1964). A catalogue of southern California earthquakes and associated electronic data processing programs, Bull. Seism. Soc. Am. 54, 1003–1011. Norris, R., C. Johnson, L. Jones, and K. Hutton (1986). The southern California Network Bulletin, U.S. Geol. Surv. Open-File Rept. 86-96, 32 pp. Reasenberg, P., and D. Oppenheimer (1985). FPFIT, FPPLOT and FPPAGE: Fortran computer programs for calculating and displaying earthquake fault-plane solutions, U.S. Geol. Surv. Open-File Rept. 85-739, 46 pp. Rebollar, C., and M. S. Reichle (1987). Analysis of the seismicity detected in 1982–1983 in the Northern Peninsular Ranges of Baja California, Bull. Seism. Soc. Am. 77, 173–183. Reyes, A., J. Brune, T. Barker, L. Canales, J. Madrid, J. Rebollar, and L. Munguıá (1975). A microearthquake survey of the San Miguel fault zone, Baja California, Mexico, Geophys. Res. Lett. 2, 56–59. Richter, C. F. (1958). Elementary Seismology, W. H. Freeman & Co., San Francisco, 768 pp. Shen, Z., D. Dong, T. Herring, K. Hudnut, D. Jackson, R. S. McClusky, and L. Sung (1997). Crustal deformation measured in southern California, EOS 78, 477–482. Shor, R. R., and E. Roberts (1958). San Miguel Baja California Norte earthquakes of February 1956, A field report, Bull. Seism. Soc. Am. 48, 101–116. Soares, J. J. (1981). Estudio de microsismicidad a lo largo de dos sectores de la falla de San Miguel, B.Sc. Thesis, Universidad Autońoma de Baja California, Ensenada, B.C., Mexico, 60 pp. Sua´rez-Vidal, F., R. Armijo, G. Morgan, P. Bodin and R. G. Gastil (1991). Framework of recent and active faulting in Northern Baja California, in The Gulf and Peninsular Province of the California, B. Simoneit and J. P. Dauphin (Editors), American Association of Petroleum Geologists Memoir 47, American Association of Petroleum Geologists, Tulsa, Oklahoma, 285–300. Vidal, A., and L. Munguıá (1999). The ML scale in northern Baja California, Mexico, Bull. Seism. Soc. Am. 89, 750–763. CICESE Seismology Dept. Km. 107 Carretera Tijuana-Ensenada Ensenada, B.C. 22860, Mexico Email: [email protected] (J. F., J. J. G., J. G. A., F. A. N., I. M., J. C., R. E. G.-A.) IRIS/PASSCAL Instrument Center New Mexico Tech 801 Levoy Place Socorro, New Mexico 87801 (M. A.) Manuscript received 16 December 1999.