Geology topics

W. Thatcher

Publications and source records attributed to W. Thatcher.

24 records · Page 2Linked to original sources

USGS1989

Earthquake recurrence and risk assessment in circum-Pacific seismic gaps

THE development of the concept of seismic gaps, regions of low earthquake activity where large events are expected, has been one of the notable achievements of seismology and plate tectonics. Its application to long-term earthquake hazard assessment continues to be an active field of seismological research. Here I have surveyed well documented case histories of repeated rupture of the same segment of circum-Pacific plate boundary and characterized their general features. I find that variability in fault slip and spatial extent of great earthquakes rupturing the same plate boundary segment is typical rather than exceptional but sequences of major events fill identified seismic gaps with remarkable order. Earthquakes are concentrated late in the seismic cycle and occur with increasing size and magnitude. Furthermore, earthquake rupture starts near zones of concentrated moment release, suggesting that high-slip regions control the timing of recurrent events. The absence of major earthquakes early in the seismic cycle indicates a more complex behaviour for lower-slip regions, which may explain the observed cycle-to-cycle diversity of gap-filling sequences.

Nature↗

USGS1988

Scientific goals of the Parkfield earthquake prediction experiment

Several unique circumstances of the Parkfield experiment provide unprecedented opportunities for significant advances in understanding the mechanics of earthquakes. to our knowledge, there is no other seismic zone anywhere where the time, place, and magnitude of an impending earthquake are specified as precisely. Moreover, the epicentral region is located on continental crust, is readily accessible, and can support a range of dense monitoring networks that are sited either on or very close to the expected rupture surface. As a result, the networks located at Parkfield are several orders of magnitude more sensitive than any previously deployed for monitoring earthquake precursors (a preearthquake change in strain, seismicity, and other geophysical parameters). In this respect the design of the Parkfield experiment resembles the rationale for constructing a new, more powerful nuclear particle accelerator:in both cases increased capabilities will test existing theories, reveal new phenomena, and suggest new research directions.

Earthquakes & Volcanoes (USGS)↗

USGS1983

Correlation of changes in gravity, elevation, and strain in southern California

Measurements made once or twice a year from 1977 through 1982 show large correlated changes in gravity, elevation, and strain in several southern California networks. Precise gravity surveys indicate changes of as much as 25 microgals between surveys 6 months apart. Repeated surveys show that annual elevation changes as large as 100 millimeters occur along baselines 40 to 100 kilometers long. Laser-ranging surveys reveal coherent changes in areal strain of 1 to 2 parts per million occurred over much of southern California during 1978 and 1979. Although the precision of these measuring systems has been questioned, the rather good agreement among them suggests that the observed changes reflect true crustal deformation.

California↗

USGS1979

Initiation and development of the southern California uplift along its northern margin

Analysis of three first-order leveling lines that traverse the White Wolf fault (site of the 1952 M = 7.7 earthquake), each resurveyed nine times between 1926 and 1974, reveals probable preseismic tilting, major coseismic movements, and a spatial association between these movements and the subsequently recognized southern California uplift. In examining the vertical control record, we have both searched for evidence of systematic errors and excluded from consideration portions of the lines contaminated by subsurface fluid and gas extraction. Movements have been referred to an invariant datum based on the 1926 position of tidal BM 8 in San Pedro, corrected for subsequent eustatic sea-level change. An 8 μrad up-to-the-north preseismic tilt (6 cm/7.5 km) was apparently recorded on two adjacent line segments within 10 km of the 1952 epicenter between 1942 and 1947. It is possible, however, that this tilt was in part caused by extraction-induced subsidence at one of the six releveled benchmarks. Data also show evidence of episodic tilts that are not earthquake related. At the junction of the Garlock and San Andreas faults, for example, an ≥5 μrad up-to-the-north tilt (7.2 cm/≤16 km) took place between Lebec and Grapevine within three months during 1964. Comparison of the 1947 and 1953 surveys, which includes the coseismic interval, shows that the SW-fault end (nearest the epicenter) and the central fault reach sustained four times the uplift recorded at the NE end of the fault (+72 cm SW, +53 cm Central, +16 cm NE). A regional postseismic uplift of 4 cm extended ≥25 km to either side of the fault after the main event, from 1953 to 1956. An interval of relative quiescence followed at least through 1959, in which the elevation change did not exceed ±3 cm. The detailed pattern of aseismic uplift demonstrates that movement proceeded in space—time pulses: one half of the uplift at the SW-fault end and extending southward occurred between 1959 and 1961, one half of the uplift at the NE-fault end and extending eastward occurred between 1961 and 1965, while the central fault reach sustained successive pulses of subsidence, uplift, and collapse (−4 cm, 1953–1960; +7 cm, 1960–1965; −2 cm, 1965–1970). In addition, the number of aftershocks concentrated near the fault ends increased in the NE relative to the SW from 1952 to 1974. These observations suggest that the aseismic uplift may have migrated northeastward from 1959 to 1965 at an approximate rate of 7–16 km/yr. Evidence for a mechanical coupling between the earthquake and the subsequent aseismic uplift is equivocal. At both fault ends, the major NWbounding flexure or tilted front of the southern California uplift is spatially coincident with the coseismic flexure that preceded it. In addition, the postulated migration of vertical deformation is similar to the 1952 seismic event in which the rupture initiated at the SW end of the fault and then propagated to the NE-fault end. However, the spatial distribution of aseismic uplift, nearly identical at both fault ends and to the south and east, and near zero in the central fault reach, is distinctly different from the nonuniform and localized coseismic deformation.

California↗

USGS1976

Episodic strain accumulation in Southern California

Reexamination of horizontal geodetic data in the region of recently discovered aseismic uplift has demonstrated that equally unusual horizontal crustal deformation accompanied the development of the uplift. During this time interval compressive strains were oriented roughly normal to the San Andreas fault, suggesting that the uplift produced little shear strain accumulation across this fault. On the other hand, the orientation of the anomalous shear straining is consistent with strain accumulation across northdipping range-front thrusts like the San Fernando fault. Accordingly, the horizontal and vertical crustal deformation disclosed by geodetic observation is interpreted as a short epoch of rapid strain accumulation on these frontal faults. If this interpretation is correct, thrust-type earthquakes will eventually release the accumulated strains, but the geodetic data examined here cannot be used to estimate when these events might occur. However, observation of an unusual sequence of tilts prior to 1971 on a level line lying to the north of the magnitude 6.4 San Fernando earthquake offers some promise for precursor monitoring. The data are adequately explained by a simple model of up-dip aseismic slip propagation toward the 1971 epicentral region. These observations and the simple model that accounts for them suggest a conceptually straightforward monitoring scheme to search for similar uplift and tilt precursors within the uplifted region. Such premonitory effects could be detected by a combination of frequently repeated short (30 to 70 km in length) level line measurements, precise gravity traverses, and continuously recording gravimeters sited to the north of the active frontal thrust faults. Once identified, such precursors could be closely followed in space and time, and might then provide effective warnings of impending potentially destructive earthquakes.

California↗

USGS1974

Strain release mechanism of the 1906 San Francisco earthquake

Reexamination of geodetic data has shown that aseismic slip occurred on or near the San Andreas fault in the period of about 20 years after 1906. The inferred displacements are comparable to but at greater depths than the sudden slip that occurred at the time of the earthquake. The postseismic slip is constrained only between late 1906 and 1925, and data are insufficient to determine the movements, if any, below about 20 kilometers on the fault. Two independent observations also indicate substantial anomalous crustal deformation away from the fault at least 30 years before the earthquake.

Science↗