1. Introduction

The ANZA seismic network was installed to monitor micro-earthquakes occuring near the San Jacinto fault zone near the town of Anza in Riverside County, California. This network was installed to study the scaling laws of seismic body-wave spectra, the character of high frequency ground motion, the physical interpretation of seismic stress drops, and the interaction of earthquakes. The topic of this thesis is the the analysis of earthquake data recorded by this network and the development of new spectral analysis techniques to study these data. Of particular interest are the spectra, coherence, and polarization of seismic wavefields and the influence of site effects on these measurements.

The remainder of this chapter will discuss the structure and the historical seismicity of the San Jacinto fault zone. The design and system characteristics of the ANZA seismic network are covered in Chapter II. In the third chapter the distributions of earthquake hypocenters and source parameters recorded by the ANZA seismic network are investigated. Chapter IV presents the application of multiple taper spectral analysis to seismic data. The development of multitaper coherence techniques and the study of the coherence of data recorded on a small aperture seismic array is the subject of Chapter V. The last chapter presents the derivation of a method to determine the polarization of a seismic wave as a function of frequency from three component seismic data.

2. Structure Of The San Jacinto Fault Zone

The San Jacinto fault zone is one of the major components of the San Andreas system in southern California. It originates in Cajon Pass where it splits away from the main trace of the San Andreas fault along a strike of approximately S40E. The fault crosses through the San Bernardino and San Jacinto Valleys, makes a transect through the Peninsular Range batholith, and terminates in the westernmost Imperial Valley near the Mexican Border (Figure 1).

The San Jacinto fault does not have a continuous linear surface trace. Instead there can be from one to three subparallel surface strands at various places along the zone which vary in length from 30 to 70 km. These strands can form complicated substructures within the general linear trend of the whole fault zone which can make the identification of the active strand or strands difficult. The first geological study of the fault structure was by gLawsong [1908]. Since then, various names have been used by different authors for these strands. In this thesis the term "San Jacinto" will be used in reference to the whole fault zone, following the usage of gSharpg [1972]. Each strand will be identified by a different name, as shown in Figure 1.

The northwestern part of the San Jacinto fault zone, which is known as the Claremont strand, starts in Cajon Pass, crosses the San Bernardino Valley, continues over the hills southwest of Loma Linda and into the San Jacinto Valley. In the San Jacinto Valley there are two active strands, the Claremont which runs along the northeast edge, and the Casa Loma which is parallel several kilometers to the southwest. The Casa Loma fault can be traced to the southeast end of the San Jacinto Valley near Hemet. The Hot Springs fault splays from the Claremont fault on the northeast side of the San Jacinto Valley and terminates near Lake Hemet.

From Hemet to the town of Anza the fault zone has only one active strand, the Clark, which is aligned with the Casa Loma. It is possible that these two faults are actually one strand because the gap between them is culturally disturbed on the surface (R. V. Sharp, personal communication, 1987). Just east of the town of Anza, the San Jacinto trifurcates with two new strands starting, the Buck Ridge and the Coyote Creek. The Buck Ridge is the northeast strand and can be mapped as far as the south flank of Toro Peak. The middle strand is the continuation of the Clark Fault which can be followed to the San Felipe Hills near the Salton Sea. The Coyote Creek fault is the other strand and is traced southeast through Ocotillo Wells. Southeast of the termination of the Coyote Creek strand the southern end of the San Jacinto fault zone appears to be defined by the parallel Superstition Hills and Superstition Mountain faults.

3. Geologic Slip Rates

The total displacement across the San Jacinto fault zone has been determined by examining offsets of contacts or distinctive rock types in the batholithic rocks. gSharpg [1967] measured a total right lateral slip of 24 km for the San Jacinto fault (Clark strand) and the Thomas Mountain fault. gHillg [1984] was able to add another 4.5 km of slip from the Hot Springs fault and .7 from an unnamed fault to bring the total offset to about 29 km. Faulting probably initiated approximately 4 million years ago which leads to an estimate of the average slip rate of 7 mm/yr [gHillg, 1984]. Sharp [1981] places a minimum bound of 8 to 12 mm/yr measuring the displacement of a 730,000 year old Bishop Tuff ash bed. On the flank of Thomas Mountain, which is north of the town of Anza, a slip rate of 91 mm/yr over the period of the last 9000 years has been determined by gRockwell et alg [1986].

The last two measurements were made along the section of the Clark strand where the San Jacinto fault zone has no other parallel strands. There are estimates of the slip rate for the Claremont strand of 5.41 mm/yr for the past 2000 years by gWesnousky et alg. [1987] and for the Coyote Creek fault of 3 to 5 mm/yr for the last 400 years [gSharpg, 1981] . These values, while valid for each particular strand, may not reflect the total slip rate across the whole San Jacinto fault zone. The former measurement might be biased by the complicated structure of the fault in the area and the proximity of the San Andreas fault, while the latter does not account for slip on the Clark or Buck Ridge faults southeast of the trifurcation of the San Jacinto fault zone [gWorking Group on California Earthquake Probabilities (WGCEP)g, 1988].

Near the town of Anza the slip rate determinations are remarkably consistent for several time scales at a value of about 10 mm/yr. gKing and Savageg [1983], using a trilateration network which spans the San Jacinto fault, found a strain rate maximum centered over the fault zone with a value that is compatible with the geological slip rates.

4. Historical Earthquakes

The San Jacinto fault zone is one of the most seismically active in California for moderate to large earthquakes and also for microearthquakes. Figure 2 shows the epicenters and the rupture areas for the largest known earthquakes (M ≥ 5.5) associated with this fault. The catalog of instrumental data from earthquakes on the San Jacinto fault has been recently extended back to 1899 by reexamination of archived records. The continued development and deployment of seismic instrumentation in southern California through the present has lowered the minimum magnitude of events which can be located. A nearly complete catalog of magnitude greater than 2.0 events is available starting about 1975. The distribution of seismicity is not spatially uniform along the fault in any magnitude range. The large events appear to have ruptured some parts of the fault but not others. The microseismicity appears to be organized into clusters which are separated by large regions with occasional events.

The study of the seismicity of this fault zone is based on three types of information: paleoseismic, historical, and instrumental. Paleoseismic research consists of the identification and dating of specific surface slip events using geological evidence from a fault trace. Until recently, there has not been any paleoseismic data for prehistoric events associated with this fault. Historical data consists of reports of felt earthquakes which are generally found in newspapers, weather bureau information, personal letters, or other archive materials. These data are useful for determining the approximate time of events, but do not lead to the reliable estimation of magnitudes or locations due to effects such as population distribution which can severely bias results. In southern California the population density, particularly in the coastal areas, increased to where the historical records begin to be useful in about 1850. The best information for the evaluation of earthquakes is data recorded on distant and local seismometers. The former provide excellent evidence the size of the event while the later allow it to be precisely located. California during 1887 at Mt. Hamilton (MHC) and Berkeley (BRK) [gLouderbackg, 1942], and recording at local sites in southern California started in the 1920's.

4.1. Paleoseismic Data

The first paleoseismic study of the San Jacinto fault zone was conducted by gClark et al.g [1972] who studied trenches across the surface rupture of the 1968 Borrego Mountain magnitude ML &sim 6.8 earthquake. They estimated a recurrence interval for this size event to be approximately 195 years along the Coyote Creek fault.

A paleoseismic study of the Clark fault is being conducted near the town of Anza by gKlinger and Rockwellg [1989]. Their site is at Hog Lake, which is located on the trace of the Clark strand on the south flank of Thomas Mountain near the town of Anza. This site is located near the northwest terminus of the Anza seismic slip gap [gThatcher et al.g, 1975]. The Clark strand is the only active strand along this particular section of the San Jacinto fault zone, removing the possibility of slip on parallel strands. Hog Lake exists because a Holocene alluvial fan forms a small dam across a drainage channel caused by the fault trace. Runoff from Thomas Mountain fills the lake intermittently and the resulting sedimentation creates a well-stratified section. gKlinger and Rockwellg [1989] used carbon-14 dating techniques on peats and other organic detritus to determine the ages of different strata. They observed offsets of these dated strata to estimate a recurrence interval of approximately 250 years for slip events in the sediments of Hog Lake.

gKlingerg [1989] has been able to identify four slip events which have occurred since 1100 A.D. The most recent slip event caused a small observable vertical offset in a 180-year-old Hog Lake sediment. This event does not have a minimum age and might be associated with the 1899 or the 1918 earthquakes. The next earlier slip event appears to be constrained by sedimentary layers which were deposited about 1669 and 1776. It also shows a small amount of vertical offset. The slip event which disturbed the Hog Lake sediments with a large amount of vertical offset appears to have occurred between 1394 and 1496. There is evidence for another small amount of offset which occurred prior to 1200 A.D. Unfortunately, these measurements do not constrain the amount of strike slip motion caused by each event.

The vertical offsets may however serve as a relative scale for the slip caused by the various events. In the last 100 years the San Jacinto fault zone has produced more moderate to large earthquakes (M > 6.0) than any other fault in southern California. Table 1 gives a listing of the most significant right-lateral events along this fault. The data presented are a compilation of the most recent size and location estimates combined from various sources. The earthquakes can be sorted into four different sets. The first contains two events (1890, 1892) and uses intensity data for location and magnitude estimates. The second uses intensity data for location and instrument data for magnitude estimation and contains four events (July and December 1899, 1918, 1923). The next set uses instrumental data for calculating the epicenters and magnitudes and consists of five events between 1937 and 1969. The last two events (1980, 1987) happened after the densification of the southern California se

4.2. 1890 and 1892 Earthquakes

Studies of the historical data [gTownley and Alleng, 1939; gToppozada et al.g, 1981] show that since the mid-1800's the first large events which might be associated with the San Jacinto fault zone occurred on February 9, 1890, and May 28, 1892. Utilizing newspaper reports of the intensity of the ground motion, gToppozada et al.g [1981] decided that these temblors should be located near the San Jacinto fault zone and be assigned a magnitude of MI = 6.3. gHanks et al.g [1975], using intensity data from Townley and Alleng [1939], assigned a moment of 15 x 1018 N-m to the 1890 event.

These events illustrate the difficulty in assigning magnitudes and locations from sparse intensity data. Both of these events occurred in the early morning (04:06 PST for 1890, 03:15 PST for 1892) and were felt in parts of the Los Angeles basin, as well as in Riverside, San Diego, and Yuma [gToppozada et al.g, 1981]. The places where the strongest shaking was felt for the 1890 event were the towns of Colton and Chino. The designation of 33.4N and 116.3W as the epicenter is strongly influenced by the newspaper report in the Los Angeles Herald on February 10 which states:

Communications by wire from all the towns on the line of the Southern Pacific, between here and Yuma, convey the information that the shock was felt with equal severity in each, as far as the Colorado River.

This report yields little substantive information about which towns the earthquake was felt in, nor about the strength of shaking in those towns. By 1880 the Southern Pacific railroad was providing service between Yuma and Los Angeles; however, the Coachella and Imperial Valleys remained unsettled until about 1900. To get more information about this event it will be necessary to find out where the telegraphers were stationed along the railroad. Recently, Duncan Agnew (personal communication, 1989) found a report in the San Diego Union on March 22, 1890, which records that springs started as a result of the shaking on the Thompson ranch in the Wilson and Lancaster Valleys. These valleys are located northwest of the settlement of Aguanga and are equidistant from the Elsinore and the San Jacinto faults. A search of the weather bureau reports yields the curious information that the earthquake was quite severe in Los Angeles, light in Riverside, and not reported in Julian, San Diego, and Yuma (Duncan Agnew, personal communication, 1989). Reports on the 1892 event are similar to those from 1890 except that the weather bureau files from Julian indicate a severe shock was felt, while San Diego noticed two light shocks, and Yuma mentioned a slight tremor at 06:14 (???).

The historical intensity data from southern California does not put many constraints on the magnitudes or the locations of the 1890 and the 1892 mainshocks, but it is reasonable to assume that they have a magnitude of at least 5.5 and less than 7.0 and are located somewhere in or near the Peninsular Ranges. To date, no felt reports have been found for either event from Escondido, or Carrizo and Laguna Stations (near the Mexican border), which reported moderate to strong shaking during the February 24, 1892, Laguna Salada earthquake. If the new information is considered, the May 1892 event could be located near Julian and is probably the southernmost of the two. One significant finding from the historical record is that since 1850, neither the San Jacinto nor the Elsinore faults has produced a great (M ≥ 8) earthquake in southern California such as the 1857 Fort Tejon and the 1872 Owen's Valley earthquakes.

4.3. 1899, 1918, 1923 Earthquakes

The first known instrumental recordings of earthquakes located near the San Jacinto fault zone were made in 1899. These events were measured on undamped Milne seismometers, the closest of which were located in Victoria, British Columbia, and Toronto, Ontario. The first event (20:32 GMT, July 22) was strongly felt in the Cajon Pass area and could be associated with either the San Andreas or the San Jacinto fault [gHanks et al.g, 1975; Sanders, 1986]. Toppozada et al.g [1981], using intensity data, assigned a magnitude of 6.5 to this event while gHanks et al.g [1975] assigned a moment of 4 x 1018 N-m based on the intensity reports of gTownley and Alleng [1939]. gAbe [1988] was able to calculate a surface wave magnitude from the Milne seismograms of MS = 5.6 which suggests that the previous size estimates for this event are biased high. This estimate of magnitude was made from recordings at only one or two stations which is an indication of the size. A

The Christmas day, 1899, and the April 21, 1918, earthquakes are the two most significant events along the San Jacinto fault between Cajon Pass and the town of Anza. Both of these events caused extensive damage to the towns of Hemet and San Jacinto and were felt as far away as Santa Barbara and Needles [gTownley and Alleng, 1939; Toppozada et al.g, 1981; gToppozada et al., 1982]. The 1918 event produced the strongest reported ground shaking in the Anza valley since the first settlers homesteaded that area in the early 1900's (Emily Lauridsen, personal communication, 1985). There is no conclusive evidence of surface rupture for either event, although the main trace of the fault was searched from the town of San Jacinto to Burnt Valley (approximately 40 km southeast) after the 1918 earthquake [gTownleyg, 1918; Rolfe and Strong, 1918].

R. E. Klinger (personal communication, 1989) suggests that the small offset found in the 180-year-old Hog Lake sediments may have been a result of one of these events which was missed by the early observers. The two epicenters are effectively colocated and assigned similar magnitudes, 6.7 and 6.8 respectively, due to the high correlation between the two sets of intensity reports. These earthquakes could have occurred on either the Casa Loma or the Claremont strand. gHanks et alg. [1975] estimated a moment of 15 x 1018 N-m for the 1918 event from a sparse set of instrumental data and the intensity information. They inferred a similar value for the 1899 earthquake based on the distribution of intensities. Recently, a surface wave magnitude of 6.4 was determined from seismograms recorded on the Milne seismometers for the 1899 shock [gAbeg, 1988]. This new datum lowers the estimate of the seismic moment to 4 x 1018 N-m [gWGCEPg, 1988].

The last event of this set occurred on July 23, 1923 and was located near Loma Linda on the Claremont fault by gSandersg [1986] who used intensity data along with P and S wave arrival times recorded at Pasadena. A magnitude of 6 1/4 was given to this event by gRichterg [1958] who calibrated regional recordings of this event to later events [gSandersg, 1986].

These four events are the largest known events along the San Jacinto fault zone northwest of Anza since 1893, and possibly since before 1850.

4.4. 1937, 1942, 1954, 1968, 1969 Events

The era of instrumental data acquisition started in southern California in 1926 with the installation of the first seismic station at Riverside on October 19. By 1930, the southern California-network consisted of seven stations: Haiwee, La Jolla, Mount Wilson, Pasadena, Riverside, Santa Barbara and Tinnemaha. These new stations allowed local earthquakes to be located with errors of approximately ten kilometers with little depth control. This resolution would not improve significantly for the next 43 years until the installation of denser seismic networks. The seismic information from the new stations also allowed better estimates of earthquake magnitudes, particularly when coupled with data from more distant stations. One of the most important contributions of this period of seismic recording is the generation of a complete catalog of events and seismograms. This allows these events to be reanalyzed and compared to new ones as more sophisticated instruments are developed and installed.

The March 19, 1937 earthquake was the first large event located along the San Jacinto fault zone by using instrumental data. gWoodg [1937] examined P wave arrivals at seven seismic stations in southern California to deduce an epicenter of 33.46N 116.58W which is near the trifurcation of the San Jacinto fault. This event is the first in a set of five with magnitudes of 5.8 or larger in the next 32 years. These events all had epicenters between the trifurcation of the San Jacinto fault zone and the northern end of the Superstition Hills fault and have been the subject of numerous studies (see gSandersg [1986] for references). gSandersg [1986] made a careful study of these sources, using information from more recent events and an improved velocity structure, to provide better locations and magnitude estimates. These results are used in Table 1.

The 1937 event was named the Terwilliger Valley earthquake by gWoodg [1937] even though his location was in the adjacent Tule Canyon. Terwilliger Valley was chosen because it was the closest inhabited area, (a somewhat shortsighted naming convention considering that the next closest settlement to the southeast is Borrego Springs, about 20 km away). The relocation of the 1937 event places it underneath Buck Ridge which is between the surface traces of the Buck Ridge and the Clark faults. It has a revised magnitude of 5.9 and a moment of 0.3 x 1018 N-m. The October 21, 1942, magnitude ML &sim 6.3 mainshock has been relocated near the Fish Creek Mountains to the southeast of the Superstition Mountain fault and is apparently not situated in the San Jacinto fault zone [gSandersg, 1986]. The Arroyo Salada earthquake ended the next 11 1/2 year quiescence of large temblors on March 19, 1954. It weighed in at 6.2 on the Richter scale and was located at the southeast end of the Clar

The last pair of events of this set are the magnitude 6.8 Borrego Mountain mainshock of April 9, 1968, and the April 28, 1969, Coyote Mountain aftershock with a magnitude of 5.8. The Borrego Mountain event produced a surface rupture over 30 km in extent along the Coyote Creek strand, and the results of many scientific investigations related to this event were published in USGS Professional Paper No. 787. The aftershock in 1969 actually was located approximately 10 km to the northwest of the limits of the other aftershocks of the 1968 mainshock. This set of five events shares one other feature in common besides the instrumentation which were used to record them. All of these events are located to the southeast of the trifurcation of the San Jacinto fault zone and each one, with the possible exception of the 1942 event, is apparently related to one of the three trifurcated strands. The 1954 Arroyo Salada earthquake is associated with the Clark fault [gSandersg, 1986], while the Coyote Creek strand is the causative fault for the 1968 and 1969 events. The Terwilliger Valley probably was on the Clark fault or could have been on the Buck Ridge strand.

4.5. 1980 and 1987 Events

The two most recent moderate to large events connected with the San Jacinto fault zone are the magnitude ML = 5.5 Whitewash earthquake at 10:47 GMT on February 25, 1980, and the MS = 6.6 Superstition Hills mainshock at 13:15 GMT on November 24, 1987. These events have epicenters at the opposite ends of the southeastern half of the San Jacinto fault zone. They are the largest events along the whole fault which have reliable depth control, since they occurred after the upgrade of the southern California seismic network in 1975, and were also recorded by local strong motion instruments.

Investigators of the Whitewash earthquake have made estimates of its moment from strong motion records of approximately 0.025 x 1018 N-m [gFrankelg, 1984; Fletcher et al., 1987] and of 0.056 x 1018 N-m from Love waves [gFrankelg, 1984]. A small source dimension (less than one kilometer) was determined from the strong motion data. This event was near the northwest edge of the 43 year old aftershock zone for the Terwilliger Valley at a depth of 13.5 km. The origin time of the Whitewash earthquake was early in the morning and this event awakened people who were as far away as San Diego (personal experience, 02:47 PST).

The 1987 Superstition Hills right-lateral earthquake was preceded by the MS = 6.2 Elmore Ranch foreshock about twelve hours before on a left lateral conjugate fault. The surface wave magnitudes are significantly higher than the local magnitude estimates of 5.8 and 6.0 for the foreshock and mainshock respectively. The main shock moment for the Superstition Hills earthquake was estimated to be 8.25 x 1018 N-m [gBent et alg., 1988]. This is the second event southeast of Anza in the San Jacinto fault zone with an observed surface rupture. The surface rupture was traced for approximately 28 km with a maximum observed coseismic slip of about 80 cm [gBudding and Sharpg, 1988].

5. Seismic Slip, Seismic Gaps And Earthquake Probabilities

The expected locations and magnitudes of future large earthquakes along the San Jacinto fault zone have important implications for seismic risk assessment in southern California. All studies of this topic are limited by the available data at the time of each study. Since the historical record is not long enough to examine recurrence intervals along the various strands of the fault, the probable locations of the next large events on the fault must be inferred from the geologic slip rates and the estimated seismic moments and locations of known earthquakes. The largest earthquakes examined above, with the possible exception of the December 25, 1899, and the 1918 events, have each ruptured different sections of the San Jacinto fault zone. Due to the lack of local instrumental records for those two events, it is not possible to prove conclusively whether they have overlapping or adjacent source areas. Estimates for some of the older events have recently been improved by recovering unused instrumental data or by calibrating the original records to recent events.

An estimate for the average slip distributed over the whole fault zone can be made by summing the seismic moments of the measured earthquakes. gBruneg [1968] determined total moment of 1.4 x 1020 N-m summed from the known sources between 1912 and 1963. A seismic slip rate of 1.5 cm/yr was calculated when this total moment was distributed over a 280 km x 20 km area [gBruneg, 1968]. gThatcher et alg. [1975] calculated a rate of 0.8 cm/yr for a 240 x 15 km2 fault zone for events between 1890 and 1973. If only moments from seismic records are used and distributed over the 240 x 15 km2 area, then the seismic slip rate for 1899 to 1988 is 0.5 cm/yr.

The first study of the distribution of slip on the San Jacinto fault was by gThatcher et alg. [1975] who used seismic moment estimates of the nine largest earthquakes [gHanks et al.g, 1975]. gThatcher et alg. [1975] observed that seismic slip between 1890 and 1973 occurred in two sections of the fault zone and identified two corresponding seismic slip gaps. On the northwestern part of the fault an average slip of about 130 cm was determined for a 75 x 15 km2 area of fault from the 1890, December 1899, 1918, and 1923 events. This area extends from just north of Riverside to near Anza. One assumption used in this analysis was that the 1890 and 1899 events had moments similar to the 1918 earthquake (15 x 1018 N-m) based on intensity data. At the southeastern end of the fault, an average slip of 80 cm between Coyote and Superstition Mountains was calculated for a 75 x 15 km2 area. All the moments in the southeast area were determined from instrumental data.

The important observation from gThatcher et alg. [1975] was the identification of two seismic slip gaps, one between Anza and Coyote Mountain while the other is between Cajon Pass and Riverside. The Anza gap may be the more significant of the two since in the past 100 years large amounts of slip have accumulated on either side of it while the slip near the Cajon gap may be absorbed by motion on the adjacent San Andreas.

When the new data from gAbeg [1988] and more recent events are considered, the seismic gaps of gThatcher et alg. [1975] still exist. However, the distribution of the amount of slip along the San Jacinto fault zone does change appreciably. The reduction in magnitude of the Christmas 1899 event from 7 used by gThatcher et alg. [1975] to 6.4 by using the Milne seismograms [gAbeg, 1988] lowers the average slip in the northwestern slip region. The validity of the 7 magnitude which gThatcher et alg. [1975] used for the 1890 temblor is also suspect in view of this new information and the intensity data. The estimate of the amount of slip in the northwestern slip region is reduced from 130 to 60 cm if only instrumentally determined moments are used, excluding the 1890 and the 1892 events whose moments and locations are too poorly constrained. In the southeastern slip area the Superstition Hills earthquake has a moment equal to the 1968 Borrego Mountain event and will bring the average slip from 80 to 90 cm.

Recently the gWGCEPg [1988] conducted a study to evaluate the conditional probabilities for large earthquakes along the San Andreas fault system including the San Jacinto fault zone. The San Jacinto was divided into five segments (Figure 3) based on slip rates, various earthquake locations, and background seismicity. Three of these segments have ruptured since the turn of the century, the San Jacinto Valley segment in 1899 and 1918, the Borrego Mountain segment in 1968, and the Superstition segment in 1987. Low conditional 30-year probabilities (&le .1) for large events were assigned to each of these three segments. The Superstition segment was not initially even rated due to the lack of information. The day after it was discussed by the WGCEP, the segment filling earthquake occurred (MS = 6.6). The remaining two segments, Anza and San Bernardino Valley, could possibly generate a magnitude 7 earthquake [gWGCEPg, 1988]. While neither segment has had an

Probably the most important new information regarding the recurrence intervals of earthquakes in the Anza seismic slip gap are the results of gKlinger and Rockwellg [1989] and Klinger [1989]. gKlinger and Rockwellg [1989] suggest a recurrence interval of about 250 years for slip events at Hog Lake on the Clark strand of the San Jacinto fault. gKlingerg [1989] proposes that there was small amounts of offset ( <~ 20 cm) associated with each of the last two slip events. The first event occurred approximately 250 ± 50 years ago and the most recent was during the past 180 years and possibly was caused by the 1899 or 1918 earthquakes. These two events do not indicate offsets which are comparable with the geologic slip rate of 1 cm/year. Probably the last slip event which caused a large offset in the Hog Lake sediments occurred approximately 550 ± 50 years ago [gKlingerg, 1989]. If this was the last segment filling earthquake in the Anza segment (Figure 3), then we can calculate from the geologic slip rate that there may be as much as 5 meters of slip deficit along this part of the San Jacinto fault zone.

6. Conclusions

The review of the available data about the San Jacinto fault zone yields the following results: The geological and geodetic slip rates are consistent with a value of 1 cm/yr for most of the length of the fault. The aggregate seismic moment for all events since 1895 when distributed over an area of 240 x 15 km2 yields an average seismic slip of 0.5 cm/yr. The San Bernardino and Anza seismic slip gaps, identified by gThatcher et alg. [1975], last ruptured before 1895 and most likely before 1850. The slip deficit estimate was a lower bound of 1 meter. The results of the paleoseismic study by gKlingerg [1989] indicate then there may be a slip deficit of 5 meters in the Anza seismic slip gap which would give a 30-year conditional probability for a segment filling earthquake of .3. The Borrego Mountain and Superstition Hills segments of the fault have essentially no slip deficit. The source area for the 1918 earthquake has 0.7 meters of slip deficit.


Abe, K., Magnitudes and origin times from Milne seismograph data: earthquakes in China and California, in gHistorical Seismograms and Earthquakes of the Worldg, edited by W.H.K. Lee, H. Meyers and M. Shimazaki, pp. 37-50, Academic Press, London, 1988.

Bent, A., P. Ho-Liu, and D. Helmberger, The November 1987 Superstition Hills earthquake and comparisons with previous neighboring events, gSeis. Res. Lett., 59g, no. 1, 49, 1988.

Brune, J. N., Seismic moment, seismicity, and rate of slip along major fault zones, gJ. Geophys. Res., 73,g 777-784, 1968.

Budding, K. E., and R. V. Sharp, Surface Faulting Associated with the Elmore Desert Ranch and Superstition Hills, California, Earthquakes of 24 November 1987, gSeis. Res. Lett., 59g, no. 1, 49, 1988.

Clark, M. M., A. Gianz, and M. Rubin, Holocene activity of the Coyote Creek fault as recorded in sediments of Lake Cahuilla, in gThe Borrego Mountain Earthquake of April 9, 1968g, U.S.G.S. Professional Paper 787, pp. 112-130, 1972.

Fletcher, J., L. Haar, T. Hanks, L. Baker, F. L. Vernon III, J. Berger, and J.N. Brune, The digital array at Anza, California: Processing and initial interpretation of source parameters, gJ. Geophys. Res. 92,g 369-382, 1987.

Frankel, A., Source parameters of two ML approx 5 earthquakes near Anza, California, and a comparison with an Imperial Valley aftershock, gBull. Seis. Soc. Am., 74(g5), 1509-1527, 1984.

Given, D. D., L. K. Hutton, and L. M. Jones, gThe Southern California Network Bulletin July--December 1986g, U.S.G.S. Open File Report 87-488, 1987.

Gutenberg, B., and C. F. Richter, gSeismicity of the Earthg, 2nd ed., 310pp, Princeton University Press, Princeton, NJ, 1954.

Hanks, T. C., J. A. Hileman and W. Thatcher, Seismic moments of the larger earthquakes of the southern California region, gGeol. Soc. Am. Bull. 86g, 1131-1139, 1975.

Hill, R. I., Petrology and petrogenesis of batholithic rocks. San Jacinto Mountains, southern California, gPh.D. dissertationg, California Institute of Technology, Pasadena, 1984.

Kanamori, H., and P. C. Jennings, Determination of local magnitude, ML, from strong-motion accelerograms, gBull. Seis. Soc. Am., 68g (2), 471-485, 1989.

King, N. E., and J. C. Savage, Strain rate profile across the Elsinore, San Jacinto, and the San Andreas faults near Palm Springs, California, gGeophys. Res. Lett., 10,g 55-57, 1983.

Klinger, R. E., gMasters Thesisg, San Diego State University (expected completion 1989).

Klinger, R. E., and T. Rockwell, Recurrent Late Holocene faulting at Hog Lake in the Anza seismic gap, San Jacinto fault zone, southern California, Abstracts with Programs, gGeol. Soc. Am., 21g, no. 5, 102, 1989.

Louderback, G. D., History of the University of California seismographic stations and related activities, gBull. Seism. Soc. Am., 32g, 205-229, 1942.

Lawson, A. C., ed., gThe California earthquake of April 18, 1906: Report of the State Earthquake Investigation Commissiong, Carnegie Inst. of Washington, Washington, D.C., 1908.

Norris, R., L. M. Jones, and L. K. Hutton, gThe Southern California Network Bulletin July 01 through December 31, 1985g, U.S.G.S. Open File Report 86-337, 1986.

Richter C. F., gElementary Seismologyg, W.H. Freeman, San Francisco, Calif., 1958.

Rockwell, T., P. M. Merrifield and C. C. Loughman, Holocene activity of the San Jacinto in the Anza seismic gap, southern California, Abstracts with Programs, gGeol. Soc. Am., 18g, no. 2, 177, 1986.

Rolfe, F., and A. M. Strong, The earthquake of April 21, 1918 in the San Jacinto mountains, gBull. Seis. Soc. Am., 8g(2-3), 63-67, 1918.

Sanders, C. O., Seismotectonics of the San Jacinto fault zone and the Anza seismic gap, gunpublished Ph.D. dissertationg, California Institute of Technology, Pasadena, 143 pp., 1986.

Sharp, R. V., San Jacinto fault zone in the Peninsular Ranges of southern California, gGeol. Soc. Am. Bull., 78,g 705-730, 1967.

Sharp. R. V., Map of recently active breaks along the San Jacinto fault zone between the San Bernardino area and Borrego Valley, California, gU.S. Geol. Surv. Misc. Geol. Invest. Map, I-675g, 1972.

Sharp, R. V., Variable rates of late Quaternary strike-slip on the San Jacinto fault zone, southern California, gJ. Geophys. Res., 86g, 1754-1762, 1981.

Thatcher, W., and R. M. Hamilton, Aftershocks and source characteristics of the 1969 Coyote Mountain earthquake, San Jacinto fault zone, California, gBull. Seis. Soc. Am., 63g, 647-661, 1973.

Thatcher, W., J. A. Hileman, and T. C. Hanks, Seismic slip distribution along the San Jacinto fault zone, southern California, and its implications, gGeol. Soc. Am. Bull., 86g, 1140-1146, 1975.

Toppozada, T. R., C. R. Reel and D. L. Parke, Preparation of isoseismal maps and summaries of reported effects for pre-1900 California earthquakes, gCalif. Div. Mines and Geol., Open-File Rep. 81-11,g Sacramento, 182 pp., 1981.

Toppozada, T. R., and C. R. Parke, Areas damaged by California earthquakes 1900-1949, gCalif. Div. Mines and Geol., Open File Rep. 82-17 SAC,g Sacramento, 1982.

Townley, S. D., and M. W. Allen, Earthquakes in California, 1769 to 1928, gBull. Seis. Soc. Am., 29,g 1-258, 1939.

Townley, S. D., The San Jacinto earthquake of April 21, 1918, gBull. Seis. Soc. Am., 8g(2-3), 44-62, 1918.

Wesnousky, S. G., C. S. Prentice, and K. E. Sieh, Fault slip rate determination on the northern segment of the San Jacinto fault, San Bernardino, California, gEOS, Trans. Am. Geophys. Union, 68g, 1506, 1987.

Wood, H. O., The Terwilliger Valley earthquake of March 25, 1937, gBull. Seis. Soc. Am., 27,g 305-312, 1937.

Working Group on California Earthquake Probabilities, Probabilities of large earthquakes occurring in California on the San Andreas fault, gU.S. Geological Survey, Open File Report, 88-398g, pp. 1-62, 1988.


TABLE 1: Significant Right-Lateral Events Along San Jacinto Fault Zone
Date & Time Lat Lon MI ML MS MO 1018N-m
1890 FEB 9 12:06 33.4? (T81) 116.3? 6.3 (T81) - - -
1892 MAY 28 11:15 33.2? (T81) 116.2? 6.3 (T81) - - -
1899 JUL 22 20:32 34.3 (T81) 117.5 6.5 (T81) - 5.6 (A88) 0.2 (W88)
1899 DEC 25 12:25 33.8 (T81) 117.0 6.6 (T81) - 6.4 (A88) 4 (W88)
1918 APR 21 22:32 33.8 (T82) 117.0 6.8 (T82) - 6.8 (GR54) 15 (H75)
1923 JUL 23 07:30 34.1 (S86) 117.3 6.0 (T82) - 6.25 (GR54) 1 (H75)
1937 MAR 25 16:49 33.465 (S86) 116.415 6.0 (T82) 5.9 (S86) - 0.3 (H75)
1942 OCT 21 16:22 33.049 (S86) 116.088 6.5 (T82) 6.3 (S86) 6.5 (GR54) 9.4 (T73)
1954 MAR 19 09:54 33.296 (S86) 116.176 - 6.2 (S86) - 4.4 (T73)
1968 APR 9 02:28 33.19 (S86) 116.128 - 6.8 (KJ78) 6.9 (ISC) 8.3 (H75)
1969 APR 28 23:32 33.343 (S86) 116.341 - 5.8 (T73) 5.2 (1 obs.) (ISC) 0.5 (T73)
1980 FEB 25 10:47 33.50 (CT) 116.513 - 5.5 (CT) 4.7 (ISC) .025 (F84)
1987 NOV 24 13:15 33.012 (CT) 115.839 - 6.0 (CT) 6.6 (ISC) 8.25 (B88)

References: A88: gAbeg [1988]; B88: Bent et al. [1988]; F84: Frankel [1984]; H75: Hanks et al. [1975]; KJ78: Kanamori and Jennings [1978]; R58: Richter [1958]; S86: gSandersg [1986]; T73: Thatcher and Hamilton [1973]; W88: WGCEP [1988]; GR54: Gutenberg and Richter [1954]; CT: Norris et al. [1986] and Given et al. [1987].

URL: http://eqinfo.ucsd.edu/faq/thesis_vernon/chapter_1.php [Last updated: 2010-11-23 (327) 23:01:44 UTC]