Depth-Segmented Rupture of a Back-Thrust Fault During the 2022 Hormozgan (Iran) Earthquake Sequence

Highlights

What are the main findings?

• Multi-track Sentinel-1 InSAR observations and geodetic inversions reveal that the 2022 Hormozgan earthquake sequence was driven by a single, previously unmapped NE-striking, SW-dipping blind back-thrust fault.
• The multi-month sequence exhibited a distinct, partially overlapping depth-segmented rupture pattern, where the July events ruptured the deeper segment (5.5–8.0 km) and the November event ruptured the shallower segment (1.5–6.0 km) on the same fault plane.

What are the implications of the main findings?

• This depth-segmented behavior demonstrates that secondary conjugate back-thrust structures in the eastern Zagros can accommodate crustal shortening through multi-stage sequential ruptures, effectively releasing shallow slip deficits.
• Static Coulomb stress modeling implies that the 2022 sequence has significantly increased the seismic failure risk on the adjacent, unruptured eastern segments of the major Mountain Frontal Fault (MFF) and Zagros Frontal Fault (ZFF).

Abstract

Between 1 July and 30 November 2022, four spatially adjacent shallow M_W ≥ 5.7 earthquakes successively struck the Hormozgan province in southern Iran. This earthquake sequence offers a vital opportunity to clarify the subsurface seismogenic structure and rupture evolution in the eastern segment of the Zagros Fold-and-Thrust Belt (ZFTB). In the paper, we apply multi-temporal archived SAR images from the Sentinel-1 satellite to extract the high-precision coseismic surface deformation covering the July and November earthquake events, respectively, and further investigate the related seismogenic fault structure and slip distribution. Geodetic inversion results reveal that the cumulative coseismic slip of the three M_W ≥ 5.7 earthquakes in July is distributed at a downdip depth of 5.5 to 8 km on a SW-dipping thrust seismogenic fault plane, while the coseismic slip of the November M_W 5.7 earthquake is concentrated in the shallow downdip range of 1.5 to 6 km on the same fault, finely characterizing a partially overlapping depth-segmented rupture. According to a joint analysis of the regional topography and geomorphology, active fault distribution, and coseismic inversions, we conclude that this earthquake sequence nucleated on a secondary blind back-thrust fault of the Zagros Frontal Fault (ZFF). Coseismic Coulomb stress changes reveal that the July earthquake sequence triggered the occurrence of the November earthquake and that the shallow eastern segment of the Mountain Frontal Fault (MFF) and the eastern segment of the ZFF exhibit significant stress loading, indicating a high risk of future rupture.

1. Introduction

The Zagros Fold-and-Thrust Belt (ZFTB), an ~1800 km long and 200–300 km wide orogenic belt in southern Iran resulting from the Arabian–Eurasian collision, represents one of the most active continental collision zones on Earth [1,2,3] (Figure 1a,b). This tectonic belt accommodates ~10 mm/yr of crustal shortening, accounting for roughly half of the total plate convergence rate (~22 mm/yr) [4,5]. In the southeastern Zagros Mountains, tectonic deformation is primarily controlled by major W-NW-trending and N-NE dipping structures, notably the MFF and the ZFF [6,7] (Figure 1c). The MFF is a listric thrust fault structure extending over 1350 km; its dip angle is ~40–50° at shallow depths and flattens to 20–35° at greater depths, controlling the frontal topography of the Zagros Mountains [7,8]. Similarly, the ZFF is a listric thrust fault, dipping northward at a moderate angle of 30–50° in the shallow crust and gradually decreasing in dip as it extends downward to the sedimentary cover–basement interface; it exhibits high tectonic activity and substantial seismogenic potential [9,10,11]. As a typical recent moderate-to-strong earthquake sequence in the northern vicinity of the ZFF, the July–November 2022 earthquake sequence provides an excellent opportunity to resolve the regional subsurface seismogenic structure and rupture mechanisms using modern satellite geodetic data, thereby advancing our understanding of regional geodynamics and seismic hazard assessment.

On 1 July 2022 (21:32 UTC), a M_W 6.0 earthquake struck Hormozgan Province in southern Iran at a depth of 16 km according to the United States Geological Survey (USGS; https://www.usgs.gov/). This was followed approximately two hours later by a M_W 5.7 aftershock (23:24 UTC, depth 7 km) and a second M_W 6.0 mainshock just one minute later (23:25 UTC, depth 9 km). These three earthquakes devastated at least 55 villages, resulting in 5 fatalities and 84 injuries, and the most severe damage occurred in Sayeh Khosh village near the epicenter, involving total housing collapses and extensive infrastructure destruction, including power outages and traffic disruptions caused by landslides (https://reliefweb.int/). In the five months following this July sequence, the International Institute of Earthquake Engineering and Seismology (IIEES; https://epp.iiees.ac.ir/) recorded 122 aftershocks with magnitudes ranging from 3.0 to 5.0 (Figure 1c). Although somewhat scattered, these aftershocks were generally clustered around the ZFF (Figure 1c). Notably, on 30 November 2022 (15:17 UTC), another M_W 5.7 earthquake occurred at a depth of 6 km, located within 10 km of the July earthquake sequence epicenter; however, this moderate event caused no casualties. Tectonically, all four moderate-to-strong earthquakes between July and November 2022 occurred in the hanging wall region of the ZFF at the eastern termination of the ZFTB (Figure 1c). Focal mechanism solutions from the USGS, GEOFON Earthquake Information Service (GEOFON; http://geofon.gfz-potsdam.de), and Global Centroid Moment Tensor (GCMT; https://www.globalcmt.org/) suggest that the M_W > 5.7 earthquake sequence could have ruptured on four spatially proximate thrust faults, either steeply SW-dipping or shallow NE-dipping (Tables S1 and S2). Early geodetic studies [9,13] posited that the July 2022 sequence nucleated on a SW-dipping, high-angle thrust fault, but these studies did not deeply explore the NE-dipping seismogenic fault model. By focusing only on the coseismic deformation and source parameters of the July sequence while neglecting the November M_W 5.7 event, previous studies failed to fully characterize the rupture evolution of the 2022 sequence. In contrast, Du et al. [14] investigated the seismogenic fault structure of the November M_W 5.7 earthquake using InSAR, but their fitting residuals for ascending and descending track coseismic deformation were large, requiring further optimization of fault geometric parameters. While previous studies have analyzed the geodetic characteristics of the July and November events independently, the primary novelty of this study lies in performing a cumulative July–November deformation inversion to resolve the structural linkage between the events and identifying a partially overlapping depth-segmented rupture pattern along a single, previously unmapped back-thrust fault. Furthermore, we leverage this unified fault model to provide a comprehensive Coulomb stress assessment of the pre-existing MFF and ZFF structures, which was absent in earlier analyses.

In this study, we first process ascending and descending Sentinel-1A SAR images using D-InSAR technology to generate multiple coseismic interferograms with varying temporal and perpendicular baselines. By comprehensively evaluating phase coherence and atmospheric noise, we derive high-precision coseismic surface deformation fields for the July and November 2022 Hormozgan earthquake sequences. Subsequently, we invert the coseismic rupture geometry and slip distribution of the sequence based on the elastic half-space rectangular dislocation theory and determine the optimal seismogenic fault plane by synthesizing geodetic inversion results, regional fault distribution, and kinematic characteristics. We then conduct an in-depth analysis of the rupture mode and subsurface seismogenic structure of the sequence. Finally, we calculate regional static Coulomb stress changes to explore the triggering relationships between the events and assess the risk of potential future ruptures in the region.

2. InSAR Observations

We utilize ascending track T130A and descending track T166D SAR images from the Sentinel-1A satellite (Table S3) to derive high-precision coseismic deformation fields for the 1 July and 30 November 2022 Hormozgan earthquake sequence. All interferograms are processed using the classic two-step SAR interferometry workflow [15] within the SARscape^® 6.1 module of the ENVI 5.6 software package. The 30 m resolution ALOS World 3D Digital Elevation Model (DEM) [16] is used to mitigate topographic effects, while European Space Agency (ESA) precise orbit data are introduced to correct orbital errors and geometric distortions. During processing, multi-looking is applied with a range-to-azimuth ratio of 8:2, followed by Goldstein filtering [17] to suppress speckle noise and enhance the signal-to-noise ratio of the interferograms. Finally, phase unwrapping is performed using the Delaunay Minimum Cost Flow (MCF) algorithm [18]. To obtain high-precision coseismic deformation fields, we compare and analyze multiple interferometric pairs with different temporal baseline combinations [19,20]. For the July 2022 events, the ascending interferometric pair spanning the time from 22 June 2022 to 4 July 2022 and descending interferometric pair spanning the time from 25 June 2022 to 7 July 2022 show the best coherence (Figure S1a–f) and are thus selected for subsequent fault geometry and slip distribution inversion. Conversely, for the November 2022 event, the 12-day baseline ascending interferometric pair is severely affected by atmospheric turbulence, presenting strong atmospheric noise and widespread decorrelation (Figure S1g–t). To effectively suppress observational noise and ensure deformation data quality, we implement an adaptive temporal baseline selection strategy. Rather than relying on standard sequential pairs, we evaluate multiple pairs and ultimately select an optimal ascending interferogram with an extended 24-day temporal baseline (25 November to 19 December), which significantly minimizes atmospheric noise compared to the standard 12-day pairs. Meanwhile, the concurrent 12-day baseline descending interferogram (28 November to 10 December) is less affected by atmospheric noise and maintains good coherence, thus being selected for the descending track (Figure S1g–t). Furthermore, optimizing the coherence threshold for phase unwrapping based on interferogram quality [21] requires balancing noise suppression with the preservation of valid high-gradient near-field deformation signals, as a higher threshold mitigates unwrapping errors in noisy regions but risks masking true coseismic signals near the fault. For the July event, which shows high overall coherence, a relatively loose threshold of 0.3 is applied to maximize the spatial coverage of the deformation field and to preserve the critical high-gradient deformation gradients near the epicenter (Figure S1a,d). In contrast, for the November event, where relatively lower coherence is observed in the interferograms (probably associated with seasonal surface changes), a stricter coherence threshold of 0.5 is adopted; although this inevitably sacrifices some far-field spatial coverage, it is essential for eliminating low-quality noisy pixels, reducing unwrapping errors, and improving the reliability of the deformation results (Figure S1g,p). Finally, Generic Atmospheric Correction Online Service (GACOS) products [22,23] are used to mitigate tropospheric delay noise, and ground control points (GCPs) located far from the main deformation zone are introduced to remove residual orbital errors and topography-correlated atmospheric delays [24,25,26]. Specifically, approximately 5 to 10 highly coherent points (coherence > 0.7) distributed uniformly in the non-deforming far-field regions are carefully selected as GCPs [24,25,26]. A quadratic polynomial fitting model is then applied to these points to estimate and remove the long-wavelength phase ramps from the interferograms [24,25,26]. To quantitatively evaluate the reliability and precision of the extracted InSAR data, an uncertainty assessment is conducted by calculating the standard deviation of the residual LOS displacements within non-deforming far-field patches. The residual standard deviations for the July ascending and descending tracks are ~1.1 mm and ~1.0 mm, respectively, while those for the November tracks are ~3.6 mm and ~1.0 mm. These small far-field errors confirm that orbital ramps and topography-correlated tropospheric delays are thoroughly mitigated.

The results show that both ascending and descending coseismic interferograms for the July 2022 earthquake sequence exhibit a prominent lobate fringe pattern (Figure 2), demonstrating comparable deformation magnitudes and similar deformation patterns, both dominated by motion toward the satellite, which reveals significant uplift during this earthquake. The maximum line-of-sight (LOS) displacements for the ascending and descending tracks reach ~24.1 cm and ~27.5 cm, respectively. The November 30 event display highly similar spatial distribution characteristics to the 1 July event (Figure 2), with maximum ascending and descending LOS displacements reaching ~5.8 cm and ~5.0 cm, respectively. The pre-existing ZFF traverses the coseismic interference fringes of both tracks from west to east, indicating a potential structural association with this fault. The continuous and smooth deformation fringes in ascending and descending tracks imply that the coseismic slip did not (or only slightly) propagate to the surface.

3. Geodetic Modeling

To unravel the seismogenic structure and rupture characteristics of the July–November 2022 Hormozgan earthquake sequence, we first down-sample the coseismic LOS deformation of both the July and November events using a resolution-based quadtree method [27]. Based on the uniform elastic half-space rectangular dislocation theory [28], we invert for the fault geometry and slip spatial distribution of both events [29,30,31,32].

3.1. The July 1 Hormozgan Earthquake Sequence

Given the absence of obvious surface ruptures and the discrepancy between the nodal planes provided by agencies like the USGS and GCMT (Table S1), identifying the true seismogenic fault plane directly is challenging. Therefore, we construct both SW-dipping and NE-dipping fault models and utilize the InSAR data to invert the fault geometry and slip distribution. Because the three M_W ≥ 5.7 earthquakes on July 1 occurred within a highly condensed timeframe (~2 h), far shorter than the 12-day repeat cycle of Sentinel-1, the resulting InSAR interferograms inherently capture the cumulative coseismic deformation of these multiple events [33]. Consequently, the derived slip distribution represents the combined rupture area of the July sequence rather than a single, isolated sub-event patch. In the initial inversion phase, fault parameters are set as free variables assuming uniform slip. The unconstrained inversions for both models yield an anomalously narrow fault width (~0.1 km) with unrealistically high slip (~10.0 m), mimicking a line source, which is physically implausible. To obtain a stable and physically meaningful result, we fix the fault width and conduct a grid search at 0.5 km intervals [34,35]. The results indicate that for the SW-dipping model (strike search range: 91–179°), the root mean square (RMS) misfit ceases to improve significantly once the slip exceeds ~2.7 m (corresponding to width < 2.5 km). For the NE-dipping model (strike range: 271–359°), the RMS improvement becomes negligible when slip exceeds ~1.8 m (corresponding to width < 4.0 km). Consequently, we fix the widths of the SW-dipping and NE-dipping models at 2.5 km and 4.0 km, respectively, and re-conduct the inversion (Figure 3a and Figure S2a). The nonlinear inversion of the SW-dipping model (Table S1) yields a length of ~15.8 km, a strike of ~98.9°, a dip of ~69.5°, a rake of ~93.5°, and a slip of ~2.7 m. We then fix the fault location, strike, and dip from the nonlinear inversion, extend the length and downdip width to 25 km and 15 km, respectively, and discretize the fault plane into 1 km × 1 km sub-patches for linear inversion. Since the dip angle derived from uniform slip models is often suboptimal for fine slip distribution inversions [36], we further optimize the dip via grid search (Figure 3b). The linear inversion results (Figure 3c–i) show the earthquake ruptured on a moderately dipping thrust fault, with coseismic slip distributed at depths of 5.5–8.0 km and a peak slip of ~2.2 m, and the derived slip model reproduces the observed ascending and descending coseismic deformation with an RMS misfit of 1.7 cm. Conversely, the nonlinear inversion for the NE-dipping model (Table S1) yields a strike of ~283.3°, a dip of ~16.3°, and a rake of ~107.0°, showing predominant thrusting with a minor right-lateral strike–slip component. Fixing the fault location, strike, and dip, we expand the plane to 30 km × 25 km, discretize it into 750 sub-patches, and optimize the dip via grid search (Figure S2b). The linear inversion results (Figure S2c–i) indicate nucleation on a low-angle (9°) thrust fault, with slip concentrated at 2.0–3.5 km depth and a peak slip of 1.6 m; the resulting slip model effectively fits the main coseismic deformation, with an RMS of 1.2 cm. Although the NE-dipping model yields a lower overall RMS misfit (1.2 cm) compared to the SW-dipping model (1.7 cm), both models effectively reproduce the overall coseismic surface deformation in ascending and descending tracks (Figure 3 and Figure S2), and geomorphological analysis heavily favors the SW-dipping model. The surface trace of the SW-dipping fault is nearly parallel to the regional topographic structural belt and closely aligns with the northern mountain edge (Figure 3c), whereas the NE-dipping trace is located south of the ZFF with no prominent geomorphic markers (Figure S2c). Therefore, we infer that the SW-dipping fault is the causative fault for the July earthquake sequence.

3.2. The November 30 Hormozgan Earthquake

Similar to the July event, no obvious surface rupture is observed, and USGS/GCMT provide two distinct focal mechanism solutions (Table S2). We adopt the same analytical strategy, constructing SW-dipping and NE-dipping fault models and inverting their geometry and slip distribution using coseismic InSAR data. In the initial nonlinear inversion phase, fault parameters are free variables and uniform slip is assumed. The SW-dipping model (strike range: 91–179°) yielded a stable and physically plausible result (fault length ~ 10 km, width ~ 5.5 km, strike ~ 100.9°, dip ~ 66.9°, rake ~ 85.0°, slip ~ 0.2 m; Table S2). Fixing the fault location, strike, and dip from nonlinear inversion, we extend the plane to 20 km × 16 km, discretize it into 320 1 km × 1 km sub-blocks, and optimize the dip via grid search (Figure 4a). The distributed slip model results (Figure 4b–h) reveal a moderately dipping (54.5°) thrust fault with slip concentrated at 1.5–6.0 km depth, a maximum slip of 0.2 m, and a seismic moment of ~5.54 × 10¹⁷ Nm (M_W 5.8); the resulting slip model effectively fits the main coseismic deformation, with an RMS of 5.8 mm. In contrast, the initial NE-dipping model (strike range: 271–359°) produces physically implausible results: a ~0.1 km width with ~10.0 m slip, approximating a line source. Following the method in Section 3.1, we fix the fault width and perform a 0.5 km interval grid search; residual reduction plateaus when slip exceeds ~0.4 m (width < 3.0 km), so we fix the width at 3.0 km and re-invert (Figure S3a). Uniform slip inversion of the NE-dipping model (Table S2) produces a strike of ~280.7°, dip of ~17.8°, and rake of ~88.3°, dominated by thrust motion with a minor left-lateral strike–slip component. Fixing the fault parameters, we extend the plane to 26 km × 22 km, discretize it into 572 1 km × 1 km sub-blocks, and optimize the dip via grid search (Figure S3b). Distributed slip model results (Figure S3c–i) show a low-dipping (10°) thrust fault, with slip concentrated at 2.4–3.1 km depth and a maximum slip ~0.3 m; the derived slip model reproduces the observed ascending and descending coseismic deformation with an RMS misfit of 5.7 mm. Although the NE-dipping model yields a lower overall RMS misfit (5.7 mm) compared to the SW-dipping model (5.8 mm), both models reasonably reproduce the observed overall coseismic surface deformation in both ascending and descending tracks. However, surface damage is concentrated near the SW-dipping fault’s trace in the Sayeh Hossein village, and the SW-dipping trace aligns with the regional topographic structural belt (northern mountain margin), whereas the NE-dipping trace lies south of the ZFF with no clear geomorphic markers (Figure 1 and Figure 2). We thus conclude the 30 November 2022 earthquake ruptured the SW-dipping seismogenic fault.

4. Discussion

4.1. Depth-Segmented Rupture of the Earthquake Sequence

Accurately resolving the seismogenic structure and rupture evolution characteristics of an earthquake sequence provides a crucial basis for revealing regional tectonic deformation and accurately assessing seismic hazards. InSAR coseismic deformation inversion results reveal that the July 2022 earthquake sequence nucleated on a previously unidentified NE-striking (a strike angle of 98.9°), moderately dipping (a dip angle of 55°) thrust fault north of the Bandar-e-Lengeh anticline. Similarly, the November 2022 earthquake ruptured a NE-striking (a strike angle of 100.9°), moderately dipping (a dip angle of 54.5°) blind thrust fault (Figure 3c and Figure 4b). The spatial locations of the two seismogenic faults are highly coincident, with a horizontal separation of only ~500 m, and their strike and dip angles are nearly identical (Figure 3c and Figure 4b). Therefore, we infer that the July–November earthquake sequence nucleated on the same seismogenic fault. To verify this hypothesis, we reprocess the coseismic cumulative deformation field covering both the July and November events (Table S3) and invert for the fault geometry and coseismic slip distribution of the seismic sequence. The inversion results indicate that the July and November earthquakes rupture one NE-striking (strike 98.7°), moderately dipping (54°) thrust fault. The coseismic slip associated with the July–November earthquake sequence is concentrated within the distance of 2.5–8 km along the downdip direction (Figure 5). To further explore the best-fit seismogenic structure, we employ the fault geometry constrained by cumulative cosemsic deformation (Figure 5) to implement independent coseismic slip inversions for the July and November events. Compared to the single-rupture models proposed by early studies [9,13,14], our joint geodetic modeling clearly isolates the depth partitioning between the events. While previous models estimated generalized slip distributions leading to high residuals, our refined inversion results show that the coseismic slip of the July earthquakes is mainly confined to a distance of 5.5–8 km along the downdip direction (Figure 6), whereas the slip for the November earthquake is concentrated in a distance of 1.5–6 km along the downdip direction (Figure 7). It should be noted that there is an ~0.5 km overlap along the downdip direction (between 5.5 km and 6.0 km). Given the inherent spatial smoothing applied during geodetic inversion and the resolution limits of InSAR observations at mid-crustal depths, the boundary between these rupture patches is transitional rather than absolute. Thus, we interpret this as a partially overlapping depth-segmented rupture, wherein the November event effectively compensated for the shallow slip deficit left by the July sequence on the fault plane. This depth-segmented rupture pattern during an earthquake sequence not only resolves the discrepancies in fault geometry observed in previous independent analyses but is also widespread across other tectonically active regions. For instance, the July 1981 M_W 7.1 Sirch and March 1998 M_W 6.6 Fandoqa thrust earthquakes ruptured the deep and shallow segments of the known Gowk fault, respectively [37]. The August 1986 M_W 6.0 and January 2016 M_W 5.9 Menyuan thrust earthquakes on the northeastern margin of the Tibetan Plateau sequentially ruptured the shallow and deep structures of the pre-existing Minle–Damaying fault [38]. Similarly, two M_W 6.3 earthquakes that appeared on the northern margin of the Qaidam Basin in November 2008 and August 2009 successively ruptured the deep and shallow segments of a previously unidentified blind thrust fault [39]. Our well-constrained geodetic modeling results (Figure 6 and Figure 7) further confirm that the July and November earthquakes nucleated on an identical thrust seismogenic fault and ruptured its deep and shallow segments, respectively, exhibiting typical depth-segmented rupture. The recognition of such depth-segmented rupture behavior of the 2022 earthquake sequence offers valuable insights into regional seismogenesis and seismic hazard assessment.

4.2. Geodetic Identification of a Back-Thrust Fault in the Zagros Mountains

Accurately identifying regional seismogenic structures is a key prerequisite for exploring earthquake genesis, tectonic evolution, and seismic hazard risk assessment. Large earthquakes have long been the focus of the academic community as they directly reflect the overall tectonic morphology of main faults. However, moderate earthquakes occurring on secondary or splay faults also possess significant research value, as they effectively supplement information regarding the structure and kinematic characteristics of the main fault, thereby allowing for a more refined and complete characterization of the subsurface structure of large fault systems [40]. Near the Zagros Mountains, four moderate-to-strong earthquakes (the M_W 6.0, M_W 5.7, and M_W 6.0 events in July 2022 and the M_W 5.7 event in November 2022) successively ruptured the northern margin of the Bandar-e-Lengeh anticline. As the primary north-dipping megathrust fault on the southern margin of the mountain front, the ZFF exhibits a listric geometry and is dominated by thrust motion with strong tectonic activity. In contrast, geodetic inversion results reveal that the July–November 2022 earthquake sequence nucleated on a moderately dipping, south-dipping planar thrust fault, with a thrusting direction opposite to the general tectonic transport direction of identified regional faults. The cross-sectional profile perpendicular to the 2022 earthquake sequence (Figure 5d) shows that the coseismic slip was mainly distributed within a depth range of 2.5–8 km and intersects the ZFF at depth, forming a complex conjugate seismogenic structure consisting of a main thrust and a secondary back-thrust. In active intra-continental tectonic belts, such coupled conjugate thrust systems often develop under compressional thrust settings and exhibit multi-scale spatial distributions. Through long-term tectonic evolution, the intervening crustal block is gradually uplifted, forming localized topographic highs [41]. The high topographic features situated between the 2022 seismogenic fault and the ZFF provide strong geomorphological evidence for the uplift controlled by the conjugate main-thrust and secondary back-thrust structure. This observation also indirectly confirms the existence of the secondary back-thrust seismogenic fault. The geodetic identification of the back-thrust fault in the 2022 sequence provides important support for the detection of complex regional subsurface geological structures and seismic hazard risk assessment.

4.3. Stress Adjustment and Regional Seismic Hazard Assessment

The spatiotemporal evolution of regional seismicity is deeply influenced by the redistribution of static stress following strong earthquakes [42,43]. To investigate the stress transfer of the 2022 Hormozgan earthquake sequence, we calculate the impact of the preceding earthquakes (source faults) on subsequent events and adjacent active faults (receiver faults) through coseismic Coulomb failure stress (CFS) changes. Using PSGRN/PSCMP software [44] and a typical friction coefficient of 0.4 [45], we calculate the coseismic CFS changes; it should be noted that negative stress values generally inhibit rupture (representing fault relaxation), while positive values promote rupture (representing increased seismic hazard) [46]. We first calculate the coseismic CFS changes generated by the preceding earthquakes (source faults) on the subsequent event (receiver fault; Figure 6) [42]. The results (Figure 8a–c) show that the early July earthquake sequence generated positive stresses of ~1.1 bar, 8.9 bar, and 30.2 bar at depths of 2 km, 4 km, and 6 km, respectively, within the peak slip zone of the 30 November event. These values significantly exceed the empirical earthquake triggering threshold of 0.1 bar [42,47], indicating that the static stress transfer induced by the July earthquakes strongly promoted the rupture initiation of the subsequent November event. To further quantify the impact of the sequence on neighboring faults, we calculate the coseismic CFS changes on mapped active faults (receiver faults) at various depths based on the optimal southwest-dipping coseismic slip distribution model of the July–November sequence (source fault, Figure 5) [45]. Based on existing survey data, geometric distributions, and slip characteristics of each active fault (Figure 1), the ZFF is assumed to be a listric pure thrust fault with an average strike of 280° and a rake of 90° (dips are set to 55°, 44°, and 31° at depths of 5 km, 8 km, and 12 km, respectively). The MFF is also assumed to be a listric pure thrust fault with an average strike of 280° and a rake of 90° (dips are set to 50°, 40°, and 30° at depths of 5 km, 8 km, and 12 km, respectively). Calculation depths of 5 km, 8 km, and 12 km are used for the coseismic CFS computation. The results (Figure 8g–i) indicate that the shallow eastern segment of the MFF is in a state of stress loading. Furthermore, significant stress loading exists on the eastern segment of the ZFF (Figure 8d–f). Considering that stress loading on the shallow segments of the eastern ZFF and eastern MFF is much higher than the 0.1 bar triggering threshold and that no strong seismic activity has occurred following this sequence, we emphasize that these fault segments face high rupture risk and should be a priority for future seismic hazard monitoring.

5. Conclusions

We utilize multi-temporal InSAR coseismic observations to investigate the high-precision coseismic surface deformation, seismogenic structure, and rupture evolution of the July–November 2022 Hormozgan earthquake sequence. Through a comprehensive analysis of regional topography, active fault distribution, and inversion results, it is concluded that the earthquake sequence ruptured a moderately dipping secondary blind back-thrust fault of the main ZFF. The coseismic slip of the July 2022 event is mainly distributed between 5.5 and 8 km downdip, while the slip of the November event is concentrated in the shallower range of 1.5–6 km downdip, overall exhibiting typical depth-segmented rupture characteristics. Given the significant stress loading, attention should be focused on the potential rupture risks of the shallow eastern segment of the MFF and the eastern segment of the ZFF.