the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
The impact of pre-existing weaknesses on strike-slip fault evolution: insight into strain partitioning of the 2019 Ridgecrest earthquake
Abstract. The 2019 Ridgecrest mainshock produced a complex rupture pattern at the northwest end of the dextral rupture where a set of disconnected pre-existing faults that trend perpendicular to the strike of the mainshock fault had sinistral slip. This highly segmented geometry of active faulting may reflect immature faulting that provides insights into the early development of strike-slip faults. Physical experiments that simulate upper crustal deformation using scaled analogue materials, such as wet kaolin, allow us to control loading and material rheology, and directly document the complete evolution of strike-slip fault systems that grow in material with pre-existing weaknesses. To assess the impact of pre-existing weaknesses on strike-slip fault evolution we vary initial orientation and spacing of the vertical surfaces, and nature of basal shear loading (localized and distributed). Weaknesses oriented 60˚ and 90˚ from the applied dextral loading showed negligible slip while weaknesses oriented 120˚ developed sinistral slip and weaknesses oriented 150˚ had dextral slip. Experiments that developed sinistral slip along cross-faults (120˚) also showed distributed dextral strain between the faults that contributed to significant rotation of material, including the cross-faults, within the shear zone. When the sinistral faults were rotated to orientations unfavourable for continued slip, new dextral faults developed. This finding suggests that strike-slip systems with active cross faults represent immature stages of evolution that will become reorganized upon further strain accumulation. The amount of off-fault deformation and shear zone width depends on the presence of pre-existing weaknesses (even if they had low slip) and the persistence of fault irregularities that arose from slip along and interaction of new faults with the pre-existing weaknesses. Understanding how the orientation of pre-existing weaknesses influences the early evolution of strike-slip faults and strain localization over geologic time scales can inform future seismic hazard assessments of regions with pre-existing structures.
- Preprint
(1662 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-1948', Anonymous Referee #1, 22 Jul 2026
-
AC1: 'Reply on RC1', Michele Cooke, 10 Nov 2026
Responses are in bold
Thank you for the time and care that you gave this review. We appreciate your thoughtful suggestions.
General Comments:
- Consider using “localized” and “distributed” deformation rather than “on-fault” or “off-fault” deformation because all of the deformation happening around the active faults in the models is resulting from the fault. Therefore, it is not “off-fault” deformation, which cannot be attributed to this fault, but rather deformation attributed to the fault that is more distributed. In this framework, rupture on the primary fault trace can then be called “localized” instead of “on-fault”. See paragraph lines 158-170, and throughout the rest of the manuscript. In real surface ruptures, distributed and localized deformation are becoming the preferred terminology (see for example Antoine et al., 2026, https://doi.org/10.1016/j.epsl.2026.119868) because “off-fault” can then be reserved for deformation that occurs far enough away as to not be associated with the fresh rupture. In a hazard model sense, “off-fault” deformation is deformation in the budget that cannot be attributed to a fault given the slip rates of the faults in the model and the spatial distribution of surface strain.
This is a very helpful suggestion and perspective. I’ve noticed for a while now different researchers have meant different things when referring to ‘off-fault deformation’, which has led to confusion. While most of the distributed strain the experiments is due to the fault geometry, the viscoelastic nature of the clay exhibits an element of flow that also contributes to distributed shear. For example, if we cut a perfectly straight strike-slip fault, we still see some distributed deformation due to the viscoelasticity of the material (~5% Hatem et al. 2015 (JSG) and ~10 -15% Elston and Cooke, in prep). Note that Hatem et al., 2015 used a different threshold for fault detection which accounts for the difference in reported distributed strain. This viscous flow of the kaolin may be a proxy for the inelastic crustal processes described in Antoine et al., 2026. For this reason, we keep some (more limited) use of off-fault deformation in the text. The terms localized and distributed are very helpful and we’ve altered the text to use those terms when explain the impact of fault geometry.
- In the discussion, it would be neat to also see an analysis or commentary on the Ridgecrest Mw 7.1 rupture’s primary strand geometry and if it is considered to be in Stage 1, 2, or 3, and how that relates to the cross faults at the NW end of the rupture. Alternatively, or in addition, it would be interesting and informative to do an analysis of relative slip magnitude (or percentage) on the primary fault versus cross faults at different stages of fault evolution in the claybox models and compare it to the slip amounts on different fault strands in the Ridgecrest mainshock. This would strengthen the study even further and make it more applicable to other surface ruptures.
Good suggestion. Added a new figure of total sinistral and dextral slip on the experiment with 120˚ precut planes with 2cm spacing over localized basal shear. This experiment has many features of the Ridgecrest main shock and provides helpful framework for applying the results to this fault system.
- The section 4.1 – Repeatability seems like it would fit better in results than the discussion. Similarly and related, Table 1 would fit better in the methods section, and it would make the model setup and parameters easier to digest prior to the results.
We have implemented this helpful suggestion.
- It is a bit confusing to call the pre-cut weaknesses “surfaces” rather than “planes” (see paragraph lines 192-198, and later in the manuscript). It is common to discuss faults as planes, and using “surfaces” can be confused with surface rupture or surface faulting, which this is not. Although it is the author’s choice in the end, perhaps consider using “plane” to describe the pre-cut weaknesses.
We have implemented this helpful suggestion.
Minor Comments:
Line 6 – Add Mw 7.1 to the description of the Ridgecrest mainshock | corrected
19 – rephrase “if they had” to “those with” | corrected
22 – Might consider rephrasing the final sentence of the abstract as the pre-existing structures may be unknown prior to rupture.
Adjusted the text to reflect that the results of the study are useful where pre-existing weaknesses are known or suspected
24-25 – Name the historical earthquakes with complex fault networks that are being referred to here so that the “these recent earthquakes” and “these ruptures” in lines 26 and 28 are referring back to actual earthquakes.
Implemented the suggestion
Figure 1 caption - States the maps are “taken” from Rosa et al., 2024, and Milliner et al., 2021. Are the data from those publications or the figures themselves? If the figures themselves, make sure proper permissions are acquired and state instead “figure from”. If just the data, then state “data from”.
Wording changed. We acquired permissions as part of the submission process. Miliner is open access as is Rosa as it is a state government document (CA geologic survey).
62 – Add a comma between “clay, which” | corrected
86 – delete the “to” between “Fault systems evolve” | corrected
92 – Put the “in the crust” after “weaknesses” instead of after “bedding planes” | corrected
95 – Consider adding an introductory sentence to this paragraph something along the lines of “We test four cases of strike-slip fault evolution: with and without basal shear and with and without pre-existing weaknesses”. Then go ingo the details of the claybox model. Also, instead of “All strike-slip fault experiments…” consider “Our strike-slip” or “All our…”.
We implemented the helpful suggestion for an initial sentence. We prefer not to say ‘our models’ because the experiments and results do not belong to us in a proprietary sense.
121-122 – The sentence “We expects…dextral slip (Figure 3)” is more of a hypothesis than methods. Consider rephrase.
Text has been reworded to improve clarity
130 – “several other” is vague. Are there more relevant details?
Added a note that a section of the results presents the repeated experiments.
149 – Add a “the” in “impact the number of detected faults” | corrected
151 and 153 – “5x5 mm^2” and “(area <6 mm ^2)” both are areas already, and do not need the ^2.
We opt to keep the units as is to keep them consistent with area
176 – It seems a word is missing after “basal” | corrected
177 – It seems 1 or 2 is missing after “stage” | corrected
185 – Weaknesses (plural) | corrected
91 – in the title: “…with 2 cm weakness spacing” | implemented suggestion
202 – “experiment with 90 degree weaknesses” | corrected
Figure 4 – consider adding an arrow or few to point out the pre-cut weaknesses in one of the panels
These planes are now pointed out in the first series of photos with strain overlay maps.
Figure 5 caption – The x axis in the figure is “displacement” but the figure caption calls it “timing”. I understand how they are linked, but consider using “displacement” in the caption as the figure does not have a time component labelled.
Implemented suggestion
238 – update to “This reorganization of the active FAULT system marks the START of stage…” | corrected
247 – add a reference to Figure 6 somewhere in the first sentence of this paragraph.
Implemented suggestion
260 – Add a comma between “…faults, none…”. | corrected
267 and 290 – sinistral and dextral are misspelled | corrected
274, 281, and throughout – use “less” instead of “lesser”. | implemented
284 – what kind of fault irregularities - geometrical? | changed for clarity
295-296 – update to: “…as the applied loading was accommodated as DISTRIBUTED deformation and kinematic efficient increased as the faults evolved and accommodated MORE LOCALIZED slip” | implemented
Figure 8 – what do the light tan and dark tan regions represent in panel A? How are the onset of each stage defined – is it manual my inspecting the deformation patterns, or is there a quantitative threshold?
The light tan region is now explained in the caption as the standard deviation of kinematic efficiency. The dark tan was just highlighting the increasing fault slip. This isn’t needed so we removed for clarity.
How we assess the onset of stages is explained at the very start of the results section. We have added a sentence about our visual inspection of animations available in the data repository.
315 – replace “that” with “than” | corrected
334 – Add a comma between “displacement, slip” | corrected
341 – Add a comma after “stage 2,” | implemented
353 – missing final ) after “stage 1” | corrected
378 – add a comma after “During stage 2,” | corrected
393 – “irregular” rather than “irregularly” | corrected
400 – add a comma after “pre-cut surfaces,” | implemented
426 – “rheometer” what a great word!
432/Figure 10 caption – Where do the data for part B, “real-world orientations” come from?
We use the phrase ‘real world’ orientation to explain that the angles represented on the Mohr diagram (a) are double those in the ‘real world (b)’. This figure does not use field data. We have rewritten the caption for part b to remove this unclear qualifier.
437-440 – Completely agree!
454 – “…the shear zone that resembles the pattern…” | corrected
464 – “contribute TO curved cross-faults” | corrected
474 – “production OF simultaneously active…” | corrected
484 – “this early fault irregularity IMPACTS”. | corrected
495 – irregular rather than irregularly | corrected
500 – less rather than lesser | implemented
505 – update to “increasingly unfavorable”. | corrected
507 – consider adding a sentence or half a sentence about how these results might also inform PFDHA around immature faults – e.g., the results could support recommending wider deformation zones around immature faults.
Very helpful suggestion. We have added some text with this recommendation.
Citation: https://doi.org/10.5194/egusphere-2026-1948-AC1
-
AC1: 'Reply on RC1', Michele Cooke, 10 Nov 2026
-
RC2: 'Comment on egusphere-2026-1948', Anonymous Referee #2, 09 Oct 2026
Motivated by observations from the Ridgecrest earthquake sequence, this study investigates how pre-existing weak surfaces influence fault zone development using analogue model experiments. The models use wet kaolin, an analogue material that has been extensively tested and characterised in previous studies. The authors examine how the surface displacement field evolves with progressive shear across a suite of experimental configurations that vary the orientation and spacing of pre-cut surfaces and whether shear is imposed along a localised basal break or distributed across a zone at the base of the model.
The role of pre-existing weaknesses in fault zone development is a long-standing question with broad implications. I found the paper to be well written and well supported by the figures. The rationale for this study is clear, the experiments were well executed, and the results are presented effectively. The main area where I think the paper could be strengthened is in more clearly establishing the relationship between the experiments and the Ridgecrest setting:
- The discussion suggests that the Ridgecrest cross-faults may be near the end of their activity given the internal dextral fault development. One key difference is that the Ridgecrest faults are orthogonal, whereas the internal dextral faults in the experiments (Figs. 4k and 7h) are oblique. Do the experiments ever show the faults becoming orthogonal as they are at Ridgecrest?
- Extending the experimental results to Ridgecrest would imply that the cross-faults (the NE-SW oriented faults parallel to the foreshock) were pre-existing and the mainshock-oriented faults developed later. The Independence Dike Swarm is thought to be a significant pre-existing structure at Ridgecrest (Nevitt et al., 2023) and is NW-SE oriented, parallel to the mainshock. Is there another pre-existing structural fabric in the cross-fault orientation to support the experimental interpretation of cross-fault development at Ridgecrest? Can you speculate how the occurrence of multiple pre-existing structures in differing orientations might affect fault zone development?
Other areas where additional discussion could broaden the impact include:
- Are pre-cut surfaces or distributed basal shear novel developments for these types of analogue models? If so, I suggest highlighting these methodological advances more explicitly. It would also be helpful to provide some additional explanation of the motivation for testing distributed basal shear and what aspects of natural fault deformation this configuration might represent.
- One of the results that I find most interesting is that the occurrence of pre-existing structures can hinder fault development rather than facilitate it. Some exploration of this concept in the discussion could be useful and broaden the impact.
- Because fault ‘maturity’ plays an important role in your interpretations, it would be helpful to define what you mean by this term. Fault maturity is commonly defined in terms of cumulative displacement, and I suggest distinguishing this from other characteristics, such as the age of the structures (which could be quite old at Ridgecrest given the evidence for structural inheritance). It would also be helpful to compare how the cumulative displacement in your experiments compares with cumulative displacement at Ridgecrest (if known).
There are a number of typos and grammatical errors that I identify in the line-by-line comments. These are easy fixes and did not hinder my understanding of the manuscript.
Line-by-line comments and queries
22 – Very nice abstract!
29 – M 7.1
53 – upper crust
58 – is a citation needed for the digital image correlation techniques?
82 – With the careful scaling of the model, do you expect gravitational effects to be equivalent to those in the Earth? I am particularly interested in whether they produce flower-structure-like distributions of plastic strain near the model surface due to confining pressure effects. This may be a bit beyond the scope of this paper and more a question about scaled analogue models in general.
86-87 – incomplete sentence. I think this should read ‘Fault systems evolve with both…’
88 – Does the electrified probe impart any fabric in the materials adjacent to the cuts? In general, is there any pre-existing grain alignment/fabric, whether from scraping the clay or cutting the surfaces, that might influence the results?
102 – State that the 2.5 cm is equivalent to 2.5-5 km depth. What is the rationale for this depth? It seems quite shallow compared to average crustal thickness, but perhaps you are considering the locally high geothermal gradient at Ridgecrest.
116 – I believe this should be pixels/cm^2
122 – Capitalise Coulomb
124, 128– include what these distances, 2 cm and 5 cm, correspond to in nature. Also for 60 mm.
129 – 30 mm, not cm
140 – ‘and enlarged the…’ (delete the extra ‘an’)
142 – ‘decreasing’ may be clearer than ‘descending’
174 – I suggest that rather than providing a general description of all experiments here, make it specific to the uncut experiment. Otherwise, I don’t see a description anywhere for the uncut experiment to be compared with the pre-cut ones. The figure call-outs are for the uncut experiment, but the text isn’t as clear about this. You could start the description with ‘In our uncut experiment, prior to the emergence…’ to make this clear. Also report the fault orientations for comparison with the cut experiments.
176 – develop over the basal … missing word?
176-177 – During stage 2?
179 – ‘formed’ marks a shift to past tense, whereas the first part of the paragraph is present tense. Consider staying in present tense.
203 – ‘surface’ should be singular.
210 – ‘Red circles highlight’ … delete unnecessary ‘on’
230 – ‘greater strain’
235 – ‘within the shear zone, as documented…’ delete unnecessary ‘such’
241 – ‘With additional applied strain, the faults within the experiment’ … not ‘Within’; add ‘the’ before ‘experiment’
253-257 – I believe these figure call-outs should be to Figure 6 not Figure 4.
255 – ‘precut surfaces … the fault pattern’ instead of ‘precut surface … the faults pattern’
256 – should this be ‘some new dextral faults’?
279-280 – could you reframe this as ‘closer spacing of pre-existing surfaces impedes fault growth or development’?
296 – Exceptions are for the distributed basal shear experiments with 120 and 150 cuts, which appear to possibly become less kinematically efficient by the end of the experiment (Figure 6). At least, it’s not as clear a progression to me. Can you provide a brief explanation of this here?
308-309 – I believe this should be ‘left-stepping’
313 – consider specifying ‘off-fault deformation at the model surface.’
335 – ‘which increased the…’
Figure 9 – there is a jagged black curve that is cut off by the x-axis starting at ~25 mm. I think this should not be there.
366 – delete ‘until’
393 – ‘in the irregular geometry’
425 – You could reference experimental studies that have shown water/humidity to decrease the frictional strength of kaolinite gouges. Examples are Morrow et al (2000; https://doi.org/10.1029/1999GL008401) and Beynon and Faulkner (2020; https://doi.org/10.1016/j.jsg.2020.104094)
456 – strain partitioning instead of portioning
464 – interesting result
480 – This interpretation suggests that the cross-faults occurred along pre-existing weaknesses, whereas the pre-existing weakness proposed for the Ridgecrest region, the Independence Dike Swarm, occurs in the orthogonal, mainshock orientation. How do you reconcile this?
493 – ‘off-fault’ instead of ‘of-fault’
505 – increasingly unfavourable
Citation: https://doi.org/10.5194/egusphere-2026-1948-RC2
Data sets
Initiation of strike-slip faults in experiments with pre-existing weaknesses dataset Michele L. Cooke and Christ F. Ramos Sanchez https://doi.org/10.6084/m9.figshare.31961007
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 331 | 799 | 45 | 1,175 | 47 | 45 |
- HTML: 331
- PDF: 799
- XML: 45
- Total: 1,175
- BibTeX: 47
- EndNote: 45
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The authors present analogue “claybox” models of strike-slip fault evolution in the upper crust testing localized vs distributed basal shear, uncut vs precut weaknesses in the crust, and varying the angle and spacing of the pre-cut weaknesses relative to the primary active fault. Their results show that the angle of pre-cut weaknesses (a proxy for pre-existing faults) controls if slip occurs on those weaknesses in the early stages of fault evolution, and if the slip is sympathetic or antithetic to slip on the primary fault. The presence or absence of pre-existing weaknesses also affects the geometric complexity of the final throughgoing rupture. The authors link their results to observed deformation in the Mw 7.1 Ridgecrest earthquake, though these results could allow geologists to gain insights from surface rupture patterns on many continental strike-slip faults. The results could also impact fault displacement hazard analysis near immature strike-slip faults (recommending wider deformation zones), though this is not mentioned in the text.
In general, this study is easy to read and mostly logically organized. The figures are informative and illustrate the story well. Although there are some minor grammatical or typographic errors in the text (see minor comments), I recommend only minor edits prior to publication. Nice work!
General Comments:
Minor Comments:
Line 6 – Add Mw 7.1 to the description of the Ridgecrest mainshock
19 – rephrase “if they had” to “those with”
22 – Might consider rephrasing the final sentence of the abstract as the pre-existing structures may be unknown prior to rupture.
24-25 – Name the historical earthquakes with complex fault networks that are being referred to here so that the “these recent earthquakes” and “these ruptures” in lines 26 and 28 are referring back to actual earthquakes.
Figure 1 caption - States the maps are “taken” from Rosa et al., 2024, and Milliner et al., 2021. Are the data from those publications or the figures themselves? If the figures themselves, make sure proper permissions are acquired and state instead “figure from”. If just the data, then state “data from”.
62 – Add a comma between “clay, which”
86 – delete the “to” between “Fault systems evolve”
92 – Put the “in the crust” after “weaknesses” instead of after “bedding planes”
95 – Consider adding an introductory sentence to this paragraph something along the lines of “We test four cases of strike-slip fault evolution: with and without basal shear and with and without pre-existing weaknesses”. Then go ingo the details of the claybox model. Also, instead of “All strike-slip fault experiments…” consider “Our strike-slip” or “All our…”.
121-122 – The sentence “We expects…dextral slip (Figure 3)” is more of a hypothesis than methods. Consider rephrase.
130 – “several other” is vague. Are there more relevant details?
149 – Add a “the” in “impact the number of detected faults”
151 and 153 – “5x5 mm^2” and “(area <6 mm ^2)” both are areas already, and do not need the ^2.
176 – It seems a word is missing after “basal”
177 – It seems 1 or 2 is missing after “stage”
185 – Weaknesses (plural)
91 – in the title: “…with 2 cm weakness spacing”
202 – “experiment with 90 degree weaknesses”
Figure 4 – consider adding an arrow or few to point out the pre-cut weaknesses in one of the panels
Figure 5 caption – The x axis in the figure is “displacement” but the figure caption calls it “timing”. I understand how they are linked, but consider using “displacement” in the caption as the figure does not have a time component labelled.
238 – update to “This reorganization of the active FAULT system marks the START of stage…”
247 – add a reference to Figure 6 somewhere in the first sentence of this paragraph.
260 – Add a comma between “…faults, none…”
267 and 290 – sinistral and dextral are misspelled
274, 281, and throughout – use “less” instead of “lesser”
284 – what kind of fault irregularities - geometrical?
295-296 – update to: “…as the applied loading was accommodated as DISTRIBUTED deformation and kinematic efficient increased as the faults evolved and accommodated MORE LOCALIZED slip”
Figure 8 – what do the light tan and dark tan regions represent in panel A? How are the onset of each stage defined – is it manual my inspecting the deformation patterns, or is there a quantitative threshold?
315 – replace “that” with “than”
334 – Add a comma between “displacement, slip”
341 – Add a comma after “stage 2,”
353 – missing final ) after “stage 1”
378 – add a comma after “During stage 2,”
393 – “irregular” rather than “irregularly”
400 – add a comma after “pre-cut surfaces,”
426 – “rheometer” what a great word!
432/Figure 10 caption – Where do the data for part B, “real-world orientations” come from?
437-440 – Completely agree!
454 – “…the shear zone that resembles the pattern…”
464 – “contribute TO curved cross-faults”
474 – “production OF simultaneously active…”
484 – “this early fault irregularity IMPACTS”
495 – irregular rather than irregularly
500 – less rather than lesser
505 – update to “increasingly unfavorable”
507 – consider adding a sentence or half a sentence about how these results might also inform PFDHA around immature faults – e.g., the results could support recommending wider deformation zones around immature faults.