Unraveling carbonate fault dynamics, from friction to decarbonation, through the 1959 Mw 7.2 earthquake in Montana

Zamani, Nina; Murphy, Michael A.; Ferré, Eric C.; Barou, Fabrice

doi:10.1038/s41598-025-89071-4

Download PDF

Article
Open access
Published: 13 March 2025

Unraveling carbonate fault dynamics, from friction to decarbonation, through the 1959 Mw 7.2 earthquake in Montana

Nina Zamani¹,
Michael A. Murphy¹,
Eric C. Ferré² &
…
Fabrice Barou³

Scientific Reports volume 15, Article number: 8625 (2025) Cite this article

2227 Accesses
3 Citations
Metrics details

Subjects

Abstract

Seismic rupture in carbonate rocks influences fault friction behavior through thermal evolution and mineral reactions. Focusing on the 1959 Mw 7.2 Hebgen Lake event in western Yellowstone, Montana, the largest earthquake on a normal fault in the United States, we analyze fault rock microstructures and mineralogical changes to constrain frictional heating on the fault plane. We combine thermal maturity of organic matter, magnetic fabric, and thermomagnetic methods with scanning electron microscopy to unravel variations in peak frictional temperature along the fault slip surface. The mineral changes caused by coseismic heating (e.g., nanocalcite formation or goethite to hematite reaction) occur in patches along the fault mirror, hence reflecting considerable differences in frictional heat. While coseismic thermal heterogeneities have been reported in other rock types, this is the first time they are documented and quantified specifically in carbonates. Furthermore, these results provide new mineralogical criteria to quantify coseismic frictional heat in natural faults at temperatures lower than that of decarbonation and highlight the need to consider coseismic friction processes at a scale larger than most deformation experiments. For example, we document the critical role played by fault plane attitude (dip) at the scale of a few tens of centimeters in production of frictional heat. Our results emphasize that while coseismic decarbonation dynamically weakens carbonate-hosted faults, it may generally not occur along an entire fault plane.

Experimental production of K-rich metasomes through sediment recycling at the slab-mantle interface in the fore-arc

Article Open access 10 November 2023

The impact of melt versus mechanical wear on the formation of pseudotachylyte veins in accretionary complexes

Article Open access 27 January 2022

Mesoscopic and macroscopic investigation of a dolomitic marble subjected to thermal damage

Article Open access 12 September 2022

Introduction

Earthquake energy is dissipated mainly along the fault plane in the form of frictional heat and rock fracturing and, to a lesser extent, as radiated seismic waves^1,2 resulting in destructive effects on population centers³. Multiple meso- and microstructures indicate slip at seismic velocities⁴. Frictional heating along faults is documented through various indicators such as clay mineral reactions^5,6, fission tracks⁷, or mineral dehydration⁸. In some cases, coseismic frictional heat is so intense that it results in frictional melting, i.e., the formation of pseudotachylytes⁹. In carbonates, instead of melting, frictional heating leads to calcite (or dolomite) breakdown into lime (or lime and periclase) and carbon dioxide. Due to the transient nature of frictional heating and lack of simple methods to measure the heat budget of seismic deformation, the heat budget remains poorly understood¹⁰. To understand the seismic deformation in carbonate rocks, many deformation experiments have been performed in the laboratory on centimeter-size samples^{11,12,13,14,15,16}. These experiments clearly show that extreme comminution, combined with decarbonation, results in massive reduction of friction, further promoting seismic slip^11,12. Despite recent studies on natural carbonate-hosted faults^{17,18,19,20,21,22,23}, the scaling of deformation experiments to nature remains problematic because fault planes tend to be geometrically irregular and compositionally heterogeneous.

To advance our understanding of physico-mineralogical conditions preceding thermal decomposition and frictional behavior of seismogenic carbonate-hosted faults, we target the August 17, 1959, Mw7.2 Hebgen Lake Earthquake in the West Yellowstone area²⁴, the largest historic seismic event on a normal fault in the United States. During this double event, the Hebgen Lake Fault (HLF) and the Red Canyon Fault (Fig. 1A) produced a ~ 40 km long surface rupture, with maximum displacement up to 7.7 m, on a surface dipping 60 to 85°, resulting in 28 casualties^24,25,26,27. Overall, the surface rupture zone is covered by unconsolidated Holocene sediment, except at Section 31 Creek (S31C), where a 250 m-long bedrock fault scarp is visible within the Cambrian Meagher Limestone (Fig. 1B,C and Supplementary S1A)²⁸. Cosmogenic ³⁶Cl on samples from this fault exposure indicate at least 7 earthquake events during the past 24 ka²⁹. This remarkable exposure allows us to investigate the details of fault slip processes, such as frictional heating and mineral growth/reactions during seismic rupture along a major carbonate-hosted normal fault. Regional structures and previous seismological investigations show that the 1959 rupture initiated at 12 km-focal depth and propagated through carbonates, all the way to the S31C surface fault scarp²⁹.

We analyze microstructures and iron oxide/hydroxide assemblages to decipher the thermal fingerprints of seismic slip produced by the 1959 earthquake. We unravel the fault deformation and peak temperature by combining mineralogical and geochemical analyses, lattice-preferred orientation (LPO), magnetic fabric, and thermomagnetic analyses.

The wide variation of frictional heat, on the Hebgen Lake Fault at the scale of a few meters, suggests that further understanding the mechanical behavior of seismogenic carbonate faults will require field-based analysis, at a scale larger than that of conventional deformation experiments. The new mineral and magnetic criteria discussed in this study to evaluate variations in frictional heat could be applied to other carbonate-hosted seismogenic faults, e.g., in the Apennines of Italy³⁰, the Longmenshan Belt of China³¹, the Corinth Rift of Greece^32,33 or the Zagros Belt of Iran^34,35.

Results

The HLF footwall at the S31C locality presents four adjacent zones (Fig. 1B,C): (A) the mirror zone, a ~ 30 mm-thick surface, with glossy patches of either calcite-dolomite (CaCO₃, CaMg(CO₃)₂), goethite (αFeO(OH)) or hematite (αFe₂O₃) slickenlines, that cut abruptly through all previous structures; (B) a ~ 50 mm-thick zone with extensive layered-botryoidal goethite; (C) a ~ 1.0–1.2 m-thick yellowish damage zone, with a network of sub-parallel fractures that decrease in density away from the mirror, and (D) the least deformed greyish limestone.

Macroscopically, zone A shows three types of fault mirrors with distinct slickenlines: A1 mirrors with carbonate slickenlines (Fig. 2A); A2 mirrors with localized goethite slickenlines on top of layers of botryoidal goethite (Fig. 2B); and A3 mirrors with pervasive hematite slickenlines (Fig. 2C). Hematite exclusively occurs along the HLF 1959 slip surface as a < ~ 1 mm-thick red-maroonish slickenline layer on non-striated goethite deposits (Figs. 1B, 2B,C). All hematite patches observed on the HLF are characterized by slickenside fabric and approximately 0.3–0.6 m tall by 1–2 m wide. The hematite fault mirrors occur preferentially on topographic ridges (humps) of the S31C fault scarp (Fig. 2C). All three fault mirrors locally grade into each other with similar slickenline trends and normal sense of shear.

Microstructural details of the A1–A3 fault mirrors are described below (Figs. 1C, 2D–I, Supplementary S2 and Supplementary S3).

Carbonate fault mirrors (A1) are composed of carbonate grains, < 10 μm in diameter (Fig. 2D,G). These grains exhibit equigranular microstructures with high-angle (~ 120°) grain boundaries, consistent with thermal equilibration (Fig. 2G and inset). Euhedral pyrite cubes, ranging in size from ~ 1 to 50 μm, are found locally in aggregates of tens to hundreds of grains, regardless of their distance from the mirror (Figure Supplementary S2). Some diagenetic pyrite grains close to the carbonate fault mirror (A1) exhibit compositional and microstructural heterogeneity, reflected in backscatter electron images. The reflective part of the mirror is comprised of a discontinuous zone approximately 200 μm-thick, composed of fine, equant, calcite grains without preferred crystallographic orientation, referred to as nanocalcite (with a grain size smaller than 1 μm). The systematic presence of topographic depressions (pit) around the pyrite grains indicates that sulfides reacted with oxygen and water, leading to the localized dissolution of carbonate (refer to reactions in Figure Supplementary S2). Similar reactions also resulted in morphological pits on these A1 surfaces. A1 mirrors also exhibit prominent slickenlines, sub-parallel to each other over several cm², and a network of open shear fractures, less than 5 mm wide, sub-parallel to the fault mirror. The density of fractures notably increases towards the fault core. The nanocalcite layer that forms the fault mirror displays equant nanosize, densely packed, equigranular calcite grains (~ 1 μm).

Goethite fault mirrors (A2) are characterized by ochre-colored, macroscopically isotropic, equigranular goethite, calcite, dolomite, and a few subrounded clasts of carbonates, quartz, and rare feldspar, some showing undulose extinction (see Fig. 2E). These clasts are typically < 100 μm in size and slightly elongated parallel to the fault mirror. Goethite masses form botryoidal concretions with asymmetric compositional banding, broadly parallel to the fault mirror, composed of alternating bands of goethite-rich and goethite-poor material (Figure Supplementary S1B). These goethite masses are microcrystalline and individual grains cannot be discerned macroscopically. Goethite fault mirrors (A2) transition into hematite fault mirrors (A3) across a sheared contact zone. Moreover, a network of hematite-filled tension fractures, up to 10 mm wide, cuts through A2 mirrors at high angles to the mirrors. In some areas, the goethite-rich layer displays slickenlines and elongated grains of goethite and calcite, indicating a shape-preferred orientation parallel to the fault mirror (see Fig. 2H).

Hematite fault mirrors (A3) consist of three superimposed layers at increasing distances from the slip surface (see Fig. 2F): layer Hem—a hematite-rich domain with a few quartz clasts and no carbonate clasts; in these layer, hematite is microcrystalline and individual grains cannot be distinguished macroscopically, the hematite slickensides appear platy and parallel to the mirror; layer Gth—a ~ 3 mm-thick goethite-rich, comminuted domain bearing fewer and smaller subrounded clasts of quartz and carbonates, along with asymmetric lens-shaped domains indicative of shearing; layer Cal—a breccia containing abundant, rounded to angular clasts of quartz, feldspar, and recrystallized polycrystalline quartz clasts. Multiple quartz clasts in layer Cal present a slight elongation in the same direction, forming a subtle shape-preferred orientation parallel to the fault mirror.

The nanometer grain size of goethite and hematite in the A3 hematite fault mirror is indicated by their superparamagnetic (SP) properties (Fig. 3A–C). This fault mirror hosts diamagnetic (calcite, dolomite, quartz, kaolinite) and antiferromagnetic (goethite, hematite) minerals that have various contributions to magnetic susceptibility (Supplementary—Table S1). To increase the resolution of magnetic analyses, the samples were sliced into three adjacent 3.5 mm-thick slabs parallel to the fault mirror (Supplementary S4). Overall the two antiferromagnetic phases, goethite and hematite, dominate the magnetic susceptibility (ranging 65–100% and 0–40%, respectively). The quantitative contribution of these phases to magnetization is determined through Isothermal Remanent Magnetism (IRM) experiments (Supplementary S5) showing that, in the hematite fault mirror, goethite (mean coercivity, B_h ~ 2.94 T) and hematite (B_h ~ 2.02 T) account respectively for ~ 91.2% and ~ 8.8% of remanent magnetization.

Grain size distribution of magnetic minerals is determined through First Order Reversal Curve (FORC) analysis, a powerful method to resolve magnetic granulometry³⁵. Figure 3A shows a prominent SP fringe along the vertical axis, indicating the presence of SP hematite in slab 1 (~ layer Hem) of hematite fault mirror A3. Conversely, Fig. 3D,E show the presence of a mixture of SP and minor single-domain (SD) goethite assemblage in slabs 2 and 3.

The Anisotropy of Magnetic Susceptibility (AMS) provides information about the magnetocrystalline orientation of goethite and hematite nanograins. The three slabs 1 to 3 show distinct AMS fabric as a function of distance to the fault mirror (Fig. 3D–F and Supplementary S6). Samples from the hematite fault mirror slab 1 (mirror) display a prominent magnetic fabric that broadly coincides with the fault mirror orientation and the slickenlines direction. The degree of anisotropy Pʹ ranges from 1.106 to 1.491, which indicates a ferromagnetic s.l. origin and oblate magnetic fabrics (T ~ 0.128 to 0.236) are most likely due to the deformation process and the planar magnetocrystalline anisotropy of goethite and hematite. The electron backscatter diffraction (EBSD) data shows that the AMS-carrying phases goethite and hematite have weak lattice-preferred orientations, ~ 2.5 multiples of uniform distribution (Supplementary S7). Goethite displays maximum point distribution of its (001) and (100) crystallographic axes whereas hematite crystallographic axes distribution (0001) and (11$\overline{2}$0) form partial girdles indicative of partial activation of multiple slip systems.

Two types of thermomagnetic experiments shed light on mineral reactions of the HLF rocks during frictional heating (25–700 °C). In the Curie temperature experiment (Fig. 3G), an HLF goethite sample is continuously heated under argon flux to reduce oxidation. From 25 to 400 °C, the magnetic susceptibility (κ) decreases by ~ 30% due to goethite dehydration, while mass decreases by < 12%. From 400 to 570 °C, we observe minor κ fluctuations attributed to the breakdown of iron-bearing minerals. At ~ 570 °C, κ sharply drops, corresponding to the Néel temperature (T_N) of a low-Ti hematite.

The experiment becomes irreversible at 570 °C (Figure Supplementary S8), indicating lab-induced magnetite formation. This experiment also confirms that during slip in the natural conditions, in the hematite fault mirror (A3), peak temperature never exceeded 570 °C; otherwise, magnetite would have formed. The stepwise heating experiment (Fig. 3H) involves incremental and cyclic heating, in steps of 25 °C for 2 min, and cooling to room temperature. This 2 min duration of heating is intended to mimic the short-lived²⁵ seismic slip conditions at shallow depth. Up to 460 °C, mass-normalized magnetic susceptibility (χ) remains stable. Above 460 °C, calcium carbonate decarbonation and goethite dehydration occur concurrently until ~ T_N, causing a similar decrease in χ as in the Curie temperature experiment.

To further constrain co-seismic peak temperature along the HLF, we also applied Rock–Eval, a well-established hydrocarbon pyrolysis method used to determine the burial temperature of sediments^37,38. Based on 2 sets (5 samples each), separated by 0.1 m from each other, the results show a substantial decrease in hydrocarbon contents (S₁–S₂–S₃) by 55 to 30% between 1.5 and 7.5 mm away from the fault mirror (Supplementary S9 and Supplemental Table S2).

Discussion

Origin and significance of nanoscale iron hydroxides, oxides and carbonates

The most likely Fe source for the iron oxides and hydroxides along the HLF lies in the diagenetic sulfides of the Meagher Limestone, as shown by the carbonate dissolution pits surrounding sulfide grains (Supplementary S2). Also, close to the fault mirror, pyrite shows a heterogeneous internal structure (Supplementary S2C) indicative of incipient thermal decomposition that typically occurs around ~ 253 °C³⁹.

Goethite occurs along the HLF as botryoidal concretions cut by the fault plane (Fig. 2E); these deposits predate the 1959 earthquake. Also, the goethite botryoidal morphology with its typical concentric layers suggests formation in a stagnant aqueous fluid-rich setting⁴⁰, i.e., in between major seismic events and absence of deformation. These very fine grain size deposits (~ 25 nm) are documented primarily by their superparamagnetic behavior shown in the FORC analysis (Fig. 3B,C). In general, goethite forms through chemical weathering of iron oxides and sulfides or precipitation from iron-rich aqueous solutions⁴¹. In the case of the HLF (Fig. 2B), weathering can be ruled out because these deposits occur only along the fault plane and not elsewhere in the bedrock. Therefore, goethite must have formed through percolation and precipitation of supergene iron-rich solutions related to pyrite alteration (see Sect. 2.3). The goethite magnetic fabric most likely resulted from unidirectional fluid flow deposition, as shown elsewhere by deposits in an industrial pipe filled with precipitated calcite and goethite⁴².

Hematite in the Meagher Limestone has been observed only at the S31C outcrop, along the HLF 1959 fault mirror (A3), as slickenlines forming a < ~ 3 mm-thick layer, always associated with and directly overlying the goethite-rich layer (Fig. 2B,C). The position of this polished hematite layer atop the fault structure (Fig. 1C), its excellent preservation, and identical slickenside orientation with other slickensides indicate that the hematite slickenlines developed synkinematically during the most recent slip event of the HLF in 1959. ³⁶Cl isotopic data²⁹ shows that the samples used in our study originate from the 1959 slip surface. Therefore, the striated hematite on this slip surface must have formed through the goethite to hematite dehydration reaction during the 1959 slip event. Also, no hematite was found on the older parts of the HLF produced by previous seismic events, probably because hematite readily hydrates to goethite in the presence of fluids.

At the microscopic scale, the coseismic dehydration of goethite into hematite is documented by the hematite overgrowths surrounding goethite grains (see enlarged inset of Fig. 2I), further attesting to coseismic frictional heating along the HLF slip surface. Several studies on reddish, hematite slickenlines on carbonate fault mirrors elsewhere concur that they formed under ~ 300 °C and at seismic slip velocities^43,44,45,46. Further, in heating experiments without deformation, goethite reacts to hematite between 200 and 300 °C^47,48,49, which supports our interpretation that hematite formed through coseismic frictional heating because localized, elevated temperatures are unique to seismic slip. The difference between our results on carbonate rocks and experiments on pure natural goethite lies in the delay in oxidation of goethite into hematite caused by CO₂ release during decarbonation. We acknowledge, also, that this reaction is grain-size dependent⁵⁰, and to test the thermal stability of goethite upon heating, we conducted thermomagnetic experiments to simulate coseismic frictional flash heating on the HLF slip surface. In our experiments, the dehydration reaction began at ~ 100 °C and was completed at ~ 460 °C, while hematite continued to form up to 620 °C (Fig. 2A,B). Combined with X-ray diffraction analyses (Supplementary Table S1 and Figure S10), these experiments support goethite dehydration to hematite upon frictional heating (Supplementary Eq. 4). Furthermore, the quasi-parallel direction of the AMS magnetic lineation within the hematite fault mirror and the hematite slickenlines (Fig. 3D) support their shared coseismic origin.

In previous carbonate deformation experiments, nanoscale calcite typically forms glossy surfaces similar to natural fault mirrors. These surfaces may also bear nanocalcite that forms through decarbonation and subsequent recrystallization of portlandite^{11,12,13,15,46} or, more seldomly, through extreme comminution and sintering⁴⁴. On the HLF slip surface, carbonate slickensides occur on a few discontinuous and irregular patches, a few tens of centimeters across (Fig. 2A). The most glossy of these patches preserve a thin nano-calcite layer, hence forming a fault mirror (Fig. 2D,G). We interpret this nanocalcite layer as resulting from thermal decomposition of carbonates based on the comparison between the HLF A1 mirror microstructures in this layer with microstructures proven elsewhere to result from either decarbonation or from simple grain size reduction: (1) cataclastic comminution follows an asymmetric power-law crystal size distribution (CSD), which is clearly not the case at the HLF (Figure S3A–C); (2) grain size reduction through brittle fracturing and mechanical grinding produces rounded grains⁴⁴^{Figure 2E,F}, which is not the case at HLF (Figure S3); (3) as decarbonation takes place at temperatures ~ 400–600 °C, the recrystallized grains tend to develop equant, sintered microstructures¹⁵^{Figure 5D,}¹⁶^{Figure 13B} which is precisely what we observe on the A1 mirror at HLF (Figure S3A).

The XRD data (Supplemental Table S1 and Figure S10) shows a drastic increase in hematite concentration in the A3 fault mirror. Overall, three different fault mirrors (carbonate, goethite, and hematite) show that post-seismic mineral assemblage on the slip surface is controlled by the heterogeneous distribution of mineral precipitates along the fault domain. Also, the lack of plastic or cataclastic strain in the goethite botryoidal concretions shows that they formed during interseismic periods of the HLF history when the fault experienced limited to no creep between seismic events. Further, our observations show that previously precipitated goethite concretions were coseismically smeared into hematite slickenlines (Fig. 2B,C) due to dynamic frictional heating. Due to morphological undulation along the fault plane, goethite was unevenly sheared into slickenlines in some places but not sufficiently heated to form hematite (Fig. 2B). The patchy distribution of hematite slickenlines also suggests that frictional heat was not evenly distributed along the slip surface.

Contribution of AMS to fabric development and deformation history

We calculate the diamagnetic (calcite, quartz, dolomite, kaolinite) and antiferromagnetic (goethite, hematite) contributions to magnetic susceptibility using published intrinsic susceptibilities and XRD modal percentages measured on the hematite fault mirror (Table S1; Figure S5). These calculations highlight goethite’s dominant contribution (65–100%) to magnetic susceptibility across all slabs and hematite’s major role (~ 40%) in the fault mirror slab.

The magnetic grain size of goethite and hematite, evaluated using magnetic hysteresis and FORC data, corresponds to superparamagnetic (SP) domain state (Fig. 3C–E), i.e., nanoscale sizes of ~ 30 nm for hematite³⁶ and ~ 25 nm for goethite^51,52. Based on IRM acquisition experiments, goethite and hematite, respectively, account for ~ 91% and ~ 9% of magnetic remanence (Supplementary S5).

In general, the AMS of deformed rocks arises from the combined magnetostatic, magnetocrystalline, and interaction anisotropies of minerals^53,54. In the HLF rocks, the AMS arises mainly from the strong magnetocrystalline anisotropy of goethite and hematite (Supplementary Table S1), as shown elsewhere^53,54,55. For SD and SP goethite, the AMS tensor mimics the grain SPO, unlike in the case of MD grains that have an inverse fabric^39,56. Similarly, SP hematite shows a normal AMS fabric^41,55. Therefore, the AMS of goethite-rich and hematite-rich HLF rocks reflects the mineral SPO (Figs. 2I, 3D–F). As goethite is a hydrous mineral deposited by fluids, its fabric must have been produced through fluid precipitation-percolation, primarily without much plastic deformation, i.e., in interseismic periods (Fig. 4A,B). Moreover, the AMS long axis reflects the unidirectional fluid flow direction, as shown elsewhere by⁴² in hydrothermal deposits. The source of the HLF ferruginous fluids will be discussed in a subsequent study. The presence of slickenlines on goethite fault mirrors (A2) that seamlessly transition into hematite fault mirrors (A3) indicates that localized shear deformation of goethite, due to frictional sliding below 300 °C, likely took place in slab 2.

In slab 3, the AMS K₁ direction (magnetic lineation) represents quasi-exclusively the direction of fluid flow along the dip direction of the HLF plane (Fig. 3D–F and Supplementary S4) as shown elsewhere by⁴². In contrast, in slab 2, the pre-existing AMS is modified by shear deformation, resulting in a shallower plunging K₁ trend and dipping magnetic foliation (plane normal to K₃ axis). In slab 1, fault mirror, the AMS partially arises from the weak SPO of hematite and goethite clusters (Fig. 2I). The distinctively lower degree of anisotropy in slab 1 compared to slabs 2 and 3 is explained by the lower intrinsic anisotropy of hematite compared to goethite^41,57. We interpret the hematite microfabric in the fault mirror to result from synkinematic comminution, under moderate temperature (150-350 °C), and shear deformation of the underlying layer of precipitated SP goethite. The AMS K₁ direction matches the slip direction along the hematite fault mirror.

Implications of fault mirror development for the Hebgen Lake Fault

The Plio-Pleistocene slip history of the Hebgen Lake Fault, previously established by²⁹ using chlorine isotopes on the S31C scarp, comprised up to six events at ~ 23.8, 20.3, 7.0, 2.6, 1.7, and 0.4 ka ago, prior to the 1959 Mw7.2 event⁵⁸ confirmed these results using Lidar and Schmidt hammer methods, although conventional paleoseismic trenches documented only three main events²⁶. Finally, surveys performed immediately after the Mw 7.2 earthquake⁵⁹ showed that the S31C fault scarp resulted from the 1959 slip event. This young age also partially accounts for the sporadic preservation of a nanocalcite thin layer on the A1 carbonate fault mirror that we report here for the first time (Fig. 2D,G).

In the following, we first briefly review prevailing conditions for fault mirror development and then, in this context, discuss the temperatures inferred from the carbonate-, goethite- and hematite-mirrors specifically for the HLF.

Carbonate fault mirrors are naturally glossy and smooth surfaces formed during slip at subseismic to seismic velocities^43,60. These thin zones consist of layers of nanoscale grains^43,44,60,61. Furthermore, deformation experiments show that fault mirrors result from microstructural (e.g., comminution) and chemical processes (e.g., decarbonation) resulting in dynamic weakening^44,45,60,62, and therefore, this is a critical process to understand earthquake mechanics. We however acknowledge that decarbonation does not systematically occur along carbonate seismogenic faults although this process is well documented elsewhere^18,19,20. The continuous formation and destruction of fault mirrors throughout seismic slip generally leads to their patchy occurrence on a fault plane⁴⁴. Well-preserved glossy mirrors in carbonate faults have been attributed to the sintering of nanocalcite grains^50,63,64.

The HLF carbonate fault mirror (A1, Fig. 2A) exhibits evidence of frictional heating, which we estimate through Rock–Eval experiments by observing a significant reduction in hydrocarbon contents (S₁, S₂, S₃) ranging from 55 to 30% over a distance of 1.5–7.5 mm from the fault mirror (Supplementary Figure S9B). This decrease in hydrocarbon content is most evident in the S₁ parameter, which primarily reflects the release of methane (CH₄) at a minimum temperature range of 300–350 °C. Our analysis detected this release at a distance of 6.5 mm from the fault mirror. In contrast, under natural conditions in carbonates with the presence of O₂, the peak of katagenesis typically occurs around 120 °C⁶⁵ (Supplementary Figure S9A). However, in our pyrolysis experiments, conducted in a helium atmosphere to avoid oxidation, the peak katagenetic temperature occurred between 345 and 420 °C. Therefore, it is likely that under the shallow, oxidizing conditions of the Hebgen Lake Fault, closer to the surface, the natural peak temperature for frictional heating pyrolysis would have been around 120 °C. Using this temperature, we determine the bulk katagenesis temperature as ~ 120 °C at 4.5 mm from the slip surface, a value in agreement with our numerical modeling results (Supplementary Figure S9B and Fig. 4C). The stability of kaolinite in sample HL011S1C56 (Table S1) also places an upper limit on the maximum temperature reached during slip to ~ 400 °C⁶⁶.

Further, the uniform grain size and equant shape of nanocalcite grains (~ 1 μm) along the carbonate fault mirror (Fig. 2G inset and Supplementary S3A) most likely did not form through comminution because this process typically forms broad crystal size distributions (CSD)⁶⁷. Instead, the nanocalcite microstructure resulted from either from superplastic flow⁴⁴ or from decarbonation, a thermally and chemical-diffusion driven process⁶⁸. Also, the well-defined Ca Ka peak on EDS spectra of nanocalcite indicates high degree of crystallinity (Supplemental Figure S3A inset). We argue that the equant shape of the nanocalcite supports post-kinematic, static recrystallization of lime or portlandite whereas synkinematic superplastic flow and comminution would have resulted in slightly elongated grains. We attribute the weak lattice-preferred orientation of nanocalcite (MUD ~ 2; Supplementary S7) to static recrystallization in a thermal gradient.

In addition, the numerical modelling of frictional heating (Table 1), Rock–Eval results (Supplementary S9), and microstructures indicative of pyrite thermal decomposition (Supplementary S2C), show that coseismic frictional heating took place and most likely affected a ~ 60 mm-thick zone along the fault mirror. Also, the formation of lime and portlandite along slip surfaces is a short-lived process and these decarbonation-related phases tend to be preserved only for a short duration⁶⁹. As the earthquake took place in 1959, these minerals disappeared rapidly through weathering and fault scarp degradation.

Table 1 Calculation of frictional heat along the Hebgen Lake Fault plane and plot showing the variation of frictional heat (Q_f ) with dip angle.

Full size table

The A2 HLF goethite mirror (Fig. 2B) consists of a thin layer of slickenside grains (< 50 μm) structurally on top of a layer of botryoidal goethite (< 50 mm) in which the concentric growth layers are visible. The slickenside layer exhibits two types of microstructures similar to those described elsewhere on other carbonate fault mirrors⁷⁰: (1) the sigmoidal microstructures (Fig. 2E) that indicate high normal stress and moderate velocity (~ 100 MPa; ~ 10^–2 m/s), (2) the mirror-parallel fractures (Fig. 2H) that indicate lower normal stress and higher slip velocities (~ 10 MPa; ~ 0.3 m/s). The range of physical conditions under which these microstructures developed suggests that this mirror recorded seismic slip from the early acceleration phase through the rise of frictional heat. The microstructures above occur in domains parallel to the mirror (Fig. 2E).

The A3 HLF hematite fault mirror (Fig. 2C), based on FORC analysis (Fig. 3A), consists of nanograins of hematite, a mineral with a normal AMS. The hematite mirror bears a magnetic foliation (plane normal to K₃ axis) close to the fault slip surface and a magnetic lineation (K₁ axis) quasi identical to the hematite slickenlines (Fig. 3D). Since the thermomagnetic experiments indicate that hematite forms above 460 °C (Fig. 3H), we conclude that the hematite magnetocrystalline anisotropy (Fig. 2D) was acquired as a result of hematite shear deformation under moderate temperature, as shown elsewhere by⁷¹. This interpretation is consistent with currently available deformation mechanism maps for hematite⁷² and with the partial girdle of hematite crystallographic axes (Supplemental Material—Figure S70).

Numerical quantification of HLF coseismic frictional heat during the 1959 slip event

We calculate the thickness of host-rock that underwent decarbonation, using fault displacement, dip angle, depth, thermal diffusivity, specific heat, specific mass of limestone, enthalpy and temperature of decarbonation, while considering adiabatic heat transfer after 10 s of slip (Table 1).

The 1959 earthquake produced displacements greater than 7 m in multiple sections of the Red Canyon Fault²⁷ and the Hebgen Lake Fault, including at the S31C site (²⁹ and this study). We use this approximate 7 m displacement and the estimated depth of the outcrop (~ 7 m) to calculate several faulting parameters for the S31C locality, where the HLF is best exposed.

The energy released during faulting (W_f) is mostly dissipated as frictional heat (Q_f) and, to a lesser extent, as radiated seismic energy (E) (2).

$${\text{W}}_{{\text{f}}} = {\text{Q}}_{{\text{f}}} + {\text{E}}$$

(1)

Indeed, experiments and seismological studies (2) show that radiated seismic energy (E) is small (< 5%) compared with the total earthquake energy (W), therefore;

$${\text{W}}_{{\text{f}}} \approx {\text{Q}}_{{\text{f}}}$$

(2)

It follows that, per unit area,

$${\text{W}}_{{\text{f}}} \approx {\text{d}} \times \sigma_{{\text{s}}}$$

(3)

where d is the true displacement, and σ_s is the shear stress at the deformation time.

Using the fault plane dip angle θ = 66°, the fault parameters of Table 1, and Eq. (3), we calculate shear stress acting upon the HLF plane at 7 m depth, as σ_n = 63.8 × 10³ Pa, and frictional heat

$${\text{Q}}_{{\text{f}}} \approx {446}.{6} \times {1}0^{{3}} \;{\text{N/m}} = {446}.{6} \times {1}0^{{3}} \;{\text{J/m}}^{{2}}$$

(4)

Table 1 also shows the prominent role played by fault plane dip angle (θ) in frictional heat production across a 20° range of dip angles (from 56 to 76°). Equation 4 (Table 1) further highlights the dependency of frictional heat (Q_f) to shear stress (σ_s) and, as σ_s varies with dip angle [σ_s = (σ sin2θ)/2], it follows that Q_f varies with dip angle (θ) (see graph in Table 1). In the following, we use the 66° angle as the median value at the S31C locality. This thermal energy produced by friction during a single seismic slip (Q_f) event is mostly adiabatic because the heat-transfer-distance (D) into the host-rock is minimal during a short-lived (~ 10 s) seismic slip event:

$${\text{D}} = (kt)^{{0.{5}}}$$

(5)

With thermal diffusivity for limestone κ = 1.14 mm² s⁻¹ ⁷³ and t, time in seconds, we calculate D = (1.14 mm² s⁻¹ × 10 s)^0.5 ~ 3.38 mm after 10 s of seismic slip. This low D value agrees with the ~ 3 mm-thickness of the goethite to hematite reaction zone that we interpret to result from coseismic frictional heat.

Then, we calculate how coseismic frictional heat was used to heat and decompose the host carbonate on both sides of the slip plane (hanging wall and footwall). The thermal energy Q_d needed to both heat and decompose a dolomitic carbonate rock of one mass unit is

$${\text{Q}}_{{\text{d}}} = {\text{C}}_{{\text{p}}} \Delta {\text{T}} + \Delta {\text{H}}$$

(6)

With Cp: specific heat, ΔT: the temperature difference between ambient and decomposition temperature, and ΔH is the enthalpy of the decarbonation reaction. We use the enthalpy ΔH of the endothermic decarbonation reaction from⁷⁴:

$$\begin{aligned} & {\text{Ca}},{\text{Mg}}\left( {{\text{CO}}_{{3}} } \right)_{{2}} \leftrightarrow {\text{CaCO}}_{{3}} + {\text{MgO}} + {\text{CO}}_{{2}} \quad \Delta {\text{H}} = {187} \times {1}0^{{3}} \;{\text{J/mol}} \\ & {184}.{\text{4 g}} \leftrightarrow {1}00.{\text{1 g}} + {4}0.{\text{3 g}} + {44}.0{\text{ g}} \\ \end{aligned}$$

(7)

Decarbonation of dolomite occurs at temperatures as low as ~ 550 °C⁷⁵ and the host rock was at the time of seismic slip at a temperature of ~ 25 °C, therefore ΔT = 550 − 25 = 525 °C. For dolomite Cp = 0.854 × 10³ J/(kg ℃)⁷⁶.

The heat needed to increase the temperature of 1 kg of dolomite from 25 to 550 °C is

$${\text{Cp}}\Delta {\text{T}} = 0.{854} \times {1}0^{{3}} \times {525} \approx {447} \times {1}0^{{3}} \;{\text{J}}.$$

In addition, to decarbonate 1 kg of dolomite, the heat needed is

$$\Delta {\text{H}} \times {\text{m}} = \left( {{1}000/{184}.{4}} \right) \times {187} \times {1}0^{{3}} \;{\text{J}} = {1}0{14} \times {1}0^{{3}} \;{\text{J}}$$

where m = moles of dolomite.

Overall, the heat needed to heat and decarbonate 1 kg of dolomite from 25 to 550 °C is

$${\text{Q}}_{{\text{d}}} = \left( {{447} + {1}0{14}} \right) \times {1}0^{{3}} = {1461} \times {1}0^{{3}} \;{\text{J}}$$

(8)

The total frictional heat produced by seismic slip over 1 m² fault surface (446.6 × 10³ J) must be distributed evenly between the hanging wall and the footwall, hence, the total mass of dolomite that can be decarbonated is m = 446.6/1462/2 kg = 0.153 kg. With a 1 m² surface and a dolomite specific mass of 2500 kg/m³, the volume of decarbonated dolomite is v = 0.153/2500 ≈ 61.2 × 10⁻⁶ m³, which corresponds to a maximum decarbonated thickness t of ~ 61 μm.

This maximum calculated thickness of Meagher Limestone that underwent decarbonation on one side of the fault is in broad agreement with our microstructural observations of a thin layer of nanocalcite (~ 30 μm). This calculated thickness is most likely overestimated because as previously shown elsewhere^11,42, dynamic weakening is likely to occur due to grain size reduction, thermal pressurization and coseismic decarbonation. Similarly, we consider the initial coefficient of friction (μ) at the onset of slip (Table 1) and we acknowledge that μ will decrease during slip towards dynamic steady state value of ~ 0.2¹¹. When μ falls to its lower value of 0.2, the predicted thickness of decarbonated material falls to ~ 5 μm.

In addition, the uneven attitude and heterogeneity of the fault plane (Figs. 1 and 2) most likely resulted in uneven frictional heat distribution. The calculations above indicate that frictional heat at a distance greater than ~ 1 mm from the slip plane should not affect microstructures such as calcite twins. Finally, the Rock–Eval results constrain the width of the thermal aureole where hydrocarbons were heated by friction (~ 120 °C) along the HLF to 4.5 mm from the slip surface (Figure S7).

Heterogeneity of coseismic frictional heat and implications for fault behavior

In summary, despite differences in protolith composition, the slickenlines maintain consistent orientation across mirror types (carbonate, goethite, and hematite), which shows that they likely formed under similar dynamic stress states and slip rates, although presumably under different temperatures. The carbonate mirror consists of a thin layer of nanocalcite (< 1 mm-thick) parallel to the fault mirror that is responsible for the mirror’s glossy appearance. Decarbonation above 400 °C documented in the Curie experiment (Fig. 3G), in addition to the uniform nano-calcite grain size (Supplementary Figure S7E), supports a thermally driven reaction for this layer instead of a mechanical wear origin⁷⁷. Our temperature estimate is in excellent agreement with previous experiments¹⁶, showing that decarbonation in dolomite begins at temperatures lower than previously thought.

Further, previous experiments on natural goethite showed that it dehydrates upon heating from 25 to 300 °C^50,78. The step-wise heating experiment on the HLF rocks attests that goethite dehydration occurs between ~ 250 and 460 °C (Fig. 3H). Hence the goethite fault mirror most likely records a broad range of frictional temperatures from 25 to 460 °C. In previous heating experiments, the observed temperature of goethite to hematite reaction ranges between 200 and 300 °C^47,48,49. In our stepwise heating experiments on the HLF rocks, conducted statically without deformation, we observe that hematite forms above 460 °C (Fig. 3H), and therefore, we interpret the frictional heat associated with the 1959 slip event and the formation of the hematite fault mirror to be also above 460 °C. However, we acknowledge that this reaction is grain-size dependent⁵⁰.

The fault core before the 1959 seismic slip (~ 7 m length) comprised various minerals, including dolomite and goethite, leading to likely variations in core strength along the HLF slip plane due to distinct frictional properties. While these differences may account for variations in frictional heat, the variations in dip angle would majorly impact frictional heat, as highlighted in Table 1, underscoring its importance over other factors such as for example slip zone thickness.

Recent studies have shown the effects of fault geometry/long-wavelength fault surface topography on slip behavior^79,80,81. These geometric variations of fault attitude would also determine heterogeneous frictional heat generated along a fault⁸².

In the case of the Hebgen Lake Fault, our assessment of frictional heating variations is primarily based on presence or absence of hematite formed on top of a precursor layer of goethite. The reasons why parts of the HLF do not bear goethite is beyond the scope of this study. The goethite to hematite reaction, documented in the HLF through macroscopic field observations (Fig. 2B), microstructures (Fig. 2F), and thermomagnetic experiments (Fig. 3A,B), constrains frictional heat along the hematite fault mirror (A3). The fact that in the Meagher Limestone, hematite occurs only in patches of the S31C fault scarp (Figs. 1B, 2C,F,I), forming a < 1 mm-thick slickenside layer, adjacent to goethite slickenside patches, shows that frictional heat was heterogeneously distributed. Moreover, we interpret the three types of mirrors at the HLF to have formed during the 1959 seismic slip event. In addition, the AMS fabric acquired by the hematite fault mirror coincides with slickensides and is further testimony of its seismic origin at temperatures > 460 °C. Ultimately, the structural macroscopic record of previous seismic slip events²⁹ has most likely been overprinted or degraded to various degrees⁵⁸.

Our results introduce a significant advance in thermal constraints on carbonate fault frictional heat: the goethite to hematite reaction at around ~ 460 °C as a new indicator of seismic activity along carbonate fault planes. Unlike previous markers, such as nanocalcite and portlandite^45,46, which are transient, this new reaction offers lasting mineralogical evidence. Understanding these processes is crucial as they profoundly reduce friction along carbonate faults, further promoting seismic slip through lubrication by both comminution and decarbonation^11,12,14,15, and pore fluid pressure increase through CO₂ production¹⁶, further promoting failure.

Our understanding of seismic deformation in carbonates has greatly benefitted from friction experiments usually performed at the centimeter scale. Nonetheless, field observation-based investigations remain crucial in developing novel concept-based models. Moreover, the variations in fault dip angle play a critical role in frictional heating and decarbonation along natural seismogenic faults. Investigating and detecting the changes in friction, temperature, and mineral assemblage thus becomes critical instead of assuming a uniform maximum temperature from constant friction. Here, we show that iron hydroxide deposits, formed prior to seismic rupture during interseismic periods in carbonates, constitute valuable frictional heat markers. Our investigations of the Hebgen Lake Fault S31C scarp show that, before the 1959 earthquake, aqueous fluid flowed through fractured rock, chemically altered diagenetic iron sulfides, liberated iron, and resulted in substantial precipitation of botryoidal goethite along the fault plane. The origin of these fluids, meteoric or hydrothermal, is still under investigation. This goethite is then locally heated and reacts to hematite under well-documented thermal conditions. Also, the lowest coseismic frictional marker determined along the HLF consists in the Rock–Eval 120 °C temperature measured at ~ 4.5 mm from the slip plane on the carbonate A1 fault mirror. The second lowest record of frictional heat is indicated by the microstructural heterogeneity of diagenetic pyrite grains close to the slip plane at temperatures ca. 250 °C. Further, the presence of a thin nano-calcite layer at the surface of the carbonate fault mirror and the paucity of calcite clasts in the hematite fault mirror attest to carbonate thermal decomposition at temperatures > 400 °C. Finally, frictional heating > 460 °C is recorded along the fault plane through the decomposition of goethite to hematite, which constitutes a powerful new indicator of coseismic frictional heat. Our outcrop-based study of the Hebgen Lake Fault reveals that frictional heating, followed by decarbonation, took place at least locally during the 1959 Mw7.2 earthquake and highlights the need to advance our understanding of rupture processes in carbonate from hypocenter depth to the surface. This study highlights magnetic minerals that had previously been overlooked, which serve as important indicators of fault peak temperature in carbonate rocks.

Data availability

All original magnetic, XRD, SEM, EBSD and microstructural data are available at the Zenodo data repository (https://doi.org/https://doi.org/10.5281/zenodo.14541699).

References

Kanamori, H. Energy budget of earthquakes and seismic efficiency. in Earthquake Thermodynamics and Phase Transformations in the Earth’s Interior, 293–305 (Academic Press, 2001).
Scholz, C.H. in The Mechanics of Earthquakes and Faulting, Edn. 3. (Cambridge University, 2019).
Kanamori, H. Real-time seismology and earthquake damage mitigation. Annu. Rev. Earth Planet. Sci. 33, 195–214 (2005).
Article ADS CAS MATH Google Scholar
Curzi, M., Aldega, L., Billi, A., Boschi, C., Carminati, E., Vignaroli, G., Viola, G. and Bernasconi, S.M. (2024). Fossilchemical-physical (dis) equilibria between paleofluids and host rocks and their relationship to the seismic cycle and earthquakes. Earth-Sci. Rev. 104801.
Kitajima, H., Chester, J. S., Chester, F. M. & Shimamoto, T. High-speed friction of disaggregated ultracataclasite in rotary shear: Characterization of frictional heating, mechanical behavior, and microstructure evolution. J. Geophys. Res. Solid Earth 115, B08408 (2010).
Article ADS MATH Google Scholar
Ujiie, K., Kameyama, M. & Yamaguchi, A. Geological record of thermal pressurization and earthquake instability of subduction thrusts. Tectonophysics 485, 260–268 (2010).
Article ADS MATH Google Scholar
Tagami, T. Zircon fission-track thermochronology and applications to fault studies. Rev. Mineral. Geochem. 58, 95–122 (2005).
Article CAS MATH Google Scholar
Woo, S., Han, R. & Oohashi, K. Principal slip zone in kaolinite gouge: A seismic slip indicator?. J. Struct. Geol. 175, 104940 (2023).
Article MATH Google Scholar
Sibson, R. H. Generation of pseudotachylyte by ancient seismic faulting. Geophys. J. Int. 43, 775–794 (1975).
Article ADS MATH Google Scholar
d’Alessio, M. A., Blythe, A. E. & Burgmann, R. No frictional heat along the San Gabriel fault, California: Evidence from fission-track thermochronology. Geology 31, 541–544 (2003).
Article ADS Google Scholar
Han, R., Shimamoto, T., Hirose, T., Ree, J. H. & Ando, J. I. Ultralow friction of carbonate faults caused by thermal decomposition. Science 316, 878–881 (2007).
Article ADS CAS PubMed Google Scholar
Han, R., Shimamoto, T., Ando, J. I. & Ree, J. H. Seismic slip record in carbonate-bearing fault zones: An insight from high-velocity friction experiments on siderite gouge. Geology 35, 1131–1134 (2007).
Article ADS Google Scholar
Billi, A., Di Toro, G. & Landowe, S.J. Fault-related carbonate rocks and earthquake indicators: Recent advances and future trends. in Structural Geology: New Research, 63–86 . (Nova Science Publishing, 2008).
Boneh, Y., Sagy, A. & Reches, Z. Frictional strength and wear-rate of carbonate faults during high-velocity, steady-state sliding. Earth Planet. Sci. Lett. 381, 127–137 (2013).
Article ADS CAS Google Scholar
Violay, M., Nielsen, S., Spagnuolo, E., Cinti, D. & Di Toro, G. Pore fluid in experimental calcite-bearing faults: Abrupt weakening and geochemical signature of co-seismic processes. Earth Planet. Sci. Lett. 361, 74–84 (2013).
Article ADS CAS Google Scholar
Kim, H. N., So, B. D., Kim, M. S., Han, K. S. & Oh, S. B. Seismic fault weakening via CO₂ pressurization enhanced by mechanical deformation of dolomite fault gouges. Geology 49, 1245–1249 (2021).
Article ADS CAS MATH Google Scholar
Goren, L., Aharonov, E. & Anders, M. H. The long runout of the Heart Mountain landslide: Heating, pressurization, and carbonate decomposition. J. Geophys. Res. Solid Earth. https://doi.org/10.1029/2009JB007113 (2010).
Article MATH Google Scholar
Rowe, C. D., Fagereng, Å., Miller, J. A. & Mapani, B. Signature of coseismic decarbonation in dolomitic fault rocks of the Naukluft Thrust, Namibia. Earth Planet. Sci. Lett. 333, 200–210 (2012).
Article ADS Google Scholar
Collettini, C., Viti, C., Tesei, T. & Mollo, S. Thermal decomposition along natural carbonate faults during earthquakes. Geology 41, 927–930 (2013).
Article ADS CAS Google Scholar
Collettini, C. et al. Fault structure and slip localization in carbonate-bearing normal faults: An example from the Northern Apennines of Italy. J. Struct. Geol. 67, 154–166 (2014).
Article ADS Google Scholar
Delle Piane, C., Clennell, M. B., Keller, J. V., Giwelli, A. & Luzin, V. Carbonate-hosted fault rocks: A review of structural and microstructural characteristics with implications for seismicity in the upper crust. J. Struct. Geol. 103, 17–36 (2017).
Article ADS Google Scholar
Zamani, N., Heij, G. W., Ferré, E. C., Murphy, M. A. & Bagley, B. High-velocity slip and thermal decomposition of carbonates: Example from the Heart Mountain Slide ultracataclasites, Wyoming. J. Geophys. Res. Solid Earth 128, e2022JB026185 (2023).
Article ADS CAS Google Scholar
Witkind, I. J., Myers, W. B., Hadley, J. B., Hamilton, W. & Fraser, G. D. Geologic features of the earthquake at Hebgen Lake, Montana, August 17, 1959. Bull. Seismol. Soc. Am. 52, 163–180 (1962).
Article Google Scholar
Doser, D. I. Source parameters and faulting processes of the 1959 Hebgen Lake, Montana, earthquake sequence. J. Geophys. Res. Solid Earth 90, 4537–4555 (1985).
Article MATH Google Scholar
Johnson, K. L., Nissen, E. & Lajoie, E. L. Surface rupture morphology and vertical slip distribution of the 1959 Mw 7.2 Hebgen Lake (Montana) earthquake from airborne lidar topography. J. Geophys. Res. Solid Earth 123, 8229–8248 (2018).
Article ADS MATH Google Scholar
Andreuttiova, L., Hollingsworth, J., Vermeesch, P., Mitchell, T. M. & Bergman, E. Revisiting the 1959 Hebgen Lake earthquake using optical image correlation: New constraints on near-field 3D ground displacement. Geophys. Res. Lett. 49, 66 (2022).
Article Google Scholar
Wallace, R. E. Degradation of the Hebgen Lake fault scarps of 1959. Geology 8, 225–229 (1980).
Article ADS MATH Google Scholar
Zreda, M. & Noller, J. S. Ages of prehistoric earthquakes revealed by cosmogenic chlorine-36 in a bedrock fault scarp at Hebgen Lake. Science 282, 1097–1099 (1998).
Article ADS CAS PubMed Google Scholar
Smeraglia, L., Aldega, L., Billi, A., Carminati, E. & Doglioni, C. Phyllosilicate injection along extensional carbonate-hosted faults and implications for co-seismic slip propagation: Case studies from the central Apennines, Italy. J. Struct. Geol. 93, 29–50 (2016).
Article ADS Google Scholar
Wang, H. et al. Paleoseismic slip records and uplift of the Longmen Shan, eastern Tibetan Plateau. Tectonics 38, 354–373 (2019).
Article ADS MATH Google Scholar
Jackson, J. A. et al. Seismicity, normal faulting, and the geomorphological development of the Gulf of Corinth (Greece): The Corinth earthquakes of February and March 1981. Earth Planet. Sci. Lett. 57, 377–397 (1982).
Article ADS Google Scholar
Taylor, B. et al. The structures, stratigraphy, and evolution of the Gulf of Corinth rift, Greece. Geophys. J. Int. 185, 1189–1219 (2011).
Article ADS Google Scholar
Authemayou, C. et al. Quaternary slip rates of the Kazerun and the Main Recent Faults: Active strike-slip partitioning in the Zagros fold-and-thrust belt. Geophys. J. Int. 178, 524–540 (2009).
Article ADS Google Scholar
Nissen, E., Tatar, M., Jackson, J. A. & Allen, M. B. New views on earthquake faulting in the Zagros fold-and-thrust belt of Iran. Geophys. J. Int. 186, 928–944 (2011).
Article ADS Google Scholar
Roberts, A. P. et al. Characterization of hematite (α-Fe₂O₃), goethite (α-FeOOH), greigite (Fe₃S₄), and pyrrhotite (Fe₇S₈) using first-order reversal curve diagrams. J. Geophys. Res. Solid Earth. https://doi.org/10.1029/2006JB004715 (2006).
Article Google Scholar
Lafargue, E., Marquis, F. & Pillot, D. Rock-Eval 6 applications in hydrocarbon exploration, production, and soil contamination studies. Rev. Inst. Fr. Pet. 53, 421–437 (1998).
Article CAS MATH Google Scholar
Behar, F., Beaumont, V. D. E. B. & Penteado, H. D. B. Rock-Eval 6 technology: Performances and developments. Oil Gas Sci. Technol. 56, 111–134 (2001).
Article CAS MATH Google Scholar
Hu, G., Dam-Johansen, K., Wedel, S. & Hansen, J. P. Decomposition and oxidation of pyrite. Prog. Energy Combust. Sci. 32, 295–314 (2006).
Article CAS MATH Google Scholar
Parry, W. T. Composition, nucleation, and growth of iron oxide concretions. Sedim. Geol. 233, 53–68 (2011).
Article ADS CAS MATH Google Scholar
Dekkers, M.J. Some rock magnetic parameters for natural goethite, pyrrhotite, and fine-grained hematite. Ph.D. dissertation, Utrecht University, 249 p. (1988).
Sizaret, S. et al. Crystallization in flow–I. Palaeocirculation tracked by texture analysis and magnetic fabrics. Geophys. J. Int. 167, 605–612 (2006).
Article ADS CAS Google Scholar
Fondriest, M. et al. Mirror-like faults and power dissipation during earthquakes. Geology 41, 1175–1178 (2013).
Article ADS Google Scholar
Siman-Tov, S., Aharonov, E., Boneh, Y. & Reches, Z. E. Fault mirrors along carbonate faults: Formation and destruction during shear experiments. Earth Planet. Sci. Lett. 430, 367–376 (2015).
Article ADS CAS Google Scholar
Siman-Tov, S., Aharonov, E., Sagy, A. & Emmanuel, S. Nanograins form carbonate fault mirrors. Geology 41, 703–706 (2013).
Article ADS Google Scholar
Ohl, M. et al. Mechanisms of fault mirror formation and fault healing in carbonate rocks. Earth Planet. Sci. Lett. 530, 115886 (2020).
Article CAS Google Scholar
Goss, C. J. The kinetics and reaction mechanism of the goethite to hematite transformation. Mineral. Mag. 51, 437–451 (1987).
Article CAS MATH Google Scholar
Gialanella, S. et al. On the goethite to hematite phase transformation. J. Therm. Anal. Calorim. 102, 867–873 (2010).
Article CAS Google Scholar
Ponomar, V. P. Thermomagnetic properties of the goethite transformation during high-temperature treatment. Miner. Eng. 127, 143–152 (2018).
Article CAS MATH Google Scholar
Walter, D., Buxbaum, G. & Laqua, W. The mechanism of the thermal transformation from goethite to hematite. J. Therm. Anal. Calorim. 63, 733–748 (2001).
Article CAS Google Scholar
Guyodo, Y., Mostrom, A., Penn, R. L. & Banerjee, S. K. From nanodots to nanorods: Oriented aggregation and magnetic evolution of nanocrystalline goethite. Geophys. Res. Lett. https://doi.org/10.1029/2003GL017021 (2003).
Article Google Scholar
Till, J. L., Guyodo, Y., Lagroix, F., Morin, G. & Ona-Nguema, G. Goethite as a potential source of magnetic nanoparticles in sediments. Geology 43, 75–78 (2015).
Article ADS Google Scholar
Borradaile, G. J. & Jackson, M. Anisotropy of magnetic susceptibility (AMS): Magnetic petrofabrics of deformed rocks. Geol. Soc. Spec. Publ. 238, 299–360 (2004).
Article ADS MATH Google Scholar
Ferré, E. C., Gébelin, A., Till, J. L., Sassier, C. & Burmeister, K. C. Deformation and magnetic fabrics in ductile shear zones: A review. Tectonophysics 629, 179–188 (2014).
Article ADS Google Scholar
Martin-Hernandez, F. & Guerrero-Suarez, S. Magnetic anisotropy of hematite natural crystals: High field experiments. Int. J. Earth Sci. 101, 637–647 (2012).
Article CAS MATH Google Scholar
Dekkers, M. J. & Rochette, P. Magnetic properties of chemical remanent magnetization in synthetic and natural goethite: Prospects for a natural remanent magnetization/thermoremanent magnetization ratio paleomagnetic stability test. J. Geophys. Res. Solid Earth 97, 17291–17307 (1992).
Article Google Scholar
Martin-Hernandez, F. & García-Hernández, M. M. Magnetic properties and anisotropy constant of goethite single crystals at saturating high fields. Geophys. J. Int. 181, 756–761 (2010).
ADS CAS MATH Google Scholar
Tye, A. & Stahl, T. Field estimate of paleoseismic slip on a normal fault using the Schmidt hammer and terrestrial LiDAR: methods and application to the Hebgen fault (Montana, USA). Earth Surf. Process. Landf. 43, 2397–2408 (2018).
Article ADS MATH Google Scholar
Witkind, I. J. The Hebgen Lake, Montana, earthquake of August 17, 1959. Geol. Surv. Prof. Pap. 435, 37–51 (1959).
Google Scholar
Pozzi, G., De Paola, N., Nielsen, S. B., Holdsworth, R. E. & Bowen, L. A new interpretation for the nature and significance of mirror-like surfaces in experimental carbonate-hosted seismic faults. Geology 46, 583–586 (2018).
Article ADS CAS MATH Google Scholar
Pluymakers, A. & Røyne, A. Nanograin formation and reaction-induced fracturing due to decarbonation: implications for the microstructures of fault mirrors. Earth Planet. Sci. Lett. 476, 59–68 (2017).
Article ADS CAS Google Scholar
Chen, X., Madden, A. S., Bickmore, B. R. & Reches, Z. E. Dynamic weakening by nanoscale smoothing during high-velocity fault slip. Geology 41, 739–742 (2013).
Article ADS Google Scholar
Goldberg, R., Siman-Tov, S. & Emmanuel, S. Weathering resistance of carbonate fault mirrors promotes rupture localization. Geophys. Res. Lett. 43, 3105–3111 (2016).
Article ADS CAS MATH Google Scholar
Park, Y., Hirose, T. & Ree, J. H. Carbonate fault mirrors with extremely low frictional healing rates: a possible source of aseismic creep. Geophys. Res. Lett. 48, e2021GL093749 (2021).
Article ADS MATH Google Scholar
Kissin, Y. V. Catagenesis and composition of petroleum: Origin of n-alkanes and isoalkanes in petroleum crudes. Geochim. Cosmochim. Acta 51, 2445–2457 (1987).
Article ADS CAS MATH Google Scholar
Frost, R. L. & Vassallo, A. M. The dehydroxylation of the kaolinite clay minerals using infrared emission spectroscopy. Clays Clay Miner. 44, 635–651 (1996).
Article ADS CAS MATH Google Scholar
Fondriest, M., Smith, S. A., Di Toro, G., Zampieri, D. & Mittempergher, S. Fault zone structure and seismic slip localization in dolostones: An example from the Southern Alps, Italy. J. Struct. Geol. 45, 52–67 (2012).
Article ADS Google Scholar
Chen, J., Niemeijer, A. R. & Spiers, C. J. Seismic fault slip behavior predicted from internal microphysical processes. J. Geophys. Res. Solid Earth 127, e2022JB024530 (2022).
Article ADS Google Scholar
Hewitt, K. Catastrophic landslide deposits in the Karakoram Himalaya. Science 242, 64–67 (1988).
Article ADS CAS PubMed MATH Google Scholar
Mercuri, M., Scuderi, M. M., Tesei, T., Carminati, E. & Collettini, C. Strength evolution of simulated carbonate-bearing faults: The role of normal stress and slip velocity. J. Struct. Geol. 109, 1–9 (2018).
Article ADS Google Scholar
Mendes, M. & Lagoeiro, L. Microstructures, crystallographic fabric development, and deformation mechanisms in natural hematite aggregates deformed under varied metamorphic conditions. J. Struct. Geol. 40, 29–43 (2012).
Article ADS Google Scholar
Siemes, H. et al. Texture, microstructure, and strength of hematite ores experimentally deformed in the temperature range 600–1100 °C and at strain rates between 10⁻⁴ and 10⁻⁶ s⁻¹. J. Struct. Geol. 25, 1371–1391 (2003).
Article ADS Google Scholar
Hanley, E. J., Dewitt, D. P. & Roy, R. F. The thermal diffusivity of eight well-characterized rocks for the temperature range 300–1000 K. Eng. Geol. 12, 31–47 (1978).
Article CAS MATH Google Scholar
Criado, J., González, M., Málek, J. & Ortega, A. The effect of the CO₂ pressure on the thermal decomposition kinetics of calcium carbonate. Thermochim. Acta 254, 121–127 (1995).
Article CAS Google Scholar
Humphries, T. D. et al. Dolomite: A low-cost thermochemical energy storage material. J. Chem. 7, 1206–1215 (2019).
CAS Google Scholar
Stout, J. W. & Robie, R. A. Heat capacity from 11 to 300 K, entropy, and heat of formation of dolomite. J. Phys. Chem. 67, 2248–2252 (1963).
Article CAS Google Scholar
Shimamoto, T. & Nagahama, H. An argument against the crush origin of pseudotachylytes based on the analysis of clast-size distribution. J. Struct. Geol. 14, 999–1006 (1992).
Article ADS MATH Google Scholar
Prasad, P. S. R. et al. In situ FTIR study on the dehydration of natural goethite. J. Asian Earth Sci. 27, 503–511 (2006).
Article ADS MATH Google Scholar
Eijsink, A. M., Kirkpatrick, J. D., Renard, F. & Ikari, M. J. Fault surface morphology as an indicator for earthquake nucleation potential. Geology 50, 1356–1360 (2022).
Article ADS CAS Google Scholar
Rodriguez Padilla, A. M. et al. The influence of fault geometrical complexity on surface rupture length. Geophys. Res. Lett. 51, e2024GL109957 (2024).
Article Google Scholar
Sagy, A., Morad, D. & Lyakhovsky, V. Can geometrical barrier explain the Mw 7.8 earthquake in southern Türkiye in February 2023?. Seismol. Res. Lett. 95, 643–650 (2024).
Article Google Scholar
Kirkpatrick, J. D. & Shipton, Z. K. Geologic evidence for multiple slip weakening mechanisms during seismic slip in crystalline rock. J. Geophys. Res. Solid Earth 114, B12401 (2009).
Article ADS MATH Google Scholar
Harrison, R. J. & Feinberg, J. M. FORCinel: An improved algorithm for calculating first-order reversal curve distributions using locally weighted regression smoothing. Geochem. Geophys. https://doi.org/10.1029/2008GC001987 (2008).
Article MATH Google Scholar

Download references

Acknowledgements

We kindly acknowledge the insights provided by Dr. Siman-Tov (Geological Survey of Israel) through conversation while writing this manuscript. Dr. Yingfeng Xu (University of Louisiana at Lafayette) is kindly acknowledged for performing the Rock-Eval pyrolysis and X-ray diffraction measurements at the University of Louisiana at Lafayette. Dr. Gabe Casale (Appalachian State University) is gratefully acknowledged for his initial efforts to collect EBSD data. The authors are very grateful to Dr. Manuel Curzi, Dr. Luca Aldega, and a third reviewer (anonymous) for their insightful comments that improved the quality of the manuscript. We also kindly acknowledge the valuable input provided in the review process by the Editor, Dr. Andrea Billi.

Funding

This study was supported by Geological Society of America Graduate Research Grant to NZ and National Science Foundation grants EAR-0521558 and EAR-2041040 to EF.

Author information

Authors and Affiliations

Department of Earth & Atmospheric Sciences, University of Houston, Houston, TX, 77204, USA
Nina Zamani & Michael A. Murphy
Department of Geological Sciences, New Mexico State University, Las Cruces, NM, 88003, USA
Eric C. Ferré
Géosciences Montpellier UMR CNRS 5243, Université de Montpellier, 34095, Montpelier, France
Fabrice Barou

Authors

Nina Zamani
View author publications
Search author on:PubMed Google Scholar
Michael A. Murphy
View author publications
Search author on:PubMed Google Scholar
Eric C. Ferré
View author publications
Search author on:PubMed Google Scholar
Fabrice Barou
View author publications
Search author on:PubMed Google Scholar

Contributions

NZ collected field data, prepared the samples, performed all measurements and numerical modelling, processed data and interpreted all results, and drafted the manuscript. MM supervised and edited the successive versions of the manuscript. EF oversaw all magnetic measurements and SEM analyses, supervised the numerical modelling and assisted in drafting the manuscript. FB acquired and interpreted the EBSD data. All authors discussed the results and contributed to the final manuscript.

Corresponding author

Correspondence to Nina Zamani.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary Information 1.

Supplementary Information 2.

Supplementary Information 3.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Zamani, N., Murphy, M.A., Ferré, E.C. et al. Unraveling carbonate fault dynamics, from friction to decarbonation, through the 1959 Mw 7.2 earthquake in Montana. Sci Rep 15, 8625 (2025). https://doi.org/10.1038/s41598-025-89071-4

Download citation

Received: 10 September 2024
Accepted: 03 February 2025
Published: 13 March 2025
DOI: https://doi.org/10.1038/s41598-025-89071-4