William H. Bakun

Implications for prediction and hazard assessment from the 2004 Parkfield earthquake

Obtaining high-quality measurements close to a large earthquake is not easy: one has to be in the right place at the right time with the right instruments. Such a convergence happened, for the first time, when the 28 September 2004 Parkfield, California, earthquake occurred on the San Andreas fault in the middle of a dense network of instruments designed to record it. The resulting data reveal aspects of the earthquake process never before seen. Here we show what these data, when combined with data from earlier Parkfield earthquakes, tell us about earthquake physics and earthquake prediction. The 2004 Parkfield earthquake, with its lack of obvious precursors, demonstrates that reliable short-term earthquake prediction still is not achievable. To reduce the societal impact of earthquakes now, we should focus on developing the next generation of models that can provide better predictions of the strength and location of damaging ground shaking.

A bilinear source-scaling model for M-log a observations of continental earthquakes

The Wells and Coppersmith (1994) M -log A data set for continental earthquakes (where M is moment magnitude and A is fault area) and the regression lines derived from it are widely used in seismic hazard analysis for estimating M , given A . Their relations are well determined, whether for the full data set of all mechanism types or for the subset of strike-slip earthquakes. Because the coefficient of the log A term is essentially 1 in both their relations, they are equivalent to constant stress-drop scaling, at least for M ≤ 7, where most of the data lie. For M > 7, however, both relations increasingly underestimate the observations with increasing M . This feature, at least for strike-slip earthquakes, is strongly suggestive of L-model scaling at large M . Using constant stress-drop scaling (Δσ = 26.7 bars) for M ≤ 6.63 and L-model scaling (average fault slip ū = α L , where L is fault length and α = 2.19 × 10-5) at larger M , we obtain the relations \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \[\mathbf{M}=\mathrm{log}{\ }A+3.98{\pm}0.03,{\ }A{\leq}537{\ }\mathrm{km}^{2}\] \end{document} and \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \[\mathbf{M}=4{/}3{\ }\mathrm{log}{\ }A+3.07{\pm}0.04,{\ }A{>}537{\ }\mathrm{km}^{2}.\] \end{document} These prediction equations of our bilinear model fit the Wells and Coppersmith (1994) data set well in their respective ranges of validity, the transition magnitude corresponding to A = 537 km2 being M = 6.71. Manuscript received 15 April 2001.

M-logA Observations for Recent Large Earthquakes

Abstract Using a magnitude ( M )-log area ( A ) dataset augmented with seven large ( M >7.0) earthquakes occurring since Wells and Coppersmith (1994), this short note assesses the current validity of the bilinear M -log A relations for continental, strike-slip earthquakes proposed by Hanks and Bakun (2002), in particular the L -model scaling at M >7. The relations determined by Hanks and Bakun (2002) are only insignificantly altered, leaving these bilinear M -log A relations as valid now as when first proposed.

A physical model for strain accumulation in the San Francisco Bay region: Stress evolution since 1838

[1] Understanding of the behavior of plate boundary zones has progressed to the point where reasonably comprehensive physical models can predict their evolution. The San Andreas fault system in the San Francisco Bay region (SFBR) is dominated by a few major faults whose behavior over about one earthquake cycle is fairly well understood. By combining the past history of large ruptures on SFBR faults with a recently proposed physical model of strain accumulation in the SFBR, we derive the evolution of regional stress from 1838 until the present. This effort depends on (1) an existing compilation of the source properties of historic and contemporary SFBR earthquakes based on documented shaking, geodetic data, and seismic data (Bakun, 1999) and (2) a few key parameters of a simple regional viscoelastic coupling model constrained by recent GPS data (Pollitz and Nyst, 2004). Although uncertainties abound in the location, magnitude, and fault geometries of historic ruptures and the physical model relies on gross simplifications, the resulting stress evolution model is sufficiently detailed to provide a useful window into the past stress history. In the framework of Coulomb failure stress, we find that virtually all M � 5.8 earthquakes prior to 1906 and M � 5.5 earthquakes after 1906 are consistent with stress triggering from previous earthquakes. These events systematically lie in zones of predicted stress concentration elevated 5–10 bars above the regional average. The SFBR is predicted to have emerged from the 1906 ‘‘shadow’’ in about 1980, consistent with the acceleration in regional seismicity at that time. The stress evolution model may be a reliable indicator of the most likely areas to experience M � 5.5 shocks in the future. INDEX TERMS: 1206 Geodesy and Gravity: Crustal movements—interplate (8155); 1236 Geodesy and Gravity: Rheology of the lithosphere and mantle (8160); 1243 Geodesy and Gravity: Space geodetic surveys; KEYWORDS: crustal deformation, plate boundary zones, viscoelastic relaxation, San Francisco Bay Region

mmi Attenuation and Historical Earthquakes in the Basin and Range Province of Western North America

Earthquakes in central Nevada (1932-1959) were used to develop a modified Mercalli intensity (MMI) attenuation model for estimating moment mag- nitude M for earthquakes in the Basin and Range province of interior western North America. M is 7.4-7.5 for the 26 March 1872 Owens Valley, California, earthquake, in agreement with Beanland and Clarks (1994) M 7.6 that was estimated from geologic field observations. M is 7.5 for the 3 May 1887 Sonora, Mexico, earthquake, in agreement with Natali and Sbars (1982) M 7.4 and Suters (2006) M 7.5, both estimated from geologic field observations. MMI at sites in California for earthquakes in the Nevada Basin and Range appar- ently are not much affected by the Sierra Nevada except at sites near the Sierra Nevada where MMI is reduced. This reduction in MMI is consistent with a shadow zone produced by the root of the Sierra Nevada. In contrast, MMI assignments for earthquakes located in the eastern Sierra Nevada near the west margin of the Basin and Range are greater than predicted at sites in California. These higher MMI values may result from critical reflections due to layering near the base of the Sierra Nevada.

Estimating Locations and Magnitudes of Earthquakes in Southern California from Modified Mercalli Intensities

Modified Mercalli intensity (mmi) assignments, instrumental moment magnitudes M , and epicenter locations of thirteen 5.6 ≤ M ≤ 7.1 “training-set” events in southern California were used to obtain the attenuation relation mmi = 1.64 + 1.41 M − 0.00526 * Δ h − 2.63 * log Δ h , where Δ h is the hypocentral distance in kilometers and M is moment magnitude. Intensity magnitudes M I and locations for five 5.9 ≤ M ≤ 7.3 independent test events were consistent with the instrumental source parameters. Fourteen “historical” earthquakes between 1890 and 1927 were then analyzed. Of particular interest are the M I 7.2 9 February 1890 and M I 6.6 28 May 1892 earthquakes, which were previously assumed to have occurred near the southern San Jacinto fault; a more likely location is in the Eastern California Shear Zone (ecsz). These events, and the 1992 M 7.3 Landers and 1999 M 7.1 Hector Mine events, suggest that the ecsz has been seismically active since at least the end of the nineteenth century. The earthquake catalog completeness level in the ecsz is ∼ M 6.5 at least until the early twentieth century. Online material : Figures and table of empirical mmi site corrections.

The December 1872 Washington State Earthquake

The largest historical earthquake in eastern Washington occurred on 15 December 1872. We used Modified Mercalli intensity (MMI) assignments for 12 twentieth-century earthquakes to determine attenuation relations for different regions in the Pacific Northwest. MMI attenuation for propagation paths east and west of the Cascade Mountains differs significantly only for epicentral distances greater than about 225 km. We used these attenuation relations and the MMI assignments for the 15 December 1872 earthquake to conclude that its epicentral region was east of the Cascade Mountains near Lake Chelan, Washington, and most probably near the south end of Lake Chelan. The intensity magnitude, M I, is 6.8 and moment magnitude, M , is 6.5-7.0 at the 95% confidence level. Manuscript received 23 October 2001.

Seismicity of California's North Coast

At least three moment magnitude (M) 7 earthquakes occurred along California9s north coast in the second half of the nineteenth century. The M 7.3 earthquake on 23 November 1873 occurred near the California-Oregon coast and likely was located on the Cascadia subduction zone or within the North American plate. The M 7.0+ earthquake on 9 May 1878 was located about 75 km offshore near the Mendocino fault. The surface-wave magnitude ( M S ) 7.0 earthquake on 16 April 1899 was located about 150 km offshore within the Gorda plate. There were at least three M 7 north-coast earthquakes in the 35 years before 1906, two M 7 earthquakes in the 20 years after 1906, no M 7 earthquakes from 1923 until 1980, and four M 7 earthquakes since 1980. The relative seismic quiescence after 1906 for M 7 earthquakes along California9s north coast mimics the post-1906 seismic quiescence in the San Francisco Bay area for M 6 earthquakes. The post-1906 relative quiescence did not extend to lower magnitudes in either area. The 18 April 1906 earthquake apparently influenced the rate of occurrence of M 7 north-coast earthquakes as it apparently influenced the rate of M 6 earthquakes in the San Francisco Bay area. The relative seismic quiescence along the California north-coast region after 1906 should be taken into account when evaluating seismic hazards in northwest California.

M–logA Models and Other Curiosities

Abstract Magnitude ( M )–log area ( A ) relations have been the focus of considerable research in the past two decades because of their importance in estimating moment magnitude M for earthquake probability calculations and seismic‐hazard analysis. For M W , there is a strong consensus for constant stress‐drop scaling. For the larger earthquakes ( M >7) that dominate the moment balancing in continental crust, the L ‐model scaling employed by Hanks and Bakun (2002, 2008) involves fault slip growing with fault length L when L > W ∼15  km or so, requiring that static stress drops increase with increasing fault slip. Constant stress‐drop representations of the same larger‐earthquake data, such as Shaw (2009, 2013), require slip at depths significantly greater than W ∼15  km. Available evidence supports neither of these requirements leaving us perplexed as to how large‐earthquake ruptures initiate and propagate in continental crust. Deep slip M –log A models that involve an unknown amount of seismic moment/earthquake slip at unknown depths> W are not appropriate for use in earthquake probability studies governed by shallow‐slip (depths≤ W ) seismic moment/earthquake slip balancing, such as those in California during the twenty‐first century.