Rupture Segment Research Articles

Unusually large earthquakes inferred from tsunami deposits along the Kuril trench.

The Pacific plate converges with northeastern Eurasia at a rate of 8-9 m per century along the Kamchatka, Kuril and Japan trenches. Along the southern Kuril trench, which faces the Japanese island of Hokkaido, this fast subduction has recurrently generated earthquakes with magnitudes of up to approximately 8 over the past two centuries. These historical events, on rupture segments 100-200 km long, have been considered characteristic of Hokkaido's plate-boundary earthquakes. But here we use deposits of prehistoric tsunamis to infer the infrequent occurrence of larger earthquakes generated from longer ruptures. Many of these tsunami deposits form sheets of sand that extend kilometres inland from the deposits of historical tsunamis. Stratigraphic series of extensive sand sheets, intercalated with dated volcanic-ash layers, show that such unusually large tsunamis occurred about every 500 years on average over the past 2,000-7,000 years, most recently approximately 350 years ago. Numerical simulations of these tsunamis are best explained by earthquakes that individually rupture multiple segments along the southern Kuril trench. We infer that such multi-segment earthquakes persistently recur among a larger number of single-segment events.

Structure and Mechanics of the Hayward-Rodgers Creek Fault Step-Over, San Francisco Bay, California

A dilatational step-over between the right-lateral Hayward and Rodgers Creek faults lies beneath San Pablo Bay in the San Francisco Bay area. A key seismic hazard issue is whether an earthquake on one of the faults could rupture through the step-over, enhancing its maximum possible magnitude. If ruptures are terminated at the step-over, then another important issue is how strain transfers through the step. We developed a combined seismic reflection and refraction cross section across south San Pablo Bay and found that the Hayward and Rodgers Creek faults converge to within 4 km of one another near the surface, about 2 km closer than previously thought. Interpretation of potential field data from San Pablo Bay indicated a low likelihood of strike-slip transfer faults connecting the Hayward and Rodgers Creek faults. Numerical simulations suggest that it is possible for a rupture to jump across a 4-km fault gap, although special stressing conditions are probably required (e.g., Harris and Day, 1993, 1999). Slip on the Hayward and Rodgers Creek faults is building an extensional pull-apart basin that could contain hazardous normal faults. We investigated strain in the pull-apart using a finite-element model and calculated a � 0.02-MPa/yr differential stressing rate in the step-over on a least-principal-stress orientation nearly parallel to the strike-slip faults where they overlap. A 1- to 10- MPa stress-drop extensional earthquake is expected on normal faults oriented per- pendicular to the strike-slip faults every 50-500 years. The last such earthquake might have been the 1898 M 6.0-6.5 shock in San Pablo Bay that apparently pro- duced a small tsunami. Historical hydrographic surveys gathered before and after 1898 indicate abnormal subsidence of the bay floor within the step-over, possibly related to the earthquake. We used a hydrodynamic model to show that a dip-slip mechanism in north San Pablo Bay is the most likely 1898 rupture scenario to have caused the tsunami. While we find no strike-slip transfer fault between the Hayward and Rodgers Creek faults, a normal-fault link could enable through-going segmented rupture of both strike-slip faults and may pose an independent hazard of M � 6 earthquakes like the 1898 event.

Analysis of the 23 June 2001 Mw = 8.4 Peru underthrusting earthquake and its aftershocks

On 23 June 2001, a Mw = 8.4 underthrusting earthquake occurred in the southern Peru subduction zone, followed by several large aftershocks, including 26 June (Mw = 6.7) and 7 July (Mw = 7.5). Broadband analyses of seismic data for the largest of these earthquakes show southeastward rupture of 180 km along the portion of the subduction zone previously ruptured in 1868 (Mw 8.8–9). Moment release distributions determined are consistent with aftershock location patterns. Earthquake rupture mode varies along southern Peru, from larger multi‐segment rupture in 1868 to several smaller segment ruptures in 2001, similar to rupture variation observed in northern Peru and other subduction zones. Based on models of subduction zone segment interaction, the 2001 earthquake sequence may suggest a shorter recurrence time for future earthquakes along this portion of the Peru‐Chile subduction zone.

The 1999 Mw 7.1 Hector Mine, California, Earthquake Sequence: Complex Conjugate Strike-Slip Faulting

The 1999 M_w 7.1 Hector Mine mainshock showed right-lateral strike-slip faulting, with an initial strike of N6°W and vertical dip. The mainshock was preceded within 20 hours by 18 recorded foreshocks of 1.5 ≤ M ≤ 3.8 within a few kilometers distance of the mainshock hypocenter. The aftershocks delineate how the Hector Mine earthquake ruptured with strike N6°W to the south for a distance of 15 km, and possibly to the north for a distance of several kilometers. The two largest aftershocks of M 5.9 and M 5.7 occurred near the north and south ends of the first mainshock rupture segment. The second segment of rupture, starting 15 km to the south away from the mainshock hypocenter, delineated by strike-slip and thrust-faulting aftershocks, extends 10 km farther away with a strike of S140°E along the Bullion fault. The aftershocks also outline an unusual third rupture segment, extending from about 5 km south of the hypocenter with a strike of N30°W to N35°W for a distance of 20 km. Approximately 10 to 25 km farther to the north and west of the mainshock epicenter, several clusters form a complex aftershock distribution. Three-dimensional Vp and Vp/Vs models of the region exhibit only small regional changes, as is typical for the Mojave region. Nonetheless, the mainshock rupture started within a region of rapidly varying Vp, and at least three regions of low Vp/Vs are imaged within the aftershock zone. The rate of decay for the Hector Mine earthquake sequence has been slightly above the mean for both p-values and b-values in southern California. The focal mechanisms of the aftershocks and the state of stress are consistent with strike-slip faulting, including a component of normal faulting most prominent to the north. The orientation of the regional maximum horizontal stress, the variation in orientation of the mainshock fault segments by 30°, and scattered distribution of aftershocks suggest that the mainshock and aftershock deformation field exhibit volumetric shear deformation accommodated by complex conjugate sets of strike-slip faults.

Active Tectonics of Northeastern Sonora, Mexico (Southern Basin and Range Province) and the 3 May 1887 Mw 7.4 Earthquake

North–south-striking and west-dipping Basin and Range province normal faults form the western edge of the Sierra Madre Occidental plateau in northeastern Sonora. These faults and associated half-grabens extend over a distance of more than 300 km between the San Bernardino basin in the north and the Sahuaripa basin in the south. An earthquake in 1887 ruptured three neighboring segments of this major fault zone. Our field mapping in this region indicates that the surface rupture of the 1887 earthquake extends farther to the south and is considerably longer (101.4 km end-to-end length) than previously reported. A compilation of the seismicity in the epicentral region of the 1887 earthquake shows the epicenters to be distributed in well-defined clusters at the northern end of the 1887 surface rupture, in the step-overs between the three rupture segments, on a neighboring fault in the west (Fronteras fault), and on fault segments farther south along the same fault zone (Granados region). The distribution of seismicity correlates well with calculated changes in Coulomb failure stress resulting from the 1887 earthquake.

A Note on Seismic Activity Near the Eastern Termination of the Izmit Rupture in the Hours Preceding the Duzce Earthquake

We use continuous recordings at four stations installed near the eastern termination of the Izmit rupture to analyze the seismic activity there during the 5 hr leading to the Duzce earthquake. The records clearly show the presence of three groups of events. One of the groups originates from the middle of the Izmit rupture and consists of aftershocks of the M 5.7 Spanca Lake earthquake that had occurred a day earlier and was itself, to date, the second largest aftershock in the Izmit sequence. A second group of events is spread over the 30-km-long eastern (Karadere) segment of the Izmit rupture, and a comparison with the number of aftershocks expected, based on decay laws for similar sequences in other regions, suggests that seismic activity was high on the eastern segment in the hours leading to the Duzce earthquake. The third group of events occurs beyond the termination of the Izmit rupture and clusters around the hypocenter of the Duzce earthquake. It shows that, in the little over 5 hr leading to the Duzce earthquake, this destructive event was preceded by six precursory shocks ranging in magnitude from 0.9 to 2.6. Manuscript received 30 August 2000.

Surface Rupture and Slip Distribution along the Karadere Segment of the 17 August 1999 Izmit and the Western Section of the 12 November 1999 Duzce, Turkey, Earthquakes

On 17 August 1999 the Mw 7.5 I zmit, Turkey, earthquake produced surface rupture in excess of 120 km, and perhaps as much as 200 km, with up to 5 m of dextral slip, along a western portion of the North Anatolian fault zone. The 12 November 1999 Mw 7.1 Duzce, Turkey, earthquake produced a 40-km-long sur- face rupture, including 9 km of rupture overlap with the eastern end of the August event. Our mapping focused on the 40-km-long Karadere rupture segment, the east- ernmost segment of the August event, as well as on the western 20 km of the No- vember rupture. Maximum dextral slip along the Karadere segment is approximately 1.5 m, and the average slip on this segment is close to 1 m. Although slip along the Karadere segment is considerably less than that on segments to the west, this segment is of particular interest for three reasons: (1) the western boundary of the Karadere segment is defined by the most striking structural discontinuity along the entire Au- gust surface rupture (i.e., a 5-km-wide zone of no surface rupture, as well as a 25 change in trend from E-W to ENE), and such a discontinuity may have important implications for rupture dynamics; (2) surface rupture terminates at the east end of the Karadere rupture segment at a 1.5- to 3-km-wide extensional step-over at Eften Lake; and (3) the 12 November 1999 Duzce earthquake reruptured the easternmost 9 km of the Karadere segment, raising interesting questions about rupture dynamics and interactions between events on adjacent fault segments. The details of the 17 August 1999 and l2 November 1999 surface rupture traces suggest that rupture may have partially propagated across the Eften Lake extensional step-over, although this step-over seems to have acted as an effective barrier to rupture propagation.

Surface Ruptures of the 17 August and 12 November 1999 Izmit and Duzce Earthquakes in Northwestern Anatolia, Turkey: Their Tectonic and Kinematic Significance and the Associated Damage

The 17 August Izmit (Kocaeli) and 12 November Duzce earthquakes ( M w 7.4 and 7.1, respectively) represent the most recent two of a series of large earthquakes along the North Anatolian Fault, a highly active right-lateral strike-slip fault that crosses the Anatolian Peninsula in Turkey from one end to the other. The first earthquake rupture apparently initiated at a small step-over along a relatively inconspicuous fault strand following the southern coast of the Izmit Bay and propagated to the west and to the east for a distance of 125-150 km. The western segment enters the Bay in both directions after attaining the largest surface slip of about 5 m at the town of Golcuk. The eastern extension of the surface rupture associated with the Izmit earthquake is composed of five major segments with the maximum slip magnitudes from 3 to 4, 4.5, 0.5, 1.4, and 0.1 m from west to east, respectively. The segment with 0.5 m slip trends N70°W, the segment with 1.4-m slip trends N70°E, and the easternmost segment with minor slip trends east-west again. The surface manifestation of the Duzce earthquake is an extensively segmented rupture trace with dominant transtensional deformation. It has a normal slip component of more than 3.7 m and a right-lateral slip close to 1.4 m along the western part that showed minor rupture during the first earthquake. The central part has a transpressional character with about 1-m reverse slip and about 3-m right-lateral slip. The highest right-lateral slip (nearly 5 m) occurs at the central part. The eastern part of the rupture displays a typical strike-slip deformation with less than about 3 m right-lateral slip and minor local vertical slip that exhibits changing polarity. The varying orientations of the rupture segments reflect the fault mosaic of the broader region and the interaction between neighboring faults. The temporal and spatial characteristics of the rupture segments, as well as the significance of the nearly 3-month time period between the first and second earthquakes, pose highly challenging mechanical problems regarding the rupture initiation, propagation, and termination processes. The main surface rupture zone, although unusually narrow (2.5-6 m), left a path of death and destruction. Manuscript received 1 June 2000.

Postseismic Deformation near the Izmit Earthquake (17 August 1999, M 7.5) Rupture Zone

We present and interpret the results of postseismic, Global Positioning System monitoring of the first 298 days following the 17 August 1999 Izmit earth- quake. Whereas the data suggest some spatial and temporal complexity in the post- seismic motions, the overall pattern can be characterized by time-dependent relaxa- tion functions and suggests exponential decay with an estimated 57-day relaxation time. The very rapid deformation during the first few weeks following the mainshock suggests rapid afterslip on and below the coseismic rupture segments. The exponen- tial decay of the postseismic deformations through the end of the observation period suggests contributions from the lower crustal viscoelastic relaxation.

Surface Fault Breaks, Aftershock Distribution, and Rupture Process of the 17 August 1999 Izmit, Turkey, Earthquake

The results obtained from field studies of surface fault breaks on land, aftershock studies, and analysis of complex teleseismic body waves of the 17 August 1999 Izmit earthquake ( M w 7.4) indicate that the northern strand of the North Anatolian Fault Zone had been ruptured between 29.1° E and 31.1° E, with a maximum total fault rupture length of about 200 km. The 12 November 1999 Duzce earthquake ( M w 7.2) extended the surface fault break an additional 41 km to the east. Alignment of the aftershocks delineates the fault rupture on land and extends it to the west of the Hersek Delta, well into the Marmara Sea, to 29.1° E longitude. We have identified six fault rupture segments: two of them are located to the west of the Hersek Delta, and four are located to the east. Rupture process analysis suggests two stages of rupture separated in space and time, with an estimated total scalar seismic moment of 2.42 × 1020 N m. The first stage initiated at the epicenter, near Izmit, and propagated asymmetrically—first eastward and then westward with a total scalar moment of 1.6 × 1020 N m, showing predominantly right-lateral strike-slip faulting followed by a minor subevent extending the rupture further east. The second stage took place at the western and eastern ends of the source area, rupturing two segments within the Marmara Sea, west of the Hersek Delta, and two segments between Akyazi and Golyaka. The rupture process analysis indicates that six fault segments were ruptured during the mainshock: four of them on land and two within the Marmara Sea. The major moment release of 1.2 × 1020 N m is found on the 64-km-long Sapanca segments located between the easternmost tip of the Gulf of Izmit and Akyazi, with an estimated maximum slip of 5.3 m. The moment release on the Golcuk segment, with a length of 30 km, is about 3.7 × 1019 N m, and the estimated slip on this segment is 4.1 m. Rupture process analysis suggests two 20-km-long segments between Akyazi and Golyaka striking N70°E. The scalar moment releases on these segments are 2.7 × 1019 and 1.9 × 1019 N m, with estimated offsets of 2.2 and 1.6 m, respectively. The last two subevents took place to the west of the Hersek Delta on the Yalova segment and within the Marmara Sea, with estimated offsets of 1.3 and 1.0 m, respectively. Manuscript received 20 August 2000.

Time-Dependent Distributed Afterslip on and Deep below the Izmit Earthquake Rupture

Surface deformation transients measured with the Global Positioning System during the 87 days between the 17 August 1999 Izmit earthquake and the 12 November 1999 Duzce earthquake indicate rapidly decaying aseismic fault slip on and well below the coseismic rupture. Elastic model inversions for time-dependent distributed fault slip, using a network inversion filter approach, show that afterslip was highest between and below the regions of maximum coseismic slip and propa- gated downward to, or even below, the base of the crust. Maximum afterslip rates decayed from greater than 2 m/yr, immediately after the I zmit earthquake to about 1.2 m/yr just prior to the Duzce earthquake. Maximum afterslip occurred below the eastern Karadere rupture segment and near the I zmit hypocenter. Afterslip in the upper 16 km decayed more rapidly than that below the seismogenic zone. These observations are consistent with a phase of rapid aseismic fault slip concentrated near the base of the seismogenic zone. Continued loading from the rapid deep afterslip along the eastern rupture zone is a plausible mechanism that helped trigger the nearby, Mw 7.2, 12 November Duzce earthquake.

Geometry, Slip Distribution, and Kinematics of Surface Rupture on the Sakarya Fault Segment during the 17 August 1999 Izmit, Turkey, Earthquake

The M w 7.4 17 August 1999 Izmit earthquake ruptured five major fault segments of the dextral North Anatolian Fault Zone. The 26-km-long, N86°W-trending Sakarya fault segment (SFS) extends from the Sapanca releasing step-over in the west to near the town of Akyazi in the east. The SFS emerges from Lake Sapanca as two distinct fault traces that rejoin to traverse the Adapazari Plain to Akyazi. Offsets were measured across 88 cultural and natural features that cross the fault, such as roads, cornfield rows, rows of trees, walls, rails, field margins, ditches, vehicle ruts, a dike, and ground cracks. The maximum displacement observed for the Izmit earthquake (∼5.1 m) was encountered on this segment. Dextral displacement for the SFS rises from less than 1 m at Lake Sapanca to greater than 5 m near Arifiye, only 3 km away. Average slip decreases uniformly to the east from Arifiye until the fault steps left from Sagir to Kazanci to the N75°W, 6-km-long Akyazi strand, where slip drops to less than 1 m. The Akyazi strand passes eastward into the Akyazi Bend, which consists of a high-angle bend (18°-29°) between the Sakarya and Karadere fault segments, a 6-km gap in surface rupture, and high aftershock energy release. Complex structural geometries exist between the Izmit, Duzce, and 1967 Mudurnu fault segments that have arrested surface ruptures on timescales ranging from 30 sec to 88 days to 32 yr. The largest of these step-overs may have acted as a rupture segmentation boundary in previous earthquake cycles. Manuscript received 10 July 2000.

Inferring Segment Strength Contrasts and Boundaries along Low-Friction Faults Using Surface Offset Data, with an Example from the Cholame-Carrizo Segment Boundary along the San Andreas Fault, Southern California

Rupture segmentation arises from changes in fault geometry and strength. We use boundary element models of frictionless strike-slip fault segments to quantify how fault geometry and strength change earthquake surface offset distributions. Using these relationships between fault geometry, strength, and surface offsets, we can infer fault strength from the surface offsets in cases where the fault geometry can be independently constrained. This article includes normalized plots of the surface offset distribution expected from rupture along low-friction fault segments with strength contrasts of 1/4, 1/3, 1/2, 1, 2, 3, and 4 for a range of fault segment geometries. These plots may be used with offset data to constrain the strength of two coplanar, adjacent fault segments. This analysis is applied to the Cholame and Carrizo segments of the San Andreas Fault. The available surface offset data suggest that the offset increases where the fault deepens; in addition, the observed offset gradient at the segment boundary requires a 2/3–1/4 strength ratio of the Cholame to the Carrizo segment.

Probability of Rupture of Multiple Fault Segments

Fault segments identified from geologic and historic evidence have sometimes been adopted as features limiting the likely extends of earthquake ruptures. There is no doubt that individual segments can sometimes join together to produce larger earthquakes. This work is a trial of an objective method to determine the probability of multisegment ruptures. The frequency of occurrence of events on all conjectured combinations of adjacent segments in northern California is found by fitting to both geologic slip rates and to an assumed distribution of event sizes for the region as a whole. Uncertainty in the shape of the distribution near the maximum magnitude has a large effect on the solution. Frequencies of individual events cannot be determined, but it is possible to find a set of frequencies to fit a model closely. A robust conclusion for the San Francisco Bay region is that large multisegment events occur on the San Andreas and San Gregorio faults, but single-segment events predominate on the extended Hayward and Calaveras strands of segments.

A Time-Dependent Probabilistic Seismic-Hazard Model for California

For the purpose of sensitivity testing and illuminating nonconsensus components of time-dependent models, the California Department of Conservation, Division of Mines and Geology (CDMG) has assembled a time-dependent version of its statewide probabilistic seismic hazard (PSH) model for California. The model incorporates available consensus information from within the earth-science community, except for a few faults or fault segments where consensus information is not available. For these latter faults, published information has been incorporated into the model. As in the 1996 CDMG/U.S. Geological Survey (USGS) model, the time-dependent models incorporate three multisegment ruptures: a 1906, an 1857, and a southern San Andreas earthquake. Sensitivity tests are presented to show the effect on hazard and expected damage estimates of (1) intrinsic (aleatory) sigma, (2) multisegment (cascade) vs. independent segment (no cascade) ruptures, and (3) time-dependence vs. time-independence. Results indicate that (1) differences in hazard and expected damage estimates between time-dependent and independent models increase with decreasing intrinsic sigma, (2) differences in hazard and expected damage estimates between full cascading and not cascading are insensitive to intrinsic sigma, (3) differences in hazard increase with increasing return period (decreasing probability of occurrence), and (4) differences in moment-rate budgets increase with decreasing intrinsic sigma and with the degree of cascading, but are within the expected uncertainty in PSH time-dependent modeling and do not always significantly affect hazard and expected damage estimates.

Late Quaternary movement on the Ohariu Fault, Tongue Point to MacKays Crossing, North Island, New Zealand

The Ohariu Fault is one of the major active dextral strike‐slip faults in the Wellington region. It extends northeastward from offshore at Tongue Point to Waikanae and beyond, but in the north the fault trace becomes intermittent. Extrapolation of the trend of the Ohariu Fault across Porirua Harbour suggests a 1.5 km right step. South of Porirua Harbour the fault is characterised by lateral displacement of drainage features of up to 450 m, a single‐event lateral displacement of 4–5 m, an average horizontal slip rate of 1–2 mm/yr, and a recurrence interval of 2000–5000 yr. The timing of the last event was between 150 and 1130 cal. BP. North of the harbour the fault is characterised by lateral displacement of drainage features of up to 250 m, a single‐event lateral displacement of 2.9 m, a horizontal slip rate of 0.6–1.9 mm/yr, and a recurrence interval of 1530–4830 yr. The timing of the last event was between 1070 and 2310 cal. BP. Although the possibility exists that Porirua Harbour is a pull‐apart basin marking separate rupture segments on the Ohariu Fault, the data suggest there is a single rupture segment from Tongue Point to Waikanae. On this assumption, the last movement on the fault occurred 1070–1130 yr ago. The average horizontal surface displacement for the event was 3.7 m (based on seven sites) and the estimated earthquake magnitude was Mw 7.1–7.5.

Rupturing of the Awatere Fault during the 1848 October 16 Marlborough earthquake, New Zealand: Historical and present day evidence

Evidence from newspaper reports, diaries and journals, related first‐hand information, an 1854 survey map of the lower part of the Awatere Valley, and reports by geologists between 1856 and 1890, indicates that surface rupturing occurred on the Awatere Fault during the 1848 October 16 earthquake and not on the Wairau Fault as previously inferred. The rupture was initially described as a “fissure”, “crack”, and later as a “rent”, and although it extended for c. 105 km (from the coast to Barefell Pass), it was not termed a fault because displacement of the land surface or strata across the rupture could not be determined. The coincidence of the 1848 earthquake “rent” and the Awatere Fault was first demonstrated by Alexander McKay in 1885. Present day evidence of the 1848 Awatere Fault rupture is indicated by a depression between 0.6–1.5 m wide and c. 0.3 m deep that has the appearance of an infilled fissure similar to that described in early reports. The smallest and freshest displacements along the fault (15 displacements; 12 localities, most within c. 20 km of the coast) range between 4 and 8 m dextral with both the southeast and northwest sides variably upthrown between 3.5 and 0.4 m. Dextral and vertical displacement appears to decrease towards the coast, consistent with a low rate of late Quaternary slip (<1.4 mm/ yr) and the inferred termination of the fault in Cook Strait. The 1848 rupture apparently bypassed the ENE‐striking Molesworth section of the Awatere Fault, which implies that the intersection (near Molesworth Station) between this strand and the eastern section of the fault that reaches the coast may be an important mechanical boundary controlling the location of earthquake rupture segments along the Awatere Fault. Reassessment of felt intensities (based on contemporary descriptions), minimum fault rupture length (105 km), and average dextral displacement (6 m), indicates a magnitude for the 1848 earthquake of M 7.4–7.5.

Paleoseismic investigations in the Santa Cruz mountains, California: Implications for recurrence of large‐magnitude earthquakes on the San Andreas Fault

Trenching, microgeomorphic mapping, and tree ring analysis provide information on timing of paleoearthquakes and behavior of the San Andreas fault in the Santa Cruz mountains. At the Grizzly Flat site alluvial units dated at 1640–1659 A.D., 1679–1894 A.D., 1668–1893 A.D., and the present ground surface are displaced by a single event. This was the 1906 surface rupture. Combined trench dates and tree ring analysis suggest that the penultimate event occurred in the mid‐1600 s, possibly in an interval as narrow as 1632–1659 A.D. There is no direct evidence in the trenches for the 1838 or 1865 earthquakes, which have been proposed as occurring on this part of the fault zone. In a minimum time of about 340 years only one large surface faulting event (1906) occurred at Grizzly Flat, in contrast to previous recurrence estimates of 95–110 years for the Santa Cruz mountains segment. Comparison with dates of the penultimate San Andreas earthquake at sites north of San Francisco suggests that the San Andreas fault between Point Arena and the Santa Cruz mountains may have failed either as a sequence of closely timed earthquakes on adjacent segments or as a single long rupture similar in length to the 1906 rupture around the mid‐1600 s. The 1906 coseismic geodetic slip and the late Holocene geologic slip rate on the San Francisco peninsula and southward are about 50–70% and 70% of their values north of San Francisco, respectively. The slip gradient along the 1906 rupture section of the San Andreas reflects partitioning of plate boundary slip onto the San Gregorio, Sargent, and other faults south of the Golden Gate. If a mid‐1600 s event ruptured the same section of the fault that failed in 1906, it supports the concept that long strike‐slip faults can contain master rupture segments that repeat in both length and slip distribution. Recognition of a persistent slip rate gradient along the northern San Andreas fault and the concept of a master segment remove the requirement that lower slip sections of large events such as 1906 must fill in on a periodic basis with smaller and more frequent earthquakes.

A source extent analysis of the Lancang earthquake (M s=7.6) of 1988

Source extent parameters of the 1988 Lancang earthquake (M s=7.6) were estimated by computing the second-central moment of displacement pulses of far field long period SH waves. We inverted the source duration T, the fault length L and the directivity parameter D by the least squares, and obtained that T=11.77 s, D=15.05 km·s, and L=70.94 km. We also find that this event is a symmetrically bilateral rupture and fault segments in two opposite rupture directions have the same value of 35 km in length. Combining analysis of aftershock distribution, the results would imply that there is an area in the joint part of the rupture area of two main shocks, which is not broken yet during the main shocks. Due to the dislocation accompanied during the main shocks, the strain would be rearranged. The joint part has enough strength to accumulate enough strain energy to excite a larger aftershock.

Segmentation of the northern San Jacinto Fault Zone, southern California

Accurate forecasting of large earthquakes in the San Jacinto fault zone depends on the correct determination of fault segmentation. We have searched for discontinuities in the fault zone using as data geological maps of the fault traces, relocated earthquake hypocenters, focal mechanisms, slip rates, and historic large earthquake rupture zones. We identify nine principal discontinuities in the northern San Jacinto fault zone. These are characterized as structural discontinuities (bends, steps, branches, intersections), geodetic discontinuities (slip rate), and lithologic discontinuities. Most of the discontinuities have more than one defining characteristic. We also analyzed the local strain field orientation along the fault zone. Significant lateral discontinuities in the strain‐field orientation are coincident with many of the structural and geodetic discontinuities we identify. Segment lengths defined from the data vary from about 7 to 35 km. This range of segment lengths is consistent with the lengths of large historical earthquake ruptures. However, evidence from small earthquake hypocenter patterns suggests the possibility that in some areas large earthquakes may rupture multiple adjacent segments. In these cases, maximum magnitudes may be larger than those observed historically. Because of the short historic record and the sparse paleoseismic record, probabilistic estimates of earthquake recurrence for this fault zone have depended heavily on assumed characteristic lengths of individual rupture segments. Ours and previous delineations of segment lengths have varied considerably, indicating that assumed segment length is a major source of uncertainty in the recurrence calculations. This source of uncertainty should be taken into account in calculations of the reliability of the probabilistic hazard estimates.