Past Earthquakes Research Articles

Retrofit Strategies Using Braces for Pre-Northridge Steel Moment-Frame Buildings: An Analytical Study

ABSTRACT This paper focuses on retrofitting strategies for tall pre-Northridge steel moment-frame buildings with vulnerable beam-to-column moment connections. Ten retrofit schemes of a 20-story building, originally designed according to the 1994 UBC, are developed in accordance with ASCE-41 by upgrading the brittle beam-to-column moment-resisting connections, or adding conventional/buckling-restrained brace elements to the existing moment-frame bays. Using nonlinear pushover and time-history analyses under near-source ground motion records from past earthquakes and synthetic ground motion records from three large-magnitude scenario earthquakes, the retrofit schemes are compared against the baseline model. While the baseline model is a total loss in 1331 of the 4014 scenario earthquake analysis cases, the worst of the braced retrofit schemes experiences a total loss in fewer than a third of these cases (426), clearly demonstrating the effectiveness of adding braces.

Strong Shaking From Past Cascadia Subduction Zone Earthquakes Encoded in Coastal Landforms

AbstractStrong earthquakes along subduction zones are often devastating events, but sparse records along some tectonic margins limit our understanding of seismic hazards. Constraining shaking intensities is critical, especially in subduction zones with infrequent but large‐magnitude earthquakes like the Cascadia Subduction Zone (CSZ), where the lack of recorded ground motions has led to uncertainty in the severity and potential impacts of future earthquakes. Here we fill this observational gap with a novel inventory of quantitative estimates of past shaking intensities from geotechnical modeling of coastal landforms. One hundred fifty‐four deep‐seated landslides and 65 fragile geologic features constrain minimum and maximum peak ground accelerations, respectively. These estimates are broadly consistent with model predictions of M9 ruptures, suggesting strong shaking of 0.4–0.8 g during past CSZ earthquakes. Local discrepancies between our geologic shaking constraints and earthquake simulations may inform past rupture behavior, leading to better predictions of shaking intensity for future earthquakes.

Seismic fragility assessment of reinforced concrete wall buildings in Colombia: Insights and implications for earthquake-resistant design

Colombia is in a high seismicity area where the Nazca and South American plate converge. Past earthquakes in 1985 and 1999 caused considerable economic and human losses to the country. As part of a seismic risk mitigation strategy, the National Geological Service commissioned a nationwide seismic risk model in 2022, which required developing seismic fragility functions representative of the country building stock. This article presents the seismic fragility results for conventional and thin reinforced concrete (RC) wall buildings. Conventional RC walls are those fulfilling ACI 318 provisions, and thin RC wall buildings (TRCWBs) are those reinforced with a single layer of a welded wire mesh with limited ductility. Leveraging a database of 259 structural drawings and models, 91 building archetypes were created in OpenSeesPy using a modeling approach that approximates non-planar walls with the MVLEM-2D. These underwent nonlinear dynamic analyses with over 3000 hazard-consistent ground motions, recording inter-story drifts, floor accelerations, and element forces. The results were used to calculate seismic fragility functions that served as input for the application of the FEMA P-695 methodology. The findings of this study provide fragility parameters for buildings between 4 and 30 stories constructed in low, intermediate, and high seismic hazard in Colombia, which are suitable for development of seismic risk scenarios. In addition, results also suggest that the Colombian code provisions for TRCWB require adjustments.

Investigation of Passive Controlled Post-Tensioning System on the Structural Behaviour of Precast Reinforced Concrete Beam–Column Connections

Precast structures are widely used in many parts of the world. This construction technique is more commonly preferred for low-rise industrial buildings than multi-story structures. The most commonly used column–beam connection in precast buildings is the dowel connection (DC). Past earthquakes in various parts of the world have shown that these connections do not provide sufficient resistance. The main deficiencies of such connections are that they are sheared or stripped due to the shear force demand from the in-plane effects of large earthquakes, and that they do not provide sufficient resistance to the overturning moments from the out-of-plane effects of the earthquakes. Correspondingly, many prefabricated buildings have collapsed during earthquakes, causing loss of life and property. This study proposes using post-tensioning tendon (PT) systems and systems created by adding steel springs (PTS) to eliminate the weaknesses in column–beam connections in precast structures. To this end, real-sized column and beam specimens used in precast buildings were produced, and experiments were conducted under the cyclic loads defined by the American Concrete Institute (ACI) Committee, Report 374, simulating earthquake effects for three different connection types (DC, PT, and PTS). It was observed that the proposed PTS connection type dissipated approximately one-third of the energy transferred to the joint through elastic deformation in the springs, compared to the DC and PT connection types. This indicates that the PTS specimens transferred significantly less energy to the column–beam connection region. Consequently, the PTS system exhibited much less damage in the column foundation and especially the column–beam connection areas than other test specimens. In conclusion, it can be stated that the use of the PTS connection type in prefabricated structures has high potential to reduce damages due to dynamic loads such as earthquakes.

EFFECT OF EARTHQUAKE DIRECTIONALITY ON THE SEISMIC RESPONSE OF RC WALL BUILDINGS

Reinforced concrete (RC) wall buildings are widely used for their proven seismic resilience, as evidenced in past earthquakes. To understand their behavior, extensive research has been conducted using both experimental and numerical methods. Despite these efforts, the impact of earthquake directionality on these structures deserves further exploration. Addressing this gap, this paper examines the effects of earthquake directionality on three RC wall buildings. Eleven ground motion records from the far-field suite of the FEMA P695 were selected, rotated, and used to perform incremental dynamic analyses of the three buildings in OpenSeesPy. The buildings seismic performance was evaluated in terms of the roof drift ratio, axial load ratio, and bending moments. Additionally, a novel performance metric, namely the MaxRotEDPpp is introduced to assess the building’s response. The results show that the direction of ground motions has a significant impact on engineering demand parameters and in the probability of exceedance for different thresholds, which can double compared to non-rotated ground motions.

Quantitative crack evaluation in slender reinforced concrete walls with rectangular section

Past earthquakes have shown that cracking affects post-earthquake functionality and accounted for huge repair costs for reinforced concrete (RC) wall buildings, even though the code-compliant seismic design prevents collapse. Engineers should know the maximum residual flexural crack width and volume of repair material needed for the flexural cracks to determine the damage degree and the repair cost. This paper presents the experimental campaign on four RC slender walls that investigated the effect of confining reinforcement and thickness of the wall on flexural crack parameters under quasi-static reversed cyclic loading. The width of all flexural cracks was measured when reaching each cycle peak drift and when unloading to zero lateral loads. Crack widths at peak and residual states increased with increasing peak drift. Based on the experimental observations, it was found that the maximum residual crack width is obtained as a simple function of the extreme tension fiber elongation of the wall tensile fiber within ±30% error. In addition, this paper outlines methods to calculate the volume of repair material for flexural cracks from the extreme tension fiber elongation of the wall. With the fundamental rules found from the experiment in this paper, it will become possible to obtain the maximum crack width and the volume of repair material from simple numerical analysis tools such as a multi-spring line element model.

Spatiotemporal Clustering of Large Earthquakes Along the Central‐Eastern Sections of the Altyn Tagh Fault, China

AbstractThe understanding of the spatial‐temporal distribution of past earthquakes is essential to assess the event recurrence behavior and to estimate the size of potential earthquakes along active strike‐slip fault systems. However, the scarcity of paleoseismic data remains a major hurdle in this endeavor. This is the case of the longest strike‐slip fault in Asia, the Altyn Tagh Fault (ATF). We documented six very likely large earthquakes that potentially ruptured the Aksay section of the ATF. Employing a Bayesian approach, we present modeled date ranges of 6339–5220 BC, 5296–4563 BC, 3026–2677 BC, 1324–808 BC, 314–632 AD, and 915–1300 AD. The mean recurrence time is 1,329 ± 588 years with a coefficient of variation (COV) of ∼0.44. In the same fault section, 90 horizontal offsets record an average coseismic slip of 5.1 ± 1.4 m for the last event and suggest four older earthquakes plausibly with a similar slip distribution. Although at the local‐scale the COV indicates quasi‐periodic rupture behavior, the individual interevent times exhibit significant irregularity, a pattern also observed in adjacent fault sections (Xorxoli, Annanba and Tashi sections). We found that such irregularities are a natural consequence of long‐term fault interactions, which allow for synchronized ruptures along the northern and southern strands of the central‐eastern ATF. Our rupture model highlights bursty periods of seismic activity with mean interevent times of 475 ± 108 years separated by long‐lull periods of 1.1–1.6 kyr. Based on this temporal organization and considering the 401‐year elapsed time since the most recent event on the Xorxoli section, there exists a possibility of a forthcoming large earthquake occurring within the next century.

Earthquake Risk Perception and Preparedness of a Sample of Residents Following a Major Earthquake in Türkiye in 2023

The aim of this study was to investigate these residents’ levels of earthquake risk perception and preparedness following the disastrous earthquake event on 6 February 2023 near Kahramanmaraş in Türkiye. The study involved a cross-sectional descriptive design. A sample of convenience comprising 411 residents of areas not impacted directly by the 6 February 2023 earthquakes completed an online survey over a three-month period March to May 2023. There was no indication of notably elevated levels of earthquake risk perception among those residents surveyed overall. Levels of physical, or material, preparedness for earthquakes were lower than desirable. Earthquake risk perception was negatively, though weakly, related to both physical and psychological preparedness. Physical preparedness was strongly and positively correlated with psychological preparedness. Having (a)past earthquake experience (b)read or viewed earthquake safety material, (c)attended earthquake safety meetings and (d)work experience related to emergencies were all associated with significantly higher levels of residents’ preparedness. Although the information was collected a short time after a disastrous earthquake event when overall levels of community awareness of the danger posed by earthquakes were likely to be high, the findings about the levels of a possible earthquake preparedness were not satisfactory. Possible implications for improving community earthquake preparedness are discussed.

Seismic loss comparison for buildings designed with ductile steel seismic force-resisting systems and with controlled rocking braced frames

Observations from past earthquakes have highlighted the structural damage and significant residual deformations experienced by ductile steel seismic force-resisting systems (SFRSs), such as special moment resisting frames (SMRFs), special concentrically braced frames (SCBFs), and buckling-restrained braced frames (BRBFs). To mitigate these challenges, controlled rocking braced frames (CRBFs) have emerged as a promising low-damage alternative SFRS. However, concerns have been raised about whether the reduction in structural damage with CRBFs may come at the cost of increased acceleration demands and associated nonstructural damage. This study offers a comprehensive investigation of such trade-offs by analyzing three buildings of different heights, each designed with the three ductile SFRSs identified above and with CRBFs. After examining the structural response at different earthquake intensities, the focus of the paper is on earthquake-induced economic losses. Among the considered SFRSs, greater total expected annual losses (EAL) are observed in the SMRF and SCBF buildings, primarily due to demolition losses and repairable losses, including repairs of structural and nonstructural components. The total EAL is lower for the BRBFs and lowest for the CRBFs, with the losses in the BRBF buildings primarily attributed to demolition loss, considered as irreparable loss, while the losses in the CRBF buildings are mainly due to acceleration-sensitive nonstructural components, considered as reparable loss. To provide a more detailed comparison, cost-effectiveness analyses are also performed, indicating that a modest cost premium for CRBFs is justified to reduce earthquake economic costs over the building lifetime.

Psychometric Properties of Turkish Adaptation of the Environmental Risk Coping Scale

Objective: This study aims to adapt Environmental Risk Coping Scale into Turkish and examine its psychometric properties. Method: For this aim, the psychometric properties of the scale adapted into Turkish were tested in a sample of 230 participants living in 6 cities with high earthquake risk in Türkiye. Participants responded to the Environmental Risk Coping Scale, Environmental Risk Perception Scale, questions on Present Fatalistic and Future Time Orientation, and a demographic information form (i.e., age, gender, education level, city of residence, and homeowner/renter status, past earthquake experience, extent of earthquake damage). Results: The findings of the analyses (i.e., confirmatory factor analysis, reliability analysis) showed that this 12-item scale is reliable and valid in the Turkish sample. Specifically, desirable fit indices (χ2 / sd = 2.06, p < .001, CFI = .96, TLI = .95, RMSEA = .07, 90 % CI [.05, .09], SRMR = .04) demonstrated that the confirmatory factor analysis confirmed the two-factor structure (problem focused and emotion focused coping strategies) and Cronbach α values (.89 for 9-item problem focused coping strategies factor and .72 for 3-item emotion focused coping strategies factor) indicate that the internal consistency of the scale is high. In addition, the sub-dimensions of the Turkish version of the scale were correlated with variables such as risk perception, present-fatalistic time orientation, future time orientation, and demographic factors such as age, earthquake experience, and the extent of damage in earthquake(s) in line with the literature. Conclusion: The Turkish adaptation of the Coping with Environmental Risk Scale was found to be a reliable and valid scale in the Turkish sample.

Erratum to Studying Past Earthquakes with Modern Techniques: Ground-Motion Simulations for the 11 January 1693 Noto Earthquake in Italy

<i>Erratum to</i> Studying Past Earthquakes with Modern Techniques: Ground-Motion Simulations for the 11 January 1693 Noto Earthquake in Italy

Assess seismic resilience of unreinforced stone loess cave retrofitted with composite materials by shaking table tests

Unreinforced masonry (URM) structures have consistently suffered significant seismic damage in past earthquake events, leading to considerable losses in both life and property. Unreinforced stone loess cave (URSLC), a type of URM structure, are commonly used in the loess regions of northwestern China due to their affordability for low-income populations. Developing cost-effective and easily implementable reinforcement technologies to enhance the seismic performance of URSLC is therefore critical. In this study, a series of shake table tests were conducted to improve the strength and ductility of URSLC. Initially, the unreinforced structure was subjected to six progressively intense bidirectional seismic excitations, driving the structure to near-collapse. Subsequently, the damaged structure was retrofitted using composite materials, including steel bar/wire mesh mortar, steel belts, and epoxy resin. To compare the performance, a shaking table test was repeated on the retrofitted structure using the same seismic input as the initial test. Throughout the experiments, the damage progression, crack development, and failure characteristics of both the unreinforced and reinforced specimens were meticulously recorded. Additionally, dynamic properties before and after reinforcement were analyzed to assess the effectiveness of the retrofitting. The results demonstrated that the composite materials provided a strong bond between the stone layers and reinforcement, significantly improving the seismic resistance of the damaged loess caves. Moreover, the use of these materials was found to be both efficient and economical in enhancing the structural performance—particularly in terms of strength, stiffness, and damage tolerance. This study confirms that composite reinforcement can significantly improve the safety of masonry buildings, even under severe seismic loading conditions.

Limited Preservation of Strike‐Slip Surface Displacement in the Geomorphic Record

AbstractOffset geomorphic markers are commonly used to interpret slip history of strike‐slip faults and have played an important role in forming earthquake recurrence models. These data sets are typically analyzed using cumulative probability methods to interpret average amounts of slip in past earthquakes. However, interpretation of the geomorphic record to infer surface slip history is complicated by slip variability, measurement uncertainty, and modification of offset features in the landscape. To investigate how well geomorphic data record surface slip, we use offset measurements from recent strike‐slip surface ruptures (n = 39), faults with geomorphic evidence of multiple strike‐slip earthquakes (n = 29), and synthetic slip distributions with added noise (n10,000) to examine the constraints of the geomorphic record and the underlying assumptions of the cumulative offset probability distribution analysis method. We find that the geomorphic record is unlikely to resolve more than two paleo‐slip distributions, except in specific cases with low slip variability, high slip‐per‐event, and semiarid climate. In cases where site‐specific conditions allow for interpretation of more than two earthquakes, lateral extrapolation along a fault is not straightforward because on‐fault displacement and distributed deformation may be spatially variable in each earthquake. We also find that average slip in modern earthquakes is adequately recovered by probability methods, but the reported prevalence of strike‐slip faults with characteristic slip history is not supported by geomorphic data. We also propose updated methods to interpret slip history and construct uncertainty bounds for paleo‐slip distributions.

An evolutionary approach to predict slope displacement of earth embankments under earthquake ground motions

Accurate slope stability analysis of earth embankments under ground shaking is of great importance for practical use in earthquake geotechnics. This study aims to predict soil slope displacements of earth embankments subjected to earthquake loading using evolutionary algorithms. Comprehensive real case histories of slope displacement of earth embankments under past earthquakes in different areas of the world were gathered and analyzed. A robust model was then developed to predict earthquake induced soil slope displacements using gene expression programming (GEP). Characteristics of earthquake ground motion including earthquake magnitude, earthquake predominant period, maximum earthquake acceleration and also geotechnical specifications of earth embankment including yield acceleration and fundamental period of earth embankment were taken as most influential factors on the slope displacements of earth embankments under earthquakes. Subsequently, performance of developed GEP-based predictive model was assessed using a sensitivity analysis under various effective factors. Finally, the accuracy of the predictive model was evaluated through comparison with the available relationships for estimation of seismic soil slope displacements. The results clearly indicate favorable accuracy of developed GEP-based model to predict slope displacements of earth embankments subjected to earthquake ground motions.

Developing Liquefaction Prediction Equations for Gravels Based on DPT Blow Count and Shear Wave Velocity at Field Case History Sites

Gravelly soils have liquefied at multiple sites in at least 20 earthquakes over the past 130 years. These gravels typically contain more than 25% sand which lowers the permeability and makes them susceptible to liquefaction. In other cases, a low permeability layer at the surface impedes drainage and causes liquefaction. Typical SPT- or CPT-based liquefaction correlations can be affected by large gravel particles and lead to erroneous results. To deal with these problems, we have developed liquefaction triggering curves for gravelly soils based on (1) shear wave velocity (Vs) and (2) a 74-mm diameter dynamic cone penetrometer (DPT), that are less affected by gravel particles. These correlations are based on case histories where gravel did and did not liquefy in past earthquakes. The VS-based liquefaction triggering curves for gravels shift to the right relative to similar curves based on sands. Good agreement has been obtained with liquefaction resistance from the DPT and the CPT when these methods could all penetrate the gravel in Wellington New Zealand. Correlations are developed between Vs and blow count from the DPT.

Estimation of Vs30 and site classification of Bhaktapur district, Nepal using microtremor array measurement

The severe damage observed in the Kathmandu Basin, Nepal, during past earthquakes necessitates a thorough study of the seismic behavior of the basin sediments. As the shear-wave velocity is directly related to the elastic shear modulus of the material, it is essential to determine it to incorporate the behavior of the soil in the design of the structure. Hence, we determined average shear-wave velocity in upper 30 m (Vs30) of soil in Bhaktapur district in the eastern part of the Kathmandu Basin at 73 observation points, employing two methods involving the use of non-invasive microtremor array measurements (MAMs). These MAMs are widely used for determining subsurface soil characteristics by analyzing the ambient vibrations of the ground. The first method involves inversion using a genetic algorithm, and the second is a method for obtaining Vs30 directly from the dispersion curve. We found that Vs30 in the southeastern part of the study area was higher than that in other parts. Conversely, Vs30 in the western region was lower. The calculated Vs30 values were used to classify the sites. The elevated eastern and southeastern areas with high Vs30 were categorized as dense soil or soft rock, whereas the areas with low Vs30 that had suffered significant damage during the 2015 Gorkha earthquake were classified as soft soil sites.Graphical

Spatiotemporal evolution of post-seismic deformation caused by large earthquakes around the Pacific and its impact on plate movement

SUMMARY The fully relaxed deformation caused by an earthquake refers to the total deformation of the earth caused by the earthquake, including the coseismic deformation and the total effect of post-seismic stress relaxation in the mantle on crustal deformation. An in-depth investigation into post-seismic and fully relaxed deformation resulting from large earthquakes is conducive to comprehending the feedback effect of earthquakes on plate motion. In this paper, we first calculated theoretically the coseimic, post-seicmic and fully relaxed deformations caused by the Tohoku-Oki Mw9.0 earthquake using the spherical dislocation theory. The post-seismic displacement caused by the great earthquake leads to the continuous convergence of plates on both sides of the seismogenic fault, which will certainly facilitate the downward insertion of the subduction plate. As time goes by the influence range of seismic deformations becomes larger and larger. The Tohoku-Oki Mw9.0 earthquake can produce &amp;gt;5 cm of cumulative post-seismic horizontal displacement at a far place like the East Pacific Rise over 100 millenniums. Then, we built slip models of 375 earthquakes with Mw7.0 and above in the circum-Pacific seismic belt, calculated the cumulative post-seismic deformations of them, and found that significant post-seismic horizontal displacements covered the entire Pacific Ocean. The post-seismic deformation field caused by the 2011 Tohoku-Oki Mw9.0 earthquake, the 2010 Chile Mw8.8 earthquake, the 1964 Prince William Sound Mw9.1 earthquake and the 1960 Chile Mw(9.3–9.4) earthquake determines the overall distribution pattern of the deformation field in the Pacific region. Those large earthquakes around the Pacific make the total Pacific Plate present a tension strain in the northwest–southeast direction. The one at the East Pacific Rise reaches 300–400 nstr, with orientation of principle strain approximately perpendicular to the Rise. This tensile strain is bound to encourage transverse expansion of the mid-ocean ridge. By a logical extension, the post-seismic stress relaxation in the mantle caused by past earthquakes should be an important driving force for the current plate movement, in addition to the classic driving force like negative buoyancy and plate material phase transition. This study proves theoretically that there is a two-way relationship between great earthquakes and plate movement, and the viscous structure of mantle plays a key role in the relationship.

Tectonic activation and the risk of Ilisu Dam collapse to Iraq through modelling and simulation using HEC-RAS

Floods caused by dam failures can cause huge losses of life and property, especially in estuarine areas and valleys. In spite of all the capabilities and great improvements reached by man in the construction of dams and their structures, they will remain helpless before the powerful forces of nature, especially those related to tectonic activation, and the occurrence of earthquakes of different intensities.The region extending from the Ilisu Dam in Turkey to the Mosul Dam in Iraq was chosen as an area for this study, and the HEC-RAS application was used to simulate the collapse of the Ilisu Dam due to a major earthquake, to know the magnitude of the risks and losses that could result from this. The Ilisu Dam was built very close to a highly tectonically active fault system, particularly the East Anatolian Fault (EAF), which is one of the largest tectonically active faults in the world with a length of 500 km. This region has witnessed past and present earthquakes of high magnitude (M > 7), especially in the EAF, so the construction of the Ilisu Dam near the EAF fault system is of great concern, as it was built in a basin with very complex seismic activity and geology.Using the HEC-RAS simulation application, the study found that the flood resulting from the collapse of the Ilisu Dam would reach the edges of the Mosul Dam Lake in just 13 h. With a flow of more than 100,000 m3/s, more than 10 billion m3 of water will flow into the Mosul Dam Lake within four days of the disaster. This will lead to the collapse of the Mosul Dam and direct the flood wave of the collapse of these dams towards Baghdad through Mosul, Tikrit, and Samarra. This could pose risks to all Iraqi cities located within the Iraqi sedimentary plain (Mesopotamia), from south of the Mosul Dam up to Basra, in a scenario similar to Noah’s Flood.

Assessing the Adequacy of Earthquake Catalog Sampling for Long-Term Seismicity in Low-to-Moderate Seismic Regions: A Geodetic Perspective

Abstract Seismic hazard assessment in low-to-moderate seismicity regions can benefit from the knowledge of surface deformation rates to better constrain earthquake recurrence models. This, however, amounts to assuming that the known seismicity rate, generally observed over historical times (i.e., up to a few centuries in Europe), provides a representative sample of the underlying long-term activity. We here investigate how this limited sampling can affect the estimated seismic hazard and whether it can explain the disagreement between the seismic moment loading rate as seen by nowadays Global Navigation Satellite Systems (GNSS) measurements and the seismic moment release rate by past earthquakes, as is sometimes observed in regions with limited activity. We approach this issue by running simulations of earthquake time series over very long timescales that account for temporal clustering and the known magnitude–frequency distribution in such regions, and that those are constrained to a seismic moment rate balance between geodetic and seismicity estimates at very long timescales. We show that, in the example of southeastern Switzerland, taken here as a case study, this sampling issue can indeed explain this disagreement, although it is likely that other phenomena, including aseismic deformation and changes in strain rate due to erosional and/or glacial rebound, may also play a significant role in this mismatch.

Rupture direction of paleoearthquakes on the Alpine Fault, New Zealand, as recorded by curved slickenlines

Abstract We observed and further exhumed curved slickenlines on fault planes associated with paleo-surface rupture of the Alpine Fault, New Zealand's ∼30 mm/yr continental transform plate boundary. Dynamic rupture modeling indicates that the geometry of such curvature provides a record of past earthquake rupture directions. We focused our efforts on three sites that span a region known to variably halt or allow passage of past earthquakes (an “earthquake gate”) to contribute rupture direction constraints to the fault's spatiotemporally rich paleoseismic record. At Hokuri Creek and Martyr River, we observed both convex-up and convex-down curved slickenlines on and adjacent to principal slip surfaces, indicating past ruptures from both the northeast and southwest of these locations. At Martyr River, relationships suggest that the most recent event (inferred to correlate to 1717 CE) ruptured from the southwest. Our results demonstrate the utility of curved slickenlines as a valuable new paleoseismological tool for determining past rupture directions, applicable to surface-rupturing faults globally.