Abstract
The northern boundary of the African plate is one of the most complex and segmented plate boundary systems worldwide, extending from the Azores Triple Junction in the Mid-Atlantic Ocean to the Gulf of Aqaba in the eastern Mediterranean. This study investigates present-day kinematics by quantifying the seismic component of crustal deformation and its relationship with the regional stress field through the combined application of seismic moment tensor summation and stress tensor inversion. A newly compiled catalogue of 784 earthquakes (Mw ≥ 4.0), recorded between 1931 and 2025, was subdivided into 17 seismogenic zones covering the principal tectonic domains of the Africa–Eurasia and Africa–Arabia plate boundaries. Seismic moment tensor summation was used to estimate seismic deformation rates and zone-averaged seismic displacement vectors, whereas stress tensor inversion constrained the orientations of the principal stress axes and the prevailing tectonic regimes. Independent GNSS (GPS) and InSAR observations were used to evaluate the consistency of the inferred regional kinematic framework. The results reveal marked spatial variations in both seismic deformation rates and tectonic style, with a west-to-east transition from extensional regimes near the Mid-Atlantic Ridge to strike-slip deformation along the Gloria Fault and the Dead Sea Transform, and compressional to transpressional regimes across the Maghreb. Although most regions exhibit relatively low seismic deformation rates (<1 mm/yr), localized concentrations of seismic strain occur in the Gulf of Cadiz, northern Algeria, and the Gulf of Aqaba. The progressive rotation of the seismic displacement vectors closely matches the first-order regional kinematic pattern derived from independent geodetic observations. The strong agreement between seismic moment tensor summation, stress tensor inversion, and geodetic data supports the proposed tectonic framework. Rather than estimating plate velocities directly, the methodology provides robust quantitative constraints on the seismic component of crustal deformation, the segmentation of the Africa–Eurasia–Arabia plate boundary, and the tectonic processes governing regional seismic hazard.
1 Introduction
Understanding how deformation is distributed along diffuse plate boundaries remains one of the major challenges in tectonics and geodynamics. Unlike narrow plate boundaries characterised by localised deformation, the study area is expressed by a broad and highly heterogeneous zone where deformation is partitioned among oceanic spreading centres, transform faults, fold-and-thrust belts, intraplate domains, and major strike-slip systems. This complexity results from the interaction between the African, Eurasian, and Arabian plates and produces significant spatial variations in tectonic style, seismicity, and strain-accommodation mechanisms. Consequently, the Africa–Eurasia plate boundary provides a natural laboratory for investigating the relationships between plate kinematics, crustal deformation, and stress-field evolution across contrasting tectonic environments. Because deformation is partitioned between seismic and aseismic processes, plate-motion estimates alone are insufficient to characterise the present-day behaviour of such a diffuse boundary. A complementary assessment of the seismic component of crustal deformation and the associated stress field is therefore required to determine how strain is distributed and released across contrasting tectonic domains. The Africa-Eurasia plate boundary extends from the Azores Triple Junction in the Mid-Atlantic Ocean to the Gulf of Aqaba in the eastern Mediterranean region (Figures 1a,b). Quantifying the rate, orientation, and style of deformation provides valuable insights into how relative plate motion is accommodated across different tectonic domains. Seismic and geodetic observations provide complementary, but not equivalent, information on present-day deformation. GNSS and InSAR measurements describe surface crustal motion over periods ranging from a few years to several decades and include both seismic and aseismic contributions. By contrast, seismic moment tensor summation quantifies only the component of deformation released seismically within a finite deforming volume during the observation period. Their combined interpretation therefore provides a more complete description of the kinematic and mechanical behaviour of diffuse plate boundaries than either approach alone and is essential for identifying zones of strain localisation and understanding the interaction among major lithospheric plates. Recent geodetic studies have refined the interpretation of deformation along northern Africa, showing that it forms a diffuse and spatially segmented zone of Africa–Eurasia convergence (). In the western sector, from southern Iberia to the Rif and Atlas domains, GNSS measurements indicate NW-SE convergence of about 4–5 mm/yr, with deformation partitioned among the Betics, Gibraltar Arc, Rif, and Moroccan Atlas (Rodríguez Collantes et al., 2025). In the eastern sector, GPS and InSAR observations from Algeria and Tunisia similarly reveal partitioned deformation, with offshore shortening, strike-slip faulting, transpression, and distributed strain across the Atlas belt (; ; Viltres et al., 2025). Together, these results highlight the strong control exerted by inherited structures and major fault systems on present-day kinematics across the entire boundary. Viltres et al. (2025) based on InSAR investigations show almost the same values for velocity field over northern Algeria and Tunisia, with 4.9 mm/yr in western Algeria, 5.0 mm/yr in Constantine, 3.9 mm/yr in eastern Algeria and 5.2 mm/yr for Tunisia with an NW-SE oblique convergence direction respectively 55° in western Algeria, 50° in Constantine, 40° in eastern Algeria and 46.5° in Tunisia. More investigations are required to complete the scheme in Azores, Morocco, Libya, Egypt and the Gulf of Aqaba.
FIGURE 1
The northern part of the African plate is characterised by significant seismic activity, reflecting the ongoing interaction between the African, Eurasian, and Arabian plates. Several destructive earthquakes have occurred along this boundary during the historical and instrumental periods, in the west along the Gibraltar-Azores region (e.g., The 1755 Lisbon Mw 8.5 earthquake, (Grandin et al., 2007a; Grandin et al., 2007b; Zitellini et al., 2001); the 1969 offshore St. Vincent Cape Ms 8.0 earthquake (Grandin et al., 2007a; Grandin et al., 2007b; Buforn et al., 1988); and the 26 may 1975 North Atlantic earthquake Ms 7.9 (Lynnes and Ruff, 1985).
In the central part along the Maghreb region (e.g., the 1954 Orléansville Mw 6.5 earthquake, Espinoza and Lopez-Arroyo, 1984; the 1980 El-Asnam Ms 7.3 earthquake; Ouyed et al., 1981; the 2003 Zemmouri Mw 6.8 earthquake; ; and the 2004 Al-Hoceima Mw 6.3 earthquake; Stich et al., 2005), and also in the easternmost part of the plate boundary along the Dead Sea transform (e.g., The 1995 Gulf of Aqaba Mw 7.2, Hofstetter, 2003). These events underline the importance of understanding the spatial distribution of deformation along the northern boundary of the African plate to characterise its seismotectonic behaviour more accurately. They also illustrate the highly heterogeneous release of seismic moment along the plate boundary and emphasise the need to analyse deformation within distinct tectonic domains rather than treating the entire boundary as a single homogeneous system.
Several studies have investigated the present-day kinematics and deformation of the Africa–Eurasia plate boundary using complementary seismological and geodetic approaches. Early studies based on seismic moment tensor summation demonstrated that earthquake focal mechanisms and cumulative seismic moment release can be used to quantify the seismic component of crustal deformation and to identify regional patterns of strain within the Mediterranean domain (Jackson and McKenzie, 1988; Ekström and England, 1989). A subsequent regional application by Kiratzi and Papazachos (1995), extending along the Africa–Eurasia plate boundary from the Azores to the Middle East, revealed strong spatial variability in seismic deformation rates and tectonic regimes, including strike-slip deformation along major transform structures such as the Gloria Fault and the Dead Sea Transform, and compressional regimes across the Mediterranean domain. However, the resulting rates represent deformation released seismically within finite volumes and observation periods and should not be interpreted as direct estimates of long-term plate velocities, because they are sensitive to earthquake recurrence, catalogue duration, deforming-volume dimensions, rigidity assumptions, and the occurrence of large earthquakes.
In parallel, geodetic observations have considerably improved constraints on present-day plate kinematics by directly measuring crustal motion at the Earth’s surface. Global models such as NUVEL-1 () and REVEL (Sella et al., 2002) provided first-order estimates of relative plate velocities and rotation poles, whereas regional GNSS studies refined the kinematics of the Africa–Eurasia plate boundary and revealed slow convergence rates and a complex, spatially distributed deformation field (e.g., Sella et al., 2002; McClusky et al., 2003; Reilinger et al., 2006; Nocquet, 2012). Unlike seismic moment tensor summation, which estimates only the component of deformation released seismically during the catalogue period, GNSS-derived velocity fields describe total present-day surface motion, including both seismic and aseismic contributions. The two approaches should therefore be regarded as complementary rather than directly equivalent. Integrated studies combining GPS velocities and seismicity data have further demonstrated that the plate boundary is segmented into distinct tectonic domains characterised by contrasting deformation styles and strain-partitioning mechanisms (e.g., Serpelloni et al., 2007; Madarieta-Txurruka et al., 2025). In addition to these kinematic investigations, numerous studies have characterised the regional stress field through the inversion of earthquake focal mechanisms (e.g., Ousadou et al., 2014; 2024; Soumaya et al., 2015). These analyses reveal pronounced spatial variations in stress regime, ranging from extension along the Mid-Atlantic Ridge and within the Azores domain, through transtensional to strike-slip regime along the Gloria Fault and the Gorringe Bank–Gulf of Cadiz–Betics sector, to compressional and transpressional regimes across the Maghreb region (Ousadou et al., 2014; 2024; Soumaya et al., 2015). Farther east, strike-slip and locally extensional stress regimes reflect the increasingly complex interaction between the African and Arabian plates. Stress inversion studies in North Africa also highlight the influence of inherited lithospheric structures and regional slab dynamics, particularly those associated with the Calabrian Arc, on the observed stress patterns and crustal deformation.
Despite these advances, most previous studies have remained restricted either to individual segments of the plate boundary or to a single methodological approach. The present study extends these efforts both spatially and temporally by examining the entire northern boundary of the African plate, from the Azores Triple Junction to the Gulf of Aqaba, within a unified framework combining seismic moment tensor summation and stress tensor inversion. Its originality lies in the integrated interpretation of the seismic component of deformation and the regional stress field across the full Africa–Eurasia–Arabia plate-boundary system using an updated and homogenised focal-mechanism catalogue spanning the period 1931–2025. This framework allows the spatial distribution of seismic strain release, the orientations of the principal stress axes, and the tectonic segmentation of the plate boundary to be assessed consistently.
The complementary application of seismic moment tensor summation and stress tensor inversion makes it possible to investigate not only where deformation is released seismically, but also the stress conditions under which this release occurs. Whereas seismic moment tensor summation quantifies the spatial distribution and orientation of seismic strain accumulated during the observation period, stress tensor inversion constrains the orientations and relative magnitudes of the principal stresses and identifies the prevailing tectonic regimes. Their combined interpretation therefore provides a physically consistent framework for analysing the present-day behaviour of a complex, diffuse, and highly segmented plate boundary.
In this study, we investigate the present-day deformation along the northern boundary of the African plate by combining seismic moment tensor summation (Kostrov, 1974; Jackson and McKenzie, 1988) and stress tensor inversion (Michael, 1984; 1987a; Heidbach et al., 2007; Heidbach et al., 2010). A catalogue of 784 earthquakes with moment magnitude Mw ≥ 4.0, occurring between 1931 and 2025, was compiled and divided into 17 seismogenic zones extending from the Azores Triple Junction to the Gulf of Aqaba. The seismic moment tensor summation method is used to estimate seismic strain rates and zone-averaged seismic displacement vectors within each zone, whereas stress tensor inversion of the focal mechanisms constrains the orientations of the principal stress axes and the prevailing tectonic regime. By integrating these complementary approaches, the study does not seek to estimate long-term plate velocities directly. Instead, it aims to provide a robust assessment of the spatial distribution of the seismic component of crustal deformation, the regional stress field, and the tectonic segmentation of the Africa–Eurasia–Arabia plate boundary. Published GNSS and InSAR observations are used as independent constraints to evaluate the consistency of the inferred regional kinematic pattern, rather than to establish direct quantitative equivalence between geodetic plate velocities and seismic displacement rates.
2 Seismotectonics of the northern African plate boundary
The northern boundary of the African plate constitutes one of the world’s most complex and laterally variable plate-boundary systems, reflecting the long-term interaction between the African, Eurasian and, in its easternmost sector, Arabian plates. Present-day deformation is primarily driven by the ongoing convergence between the African and Eurasian plates, a process initiated during the Late Cretaceous that continues today at rates of a few millimetres per year (DeMets et al., 1994; McClusky et al., 2003; Nocquet and Calais, 2004; Meghraoui and Pondrelli, 2012). This relative motion is accommodated through a combination of oceanic spreading, transform and strike-slip faulting, crustal shortening, thrusting, and localised extension, producing a broad and highly segmented deformation zone. The resulting spatial variability in tectonic style, seismicity and stress regime makes this boundary a natural laboratory for investigating the relationships between plate kinematics, seismic strain release and lithospheric stress. It is important, however, to distinguish long-term plate motion from the seismic component of deformation, which reflects only the strain released by earthquakes within a finite observation period and deforming volume.
At the western end of the plate boundary, deformation is controlled by the Azores Triple Junction, where the North American, Eurasian and African (Nubian) plates meet. Rather than forming a simple and stable ridge–ridge–ridge configuration, this region is characterised by a diffuse plate-boundary zone involving the Mid-Atlantic Ridge, the Terceira Rift and the Azores–Gibraltar fault system. Seismicity is distributed across a broad area and reflects a combination of oblique extension, strike-slip deformation and localised normal faulting (Figure 2). The Gloria Fault, which forms the central transform segment of the Azores–Gibraltar system, accommodates predominantly dextral strike-slip motion and has hosted several large historical earthquakes, confirming its role as a major lithospheric boundary structure (Marques et al., 2025). Eastwards, the plate boundary progressively changes from a predominantly oceanic transform domain to a broader and more complex deformation zone towards the Gibraltar Arc, where oblique convergence is partitioned between strike-slip faulting, thrusting and crustal shortening (Serpelloni et al., 2007; Ousadou et al., 2014).
FIGURE 2
From the Gibraltar Arc eastwards, Africa–Eurasia convergence is largely accommodated across the Maghreb region, which comprises a series of intracontinental mountain belts, foreland basins and inherited structural domains developed during the Alpine orogeny (Doglioni et al., 1999; Frizon de Lamotte et al., 2000; Hallet, 2002; Moustafa et al., 2022). These include the Rif and the Middle and High Atlas ranges in Morocco, the Tell Atlas in northern Algeria and the Tunisian Atlas farther east. Seismicity is concentrated within a relatively narrow but internally diffuse belt along both sides of the plate boundary (Figure 1a). Most earthquakes are shallow and their focal mechanisms indicate a mixture of reverse, transpressional and strike-slip faulting (; Yelles-Chaouche et al., 2006; Serpelloni et al., 2007). Shallow seismicity (h < 70 km) dominates throughout the region, whereas intermediate-depth earthquakes (h ≥ 70 km) occur mainly beneath the Alboran Sea and locally beneath the High Atlas of northern Morocco (Udías and Buforn, 1991; Buforn et al., 2004). The Rif domain is among the most seismically active sectors and is characterised by moderate-to-strong earthquakes, predominantly strike-slip focal mechanisms and local normal faulting related to complex crustal deformation. The 2004 Al-Hoceima Mw 6.3 earthquake (Stich et al., 2005; Van der Woerd et al., 2014) illustrates the significant seismic potential of this tectonically heterogeneous region.
Farther east, the Tell Atlas of Algeria and its continuation into Tunisia are dominated by compressional and transpressional structures inherited from successive Miocene to Quaternary tectonic phases. Seismicity is largely associated with E–W- to NE–SW-trending folds, thrusts and reverse faults, many of which result from the reactivation of structures formed during earlier extensional episodes (Yelles-Chaouche et al., 2006). Some of the most destructive earthquakes in North Africa have occurred within this domain, including the 1954 Orléansville Mw 6.5 earthquake (Espinoza and Lopez-Arroyo, 1984), the 1980 El-Asnam Ms 7.3 earthquake (Ouyed et al., 1981) and the 2003 Zemmouri Mw 6.8 earthquake (). These events demonstrate the important role of active crustal shortening and transpression in accommodating Africa–Eurasia convergence. Despite relatively slow regional convergence rates (Nocquet and Calais, 2004; Serpelloni et al., 2007; ; ), deformation is strongly localised within intramontane basins and along major fault systems (Perrodon, 1957; Yelles-Chaouche et al., 2006), reflecting the pronounced mechanical and structural heterogeneity of the continental lithosphere.
East of Tunisia, the northern African margin enters a domain characterised by lower overall seismic activity, comparatively low deformation rates and a more rigid lithospheric behaviour (Sïdi et al., 2011; ). In Libya, seismicity is concentrated mainly along the Mediterranean coastal zone, particularly within the Tripolitania Basin and the Cyrenaica Platform (; Salem et al., 2008). Cyrenaica, which forms part of a Mesozoic foreland domain (), exhibits moderate seismicity associated with the reactivation of inherited ENE–WSW-trending folds and faults (). Available focal mechanisms indicate predominantly strike-slip deformation, with local normal and reverse faulting that may reflect far-field stresses transmitted from the Hellenic subduction system to the north (). The Tripolitania Basin, by contrast, shows evidence of localised extension along NW–SE-trending normal faults, although its overall seismic activity remains lower than that observed across the Maghreb region.
In Egypt and the northeastern African region, the tectonic framework becomes increasingly complex because of the interaction among the African, Arabian and Eurasian plates. Deformation is partitioned between the Gulf of Suez rift system, characterised by NNW–SSE-trending normal faults, and the Gulf of Aqaba, which forms the southernmost segment of the Dead Sea Transform (Hussein et al., 2013). The Dead Sea Transform is a major left-lateral strike-slip plate boundary extending for more than 1,000 km and plays a fundamental role in accommodating relative motion between the African and Arabian plates (Salamon et al., 1996; El-Isa, 2013). The Gulf of Aqaba is characterised by a series of pull-apart basins generated by transtensional deformation and oblique extension along the Dead Sea Transform system, making it one of the most seismically active regions in northeastern Africa (El-Isa, 2013). The 1995 Mw 7.2 Gulf of Aqaba earthquake illustrates the substantial seismic hazard associated with this plate-boundary segment (Hofstetter, 2003). Together, the Gulfs of Suez and Aqaba record the transition from intracontinental extension within the African plate to active transform deformation linking the Red Sea spreading system with the broader Africa–Eurasia–Arabia plate-boundary network.
Taken together, these contrasting tectonic domains demonstrate that the northern African plate boundary cannot be described as a single, mechanically homogeneous system. Instead, it comprises a succession of oceanic spreading centres, transform faults, convergent margins, intracontinental belts and rift–transform systems, each characterised by distinct styles of seismic strain release and stress orientation. This segmentation provides the geological basis for subdividing the study area into separate seismogenic zones and for analysing the seismic component of deformation and the regional stress field within tectonically coherent domains.
3 Data and methods
3.1 Seismic moment tensor summation method
The seismic moment tensor summation method (Kostrov, 1974; Jackson and McKenzie, 1988) provides a quantitative estimate of the seismic component of crustal deformation by integrating earthquake source parameters over a defined spatial domain and observation period. The method links earthquake source mechanics to regional deformation by assuming that the cumulative seismic moment released by all earthquakes within a defined volume represents the average seismic strain accumulated in that volume. The resulting quantities therefore describe zone-averaged seismic deformation and should not be interpreted as direct estimates of long-term plate motion or point-wise geodetic velocity.
Each investigated zone is represented by a three-dimensional deforming volume defined by its length (l1), width (l2), and thickness (l3). The longest horizontal dimension (l1) is aligned with the dominant tectonic trend and forms an angle (ξ) with geographic north. The condition l1 ≫ l2 and l3 is adopted so that deformation can be approximated as laterally homogeneous within each elongated tectonic domain. The deforming volume is therefore expressed as (). is therefore defined as . The thickness (l3) corresponds to the depth interval between the shallowest and the deepest earthquakes included in each zone. It thus represents the thickness of the seismically active deforming volume rather than the total crustal or lithospheric thickness.
For a set of N earthquakes occurring within a time interval T, the average seismic strain-rate tensor () is obtained by summing the individual seismic moment tensors and normalising by the shear modulus, deforming volume, and catalogue duration (Equation 1):
Here, (μ) is the shear modulus, (V) is the deforming volume, and (T) is the duration of the focal-mechanism catalogue for the considered zone.
Different shear-modulus (μ) values were adopted according to the tectonic setting of the investigated domains. For Zones 1–5 located within the Atlantic oceanic domain, a rigidity coefficient of μ = 4.5 × 1010 N/m2 was used to account for the comparatively higher rigidity of oceanic lithosphere and upper mantle materials. This value was inferred from seismic velocity-density relationships proposed by and Mooney et al. (2010). For the remaining zones, which mainly correspond to continental and transitional crustal domains, a standard crustal shear modulus value of μ = 3.3 × 1010 N/m2 was adopted, consistently with previous regional applications of seismic moment summation (e.g., Kiratzi and Papazachos, 1995).
For each zone, the time interval (T) corresponds to the longest period for which sufficiently reliable focal mechanism data were available. The estimated deformation rates therefore represent seismic deformation released during the analysed catalogue period rather than strictly steady stationary long-term tectonic rate. Because seismic moment increases non-linearly with earthquake magnitude, one or a few moderate-to-large earthquakes may dominate the cumulative seismic moment in zones characterised by sparse seismicity. Localised seismic displacement rates may therefore exceed geodetic or plate-kinematic rates without implying anomalously rapid plate motion. This sensitivity to catalogue duration, earthquake recurrence, event magnitude, and deforming-volume dimensions is an intrinsic limitation of seismic moment tensor summation applied to finite earthquake catalogues.
For each earthquake, the seismic moment tensor is constructed from the scalar seismic moment (M0) and the focal-mechanism parameters (strike λ, dip δ, rake θ), following the standard double-couple source formulations of . The moment tensor of the nth earthquake is written as Equation 2:where Fij is a symmetric, dimensionless second-order tensor describing fault orientation and slip direction as a function of ( , , ). Using the adopted Cartesian reference frame and the double-couple convention of , the six independent moment-tensor components are expressed as follows Equations 3–8:
The seismic strain-rate tensor is subsequently converted into zone-averaged seismic displacement-rate components by multiplying the relevant strain-rate terms by the characteristic dimensions of the deforming volume. Following Jackson and McKenzie (1988), the off-diagonal components are written as Equation 9:whereas the diagonal components are expressed as Equation 10:
In these expressions, is the duration of the earthquake catalogue, is the shear modulus, and , , and are the characteristic dimensions of the deforming zone. The quantities and represent zone-averaged seismic displacement-rate components derived from the cumulative seismic moment released during the analysed catalogue period. They should not be interpreted as point-wise surface velocities equivalent to GNSS station measurements.
All calculations are initially performed in a local reference frame aligned with the principal tectonic axis of each zone. The resulting deformation and displacement-rate tensors are then rotated into a common Cartesian reference frame (x, y, z), in which x is positive northward, y eastward, and z upward. This transformation allows the horizontal components to be interpreted consistently in terms of boundary-normal and boundary-parallel seismic deformation along the Africa–Eurasia and Africa–Arabia plate boundaries.
The horizontal deformation tensor is diagonalised to obtain its principal eigenvalues and corresponding eigenvectors. Positive eigenvalue indicate extension, whereas negative eigenvalues indicate shortening. The relative magnitudes and signs of the two horizontal principal deformation rates are used to characterise the dominant kinematic regime. A dominant negative eigenvalue indicates predominantly compressional deformation, a dominant positive eigenvalue indicates predominantly extensional deformation, and eigenvalues of comparable magnitude but opposite sign indicate a strike-slip or shear-dominated regime. The associated eigenvectors define the orientations of the principal horizontal extension and shortening axes. This eigenvalue analysis provides a concise description of the dominant seismic deformation style within each tectonic zone.
To assess the robustness of the estimated deformation tensors and seismic displacement vectors, a leave-one-out jackknife resampling analysis was performed independently for each zone. For each zone containing earthquakes, a leave-one-out jackknife resampling procedure was applied. The calculations were repeated times, each iteration excluding one earthquake while retaining the remaining events. The complete seismic moment tensor summation procedure was then repeated for each reduced dataset, and the corresponding deformation tensor, principal deformation rates, principal directions, and resultant seismic displacement vector were recalculated. The sensitivity of each parameter was quantified from the mean and standard deviation of the jackknife realisations. Circular statistics were applied to azimuthal parameters in order to account for their periodic character. This procedure provides quantitative uncertainty estimates and identifies zones in which the inferred deformation pattern is disproportionately influenced by a small number of earthquakes or by one particularly large event.
Although the seismic moment tensor summation provides an integrated estimate of seismic component of deformation rates and its principal directions, it does not directly determine the stress state responsible for fault activation. It describes the cumulative kinematic expression of earthquake slip within each deforming volume over the analysed catalogue period. The method is therefore complemented by stress tensor inversion, which constrains the orientations and relative magnitudes of the principal stresses and identifies the prevailing tectonic regime. The combined use of the two approaches provides complementary kinematic and mechanical descriptions of deformation along the Africa–Eurasia–Arabia plate-boundary system.
3.2 Stress tensor inversion
To complement the kinematic information obtained from seismic moment tensor summation, the present-day stress field along the Africa–Eurasia-Arabia plate boundary was investigated through stress tensor inversion of earthquakes. Whereas seismic moment tensor summation quantifies the seismic component of crustal deformation accumulated during the analysed catalogue period, stress tensor inversion constrains the orientation and relative magnitudes of the principal stresses responsible for fault activation. The combined interpretation of these complementary approaches provides a consistent kinematic and mechanical description of active deformation along the plate boundary.
The stress tensor was determined using the procedure of Michael (1984), (1987), subsequently extended to spatially varying stress fields by Hardebeck and Michael (2006), following the stress-regime conventions adopted by the World Stress Map project (Heidbach et al., 2018). The inversion assumes that earthquake slip occurs in the direction of maximum resolved shear stress on pre-existing faults and that the stress field remains homogeneous within each investigated tectonic zone.
The inversion relies on the same focal-mechanism dataset used for the seismic moment tensor summation, ensuring full consistency between the kinematic and mechanical analyses. Only well-constrained double-couple focal mechanisms were considered, and each tectonic zone was analysed independently to account for lateral variations in stress regime. This zonal subdivision is consistent with the tectonic segmentation adopted for the seismic moment tensor summation, allowing direct comparison between the inferred stress field and the corresponding seismic deformation tensors.
The inversion assumes: (1) the stress tensor is homogeneous within each investigated zone, (2) slip occurs in the direction of maximum resolved shear stress on the selected fault plane, (3) earthquakes are independent samples of the same regional stress field, and (4) pore-pressure variations do not significantly modify the orientation of the principal stresses at the scale considered. These assumptions are commonly adopted in regional stress inversion studies and are considered appropriate for the investigated plate-boundary segments.
The stress tensor is represented by a symmetric second-order tensor (Equation 11):
Because the stress tensor is symmetric, only six independent components are required to describe the complete three-dimensional state of stress.
Diagonalisation of the stress tensor yields the three principal stresses (σ1 ≥ σ2 ≥ σ3) together with their orientations. The stress state is further characterised by the dimensionless stress shape ratio (Equation 12):which describes the relative position of the intermediate principal stress between the maximum and minimum principal stresses (; ). Values approaching zero indicate that is close to , whereas values approaching one indicate that approaches . The tectonic regime is determined from the combined interpretation of the principal stress orientations and the stress shape ratio following the World Stress Map conventions (Heidbach et al., 2018). This parameter ranges from 0 to 1 for all three Andersonian stress regimes (compressional, strike-slip, and extensional). The primary role of the stress shape ratio R is to describe the relative magnitude of the intermediate principal stress σ2 with respect to σ1 and σ3, and it controls the relative position of σ2 between σ1 and σ3. Values such as 0 or 1 are reached when σ2 is close to σ3 or σ1 (; ). The stress regime is determined by combining the orientation of the principal stress axes, particularly the orientation of σ1 relative to the vertical, following the standard World Stress Map (WSM) conventions (Heidbach et al., 2018).
For a fault plane with unit normal vector n, the traction vector t acting on the plane is Equation 13:
The shear-stress vector τ on the fault plane is obtained by removing the normal component of traction (Equation 14):
The predicted slip direction is assumed to be parallel to the shear stress vector. The inversion seeks the stress tensor that minimises the angular misfit between the observed slip direction and the predicted direction of maximum shear stress on the fault plane. The misfit angle α is defined as Equation 15:where is the observed slip vector, and is the theoretical slip direction predicted by the stress tensor.
The objective function is therefore based on minimising the average angular misfit between observed and predicted slip directions over all focal mechanisms included in each tectonic zone.
For each focal mechanism, both nodal planes are considered as potential fault planes. The optimal stress tensor is obtained by minimising the average misfit over the entire dataset within each zone. The reliability of the inverted stress tensor is evaluated using the mean misfit angle and its standard deviation. Low average misfit values indicate a good consistency between the observed focal mechanisms and the inferred stress field. Only solutions characterised by low average misfit angles and internally consistent confidence intervals were retained for tectonic interpretation. The average misfit angle and its standard deviation therefore provide quantitative measures of inversion quality and stress-field homogeneity within each investigated zone.
Uncertainties in the orientations of the principal stress axes were estimated using the bootstrap resampling procedures implemented in Michael’s inversion package (Michael, 1987; Hardebeck and Michael, 2004; Hardebeck and Michael, 2006). Confidence intervals were computed for the orientations of , , and , providing quantitative estimates of solution robustness and allowing meaningful comparison between neighbouring tectonic domains.
The resulting stress tensors are interpreted in terms of the tectonic regime by analysing the orientation of σ1 and σ3 relative to the plate boundary trends. The resulting stress tensors were interpreted according to the orientations of the principal stresses relative to the vertical axis and to the regional tectonic trends. Zones characterised by horizontal and vertical correspond to compressional regimes; zones with vertical indicate extensional regimes, whereas zones with vertical correspond to strike-slip faulting, following the Andersonian fault classification adopted by the World Stress Map project.
By combining stress tensor inversion with seismic moment tensor summation, the present study provides complementary mechanical and kinematic descriptions of active deformation along the Africa–Eurasia–Arabia plate-boundary system. Whereas stress tensor inversion constrains the present-day regional stress field, seismic moment tensor summation quantifies the seismic component of crustal deformation accumulated during the analysed catalogue period. Their integrated interpretation provides a robust framework for investigating tectonic segmentation, deformation partitioning, and the relationships between stress, seismic strain release, and plate-boundary architecture from the Azores Triple Junction to the Gulf of Aqaba.
3.3 Database
This study is based on a newly compiled earthquake catalogue (Supplementary Table S1) specifically developed to investigate the seismic component of present-day crustal deformation and the associated stress field along the Africa–Eurasia–Arabia plate boundary, extending from the Azores Triple Junction to the Gulf of Aqaba. The catalogue comprises 784 earthquakes distributed over 17 seismogenic zones and covers the period from 1931 to 2025. The geometric characteristics of each zone are summarised in Table 1, whereas the corresponding focal mechanisms are shown in Figure 2. The catalogue forms the common database for both the seismic moment tensor summation and the stress tensor inversion, ensuring full methodological consistency between the kinematic and mechanical analyses.
TABLE 1
| Zone | Lx | Ly | Lz | θ | ΔT |
|---|---|---|---|---|---|
| 1 | 588 | 175 | 14 | 40 | 38 |
| 2 | 731 | 243 | 23 | 9 | 39 |
| 3 | 532 | 180 | 14 | 120 | 43 |
| 4 | 470 | 335 | 31 | 90 | 89 |
| 5 | 580 | 196 | 42 | 90 | 45 |
| 6 | 400 | 145 | 65 | 90 | 34 |
| 7 | 400 | 240 | 38 | 90 | 63 |
| 8 | 620 | 180 | 26 | 80 | 70 |
| 9 | 290 | 230 | 30 | 110 | 57 |
| 10 | 400 | 240 | 28 | 0 | 68 |
| 11 | 410 | 225 | 22 | 10 | 28 |
| 12 | 526 | 210 | 11 | 145 | 83 |
| 13 | 570 | 330 | 31 | 97 | 57 |
| 14 | 370 | 260 | 29 | 95 | 61 |
| 15 | 235 | 210 | 29 | 160 | 43 |
| 16 | 300 | 110 | 29 | 135 | 54 |
| 17 | 180 | 130 | 16 | 18 | 35 |
Characteristics of the selected seismogenic zones.
Lx, Length in km; Ly, Width in km; Lz, Thickness in km; θ, orientation in degrees of the Lx axis; ΔT, Time period in years between the first and last event for each zone.
Only earthquakes with moment magnitude Mw ≥ 4.0 were selected to ensure reliable focal-mechanism solutions and robust seismic moment estimates. This threshold represents a compromise between catalogue completeness and data quality and guarantees the availability of source parameters required for both seismic moment tensor summation and stress tensor inversion. Although smaller earthquakes contribute to distributed deformation, the cumulative seismic moment is overwhelmingly controlled by moderate and large earthquakes, which therefore dominate the estimated seismic deformation rates. Consequently, the adopted magnitude threshold is considered appropriate for quantifying the regional seismic component of deformation investigated in this study.
The compiled focal mechanism catalogue inevitably remains heterogeneous in both space and time, particularly during the early instrumental period, when earthquake locations, focal mechanisms, and seismic moments are generally less well constrained than those derived from modern centroid moment tensor solutions. Consequently, the derived seismic deformation tensors and seismic displacement rates should be regarded as first-order regional estimates of the seismic component of crustal deformation accumulated during the analysed catalogue period rather than as direct measures of long-term plate motion.
For each event, the catalogue includes the hypocentral parameters, focal-mechanism solutions (strike, dip, and rake for both nodal planes), moment magnitude, and, whenever available, the scalar seismic moment. The dataset was carefully homogenised to ensure consistency across tectonic domains, catalogue sources, and time periods. It was compiled primarily from the Global CMT catalogue (Ekström et al., 2012), MED_RCMT (Pondrelli et al., 2011), ISC-GEM (Storchak et al., 2013), and ETHZ (), while additional information was obtained from published studies where necessary. The source of each focal mechanism is reported in Supplementary Table S1.
To ensure methodological consistency, all earthquakes were homogenised using moment magnitude (Mw) as the common reference magnitude whenever available. For earthquakes originally reported using other magnitude scales (e.g., Ms or Mb), published equivalent Mw values or scalar seismic moments reported in the original references were adopted. Although minor inconsistencies inevitably remain because of differences in inversion procedures, velocity models, centroid-depth determinations, and waveform quality, particularly for historical earthquakes, these uncertainties are not expected to modify the first-order regional patterns of seismic deformation or the inferred stress field presented in this study.
We have divided our study area into several seismogenic zones, considering the major tectonic structures and the seismicity patterns. Each seismogenic zone was defined to encompass coherent fault systems and to ensure compatibility with the assumptions of both seismic moment tensor summation and stress tensor inversion methods. The same zoning scheme and earthquake catalogue were systematically used for both methodologies, allowing direct comparison between kinematic and stress results.
The nearly one-century duration of the catalogue allows the cumulative seismic moment released by the largest earthquakes to be incorporated into the analysis and provides a representative picture of regional seismic deformation over the investigated period. Nevertheless, the derived seismic deformation rates remain dependent on catalogue duration, earthquake recurrence, and the spatial distribution of seismic moment release. They should therefore be interpreted as estimates of the seismic component of crustal deformation rather than direct measurements of long-term plate motion. This distinction is fundamental for the interpretation of the results presented in the following sections.
Because seismic moment tensor summation integrates the cumulative seismic moment released within each deforming volume, the resulting deformation estimates are inherently sensitive to the occurrence of moderate-to-large earthquakes. This characteristic does not constitute a limitation specific to the present catalogue but reflects a fundamental property of seismic deformation estimates derived from earthquake source parameters. The robustness of the inferred deformation tensors was therefore evaluated independently for each zone using a leave-one-out jackknife resampling procedure (Section 3.1).
4 Results
4.1 Seismic moment tensor summation results
Seismic moment tensor summation was applied to the 17 seismogenic zones defined along the northern boundary of the African plate, using a catalogue of 784 earthquakes. The resulting seismic deformation tensors provide quantitative estimates of the zone-averaged seismic displacement rates and principal deformation directions for each zone. The results are summarised in Table 2, and the corresponding seismic displacement vectors are shown in Figure 3.
TABLE 2
| Z | Moment tensor summation | Stress tensor inversion | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| λ1 | Ω1 | λ 2 | Ω2 | V | Az | Tectonic regime | σ1 N°/pl° | σ2 N°/pl° | σ3 N°/pl° | Stress regime variation | Stress regime | |
| 1 | 0.38 | 355 | −0.09 | 264 | 0.39 ± 0.02 | 341 ± 1 | Extension | Vertical | 192/6 | 282/1 | 0.40 ( > >) | Pure extensional tri-axial normal stress |
| 2 | 0.14 | 302 | 0.06 | 32 | 0.15 ± 0.003 | 332 ± 3 | Extension | Vertical | 224/2 | 314/05 | 0.86 ( ∼ ) | Transtensional stress regime dominated by normal stress |
| 3 | −1.86 | 231 | 1.02 | 141 | 2.12 ± 0.2 | 260 ± 19 | Strike-slip | Vertical | N316/2 | N226/05 | 0.57 ( > >) | Pure extensional tri-axial normal stress |
| 4 | −15.71 | 315 | 12.76 | 45 | 20.40 ± 5 | 270 ± 39 | Strike-slip | N338/27 | Vertical | N253/6 | 0.8 ( ∼ ) | Transtensional strike regime dominated by strike slip |
| 5 | −18.08 | 282 | 1.46 | 192 | 18.14 ± 3.2 | 279 ± 1 | Compression | N339/05 | Vertical | N245/30 | 0.2 ( ∼ ) | Transpressional stress regime with dominated strike-slip |
| 6 | −0.006 | 296 | 0.004 | 20 | 0.01 ± 0.001 | 325 ± 4 | Strike-slip | N329/20 | Vertical | N247/20 | 0.3 ( ∼ ) | Transpressional stress regime with dominated strike-slip |
| 7 | −0.28 | 296 | 0.19 | 26 | 0.34 ± 0.02 | 330 ± 2 | Strike-slip | N342/13 | Vertical | N251/06 | 0.5 ( > >) | Pure tri-axial strike stress |
| 8 | −5.04 | 305 | 0.29 | 215 | 5.05 ± 0.34 | 308 ± 1 | Compression | N329/10 | 235/19 | Vertical | 0.0 () | Transpressional stress dominated by compressive stress regime |
| 9 | −0.1 | 265 | 0.04 | 176 | 0.11 ± 0.01 | 286 ± 2 | Strike-slip | 339N/15 | 252N/12 | Vertical | 0.2 ( ∼ ) | Transpressional stress dominated by compressive stress regime |
| 10 | −0.02 | 40 | 0.01 | 310 | 0.03 ± 0.002 | 320 ± 1 | Strike-slip | 319N/4 | 236/55 Oblic | 227/34 Oblic | 0.3 ( ∼ ) | Compressional strike-slip regime with rotation of σ2 and σ3 around σ1 |
| 11 | −0.02 | 31 | 0.01 | 301 | 0.02 ± 0.001 | 346 ± 2 | Strike-slip | 313N/3 | Vertical | 222/10 | 0.4 ( > >) | Pure strike stress tri-axial regime |
| 12 | 0.63 | 233 | −0.37 | 143 | 0.73 ± 0.14 | 298 ± 12 | Strike-slip | Vertical | 299N/25 | 209/1 | 0.83 ( ∼ ) | Transtensional stress regime with dominated normal stress |
| 13 | −0.06 | 308 | 0.004 | 218 | 0.07 ± 0.01 | 312 ± 18 | Compression | 330/18 | 220/45 oblic | 257/39 oblic | 0.3 ( ∼ ) | Transpressional regime with revolution of σ2 and σ3 around σ1 |
| 14 | −0.12 | 322 | −0.002 | 232 | 0.12 ± 0.03 | 322 ± 35 | Compression | 266/10 | 359N/20 | Vertical | 0.5 ( > >) | Pure compressional tri axial stress regime |
| 15 | 0.08 | 344 | −0.02 | 164 | 0.08 ± 0.01 | 351 ± 2 | Extension | Vertical | 294N/16 | 203/4 | 0.6 ( > >) | Pure normal tri axial stress regime |
| 16 | 0.68 | 227 | −0.07 | 317 | 0.68 ± 0.08 | 341 ± 5 | Extension | Vertical | 295N/2 | 204/4 | 0.46 ( > >) | Pure normal tri axial stress regime |
| 17 | 12.6 | 98 | −11.59 | 8.0 | 17.12 ± 3.6 | 342 ± 3 | Strike-slip | 333/30 | 306/57 | 236/12 | 0.7 ( ∼ ) | Transtensional strike regime dominated by strike slip |
Results of the moment tensor summation and stress inversion methods.
λ and Ω correspond to the eigenvalues and eigenvectors, respectively. V represent the seismic displacement rate in mm/yr, and Az represent the azimuth in degrees. σ N°/pl, Azimuth and plunge of the maximum (σ1), intermediate (σ2), and minimum (σ3) principal stress. R is the ship ratio.
FIGURE 3
The tectonic regimes inferred from the eigenvalue analysis (Table 2) further highlight the pronounced segmentation of the northern African plate boundary. Extensional deformation dominates the Atlantic spreading domain, particularly in Zones 1 and 2, whereas strike-slip deformation becomes prevalent along the Azores–Gibraltar fracture system and the Dead Sea Transform, notably in Zones 4 and 17. Zone 3 exhibits a strike-slip-dominated deformation pattern, although this result is comparatively sensitive to the limited number and distribution of available earthquakes. Compressional to transpressional regimes characterise much of the Maghreb domain, particularly northern Algeria, whereas parts of the central Mediterranean and eastern Libyan margin display more heterogeneous deformation styles. These systematic spatial variations in seismic deformation provide a coherent kinematic framework for comparison with the stress tensor inversion results.
4.1.1 From the Azores triple junction to the Gulf of Cadiz and south of Iberia (from zone 1 to zone 6):
The westernmost part of the study area, extending from the Azores Triple Junction to the Gulf of Cadiz and southern Iberia, is characterised by a progressive eastward increase in seismic deformation rates accompanied by a gradual transition in deformation style. The calculated seismic displacement rates range from very low values near the Mid-Atlantic Ridge to substantially higher values in the eastern Atlantic domain (Figure 3b).
In the vicinity of the Azores Triple Junction and the Mid-Atlantic Ridge (Zones 1 and 2, Figure 3b), the seismic displacements are relatively low, reaching 0.39 and 0.15 mm/yr, respectively. The positive principal eigenvectors indicate predominantly extensional deformation, consistent with the divergent tectonic setting of the Mid-Atlantic Ridge. The NW-directed seismic displacement vectors reflect spreading-related deformation within the oceanic lithosphere.
Eastwards along the Azores–Gibraltar fault system, the deformation pattern progressively changes towards strike-slip-dominated deformation. In Zone 3, the calculated seismic displacement rate increases to 2.12 mm/yr. The principal horizontal deformation axes indicate the combined influence of extension and shear within the transition between the Mid-Atlantic Ridge and the oceanic transform system. Because this zone contains a relatively limited number of focal mechanisms, the inferred deformation regime is more sensitive to individual earthquakes, as shown by the jackknife analysis.
A marked increase in the seismic displacement rate is observed farther east along the Gloria Fault, where the estimated value reaches 20.40 mm/yr. This segment represents the principal oceanic transform boundary between the African and Eurasian plates in the eastern Atlantic and is characterised by predominantly dextral strike-slip deformation (). This value represents the seismic component of deformation released within the adopted deforming volume during the analysed catalogue period and should not be interpreted as the total relative plate motion.
A similarly elevated seismic displacement rate of 18.14 mm/yr is obtained for the Gulf of Cadiz (Zone 5; Figure 3b), which marks the transition between the oceanic transform domain of the Azores–Gibraltar fracture system and the compressional tectonic domain of the western Mediterranean. The elevated rates in Zones 4 and 5 reflect the concentration of cumulative seismic moment release within relatively localised deforming volumes. Their magnitudes are also sensitive to the adopted volume dimensions, catalogue duration, rigidity coefficient and occurrence of moderate-to-large earthquakes. They therefore represent localised seismic deformation rates rather than long-term Africa–Eurasia plate-convergence rates.
In Zone 6, south of Iberia, the deformation pattern changes towards a transpressional strike-slip regime, reflecting the transition from the oceanic Azores–Gibraltar system to the diffuse continental deformation associated with the Gibraltar Arc and the western Mediterranean.
Overall, these results reveal strong spatial variability in the seismic component of deformation between the Mid-Atlantic Ridge, the oceanic transform domain and the eastern Atlantic margin. Although direct quantitative agreement with geodetic plate velocities is not expected, the regional orientations of the seismic displacement vectors are broadly consistent with independent plate-kinematic and geodetic models.
4.1.2 From the Rif domain to the Pelegian block (from zone 7 to zone 11):
Within the Maghreb domain, corresponding to the continental segment of the Africa–Eurasia plate boundary in the western Mediterranean, the seismic displacement rates are generally low and range between 0.01 and 0.34 mm/yr. These values indicate that the seismic component of deformation released during the analysed catalogue period is limited in several sectors. They may also reflect long earthquake recurrence intervals, distributed crustal deformation, incomplete representation of low-magnitude seismicity, or non-seismic deformation processes. Evidence for aseismic deformation has previously been reported in the El Asnam region of Algeria by Lammali et al. (1997).
A major exception is observed in the central and western Tell Atlas of northern Algeria (Zone 8; Figure 3b), where the seismic displacement rate reaches 5.05 mm/yr, the highest value obtained within the Maghreb domain. This result indicates a strong concentration of seismic deformation along the northern Algerian margin, where active reverse and transpressional fault systems accommodate Africa–Eurasia convergence.
In addition to the marked spatial variability in seismic deformation rates, the resulting seismic displacement vectors exhibit a consistent NW- to WNW-directed orientation across most of the Maghreb domain (Figure 3b). This orientation reflects the first-order kinematic expression of Africa–Eurasia convergence in the western Mediterranean. It is also consistent with GNSS-derived velocity fields reported by Serpelloni et al. (2007), Vernant et al. (2010), , and . The agreement primarily concerns the regional direction of deformation rather than the absolute rate, because seismic moment tensor summation and geodetic techniques quantify different components of crustal motion over different temporal and spatial scales.
4.1.3 From the Tripolitania Basin to the Dead Sea transform system (from zone 12 to zone 17):
The eastern part of the study area is characterised by generally low seismic deformation rates, interrupted by localised zones of substantially higher seismic moment release (Figure 3b). The seismic displacement vectors exhibit a progressive change from a WNW-directed orientation in the Tripolitania Basin to a NNW-directed orientation in the Gulf of Aqaba, reflecting the increasing influence of Africa–Arabia relative motion towards the eastern end of the study area.
From the Cyrenaica Platform in northeastern Libya (Zone 13; Figure 3b) to the Gulf of Suez region in Egypt (Zone 15), the calculated seismic displacement rates remain below approximately 0.15 mm/yr. These values indicate that only a limited amount of seismic deformation was released during the analysed catalogue period. They may reflect comparatively low seismic activity, long earthquake recurrence intervals, or deformation distributed across broad continental domains.
A noticeable increase is observed in the Tripolitania Basin (Zone 12), where the seismic displacement rate reaches 0.73 mm/yr. The corresponding vector is directed towards the northwest (N298°), broadly consistent with the regional kinematic pattern of northern Africa relative to Eurasia.
Farther east, the seismic displacement rate increases markedly in the Gulf of Aqaba (Zone 17; Figure 3b), reaching 17.12 mm/yr. This is the highest value obtained in the eastern part of the study area. The elevated rate reflects concentrated seismic moment release along the southern Dead Sea Transform, where left-lateral strike-slip faulting accommodates Africa–Arabia relative motion. As in the Gloria Fault and Gulf of Cadiz zones, this value should be interpreted as a localised seismic displacement rate integrated over the catalogue period rather than as a direct estimate of long-term plate velocity.
4.2 Stress tensor inversion results
The stress field extending from the Azores Triple Junction to the Dead Sea Transform exhibits a first-order spatial gradient controlled by interactions among the African, Eurasian, North American and Arabian plates. The inverted principal stress orientations and the stress shape ratio (R) provide complementary constraints on the tectonic regimes of the 17 seismogenic zones. The results are summarised in Table 2, and the spatial distribution of the stress tensors is shown in Figure 4.
FIGURE 4
The tectonic regime was determined from the combined interpretation of the orientations of (1, 2 and 3), together with the stress shape ratio (R). The value of R describes the relative position of the intermediate principal stress between 1 and 3, but does not independently define whether the regime is extensional, strike-slip or compressional.
4.2.1 From the Azores triple junction to the Gulf of Cadiz and south of Iberia (from zone 1 to zone 6):
The Azores Triple Junction (Zone 1) displays a triaxial normal-faulting regime with an approximately vertical (1) axis and R = 0.40, consistent with its divergent tectonic setting. The Mid-Atlantic Ridge (Zone 2) is characterised by a transtensional to normal-faulting regime, also with a near-vertical (1). The high value of R = 0.86 indicates that the magnitude of 2 approaches that of 1, while the orientation of the principal axes confirms the extensional nature of the stress regime.
The Terceira Ridge (Zone 3) exhibits a triaxial normal-faulting regime with R = 0.57, consistent with active extension within the Azores domain. This stress regime differs from the strike-slip-dominated deformation style inferred from seismic moment summation. The jackknife analysis indicates that the latter result is comparatively sensitive to the limited number and distribution of earthquakes available for this zone.
Farther east, the Gloria Fault (Zone 4) is characterised by a predominantly strike-slip stress regime with a transtensional component. The maximum and minimum principal stress axes are sub-horizontal, whereas 2 is the most steeply plunging axis. The value R = 0.80 indicates that 2 is closer in magnitude to 1 than to 3; however, the strike-slip classification is based principally on the orientations of the stress axes rather than on R alone. This configuration is consistent with the role of the Gloria Fault as a major dextral oceanic transform structure.
Zone 5, extending from Gorringe Bank to the Gulf of Cadiz, displays a transpressional strike-slip regime with R = 0.20, indicating that 2 is relatively close in magnitude to 3. A comparable transpressional strike-slip regime is obtained south of Iberia in Zone 6, where R = 0.30. These zones mark the transition from oceanic transform faulting to diffuse convergence and strain partitioning in the Gibraltar Arc and western Mediterranean domains.
The stress field therefore evolves from extension at the Mid-Atlantic Ridge and within the Azores domain to strike-slip and transpressional regimes towards the Gloria Fault, the Gulf of Cadiz and southern Iberia. This west-to-east transition reflects the progressive change from oceanic spreading to transform deformation and oblique Africa–Eurasia convergence.
4.2.2 From the Rif domain to the Pelegian block (from zone 7 to zone 11):
The Rif Mountains and Alboran domain (Zone 7) are characterised by a triaxial strike-slip regime with R = 0.50. The intermediate value of R, combined with the stress-axis orientations, indicates a balanced strike-slip stress state consistent with complex strain partitioning and lateral deformation within the western Mediterranean.
Within the Tell Atlas, stress tensor inversion indicates predominantly compressional and transpressional regimes. The central and western Tell (Zone 8) are characterised by a strongly compressional stress state with R = 0.00, indicating that 2 and 3 have similar relative magnitudes. The eastern Tell (Zone 9) displays a transpressional regime with R = 0.20. In both zones, 1 is broadly horizontal and oriented NW–SE, whereas 3 is vertical or steeply plunging. This configuration is consistent with reverse to transpressional faulting associated with Africa–Eurasia convergence.
The focal mechanisms within Zones 8 and 9 (Figure 2) show a predominance of reverse faulting accompanied by strike-slip mechanisms. The relatively low R values indicate that the intermediate principal stress is closer in magnitude to the minimum principal stress, but the compressional classification is principally determined by the horizontal orientation of 1 and the near-vertical orientation of 3.
Farther east, across eastern Algeria, Tunisia and the Pelagian domain, the stress field becomes more heterogeneous. Strike-slip and locally transpressional regimes predominate, with intermediate R values and spatial rotations of the principal stress axes. These changes may reflect the influence of inherited lithospheric structures, block rotations and regional geodynamic processes associated with the central Mediterranean and the Calabrian Arc.
4.2.3 From the Tripolitania Basin to the Dead Sea transform system (from zone 12 to zone 17):
The stress inversion results for the Libya–Egypt sector reveal a more heterogeneous stress field than that observed across the Maghreb, reflecting the interaction among the African, Arabian and Eurasian plates.
In the Tripolitania Basin (Zone 12; Figure 4b), the stress tensor indicates an extensional to transtensional regime. The relatively high R value of approximately 0.80 indicates that 2 approaches 1 in magnitude, whereas the principal-axis orientations demonstrate that normal-faulting deformation is dominant.
In the Cyrenaica Platform (Zone 13; Figure 4b), the stress field changes to a strike-slip to transpressional regime. The principal stress axes rotate relative to those in the Tripolitania Basin, with 1 becoming predominantly horizontal. The lower R value of approximately 0.30 indicates that 2 is closer in magnitude to 3. This change reflects the transition between local extension in western Libya and more compressional or transpressional deformation towards the eastern Libyan margin.
Northern Egypt (Zone 14) exhibits a triaxial reverse-faulting regime with R = 0.50, characterised by a horizontal 1 and a near-vertical 3. The Gindi Basin region (Zone 15) displays a triaxial normal-faulting regime with R = 0.60, whereas the Red Sea Hills (Zone 16) are also characterised by normal faulting, with R = 0.46. These results identify a marked transition from compression in northern Egypt to extension towards the Gulf of Suez and Red Sea rift systems.
In the Gulf of Aqaba (Zone 17; Figure 4b), the stress inversion indicates a predominantly strike-slip regime with a transtensional component. The horizontal maximum principal stress axis is oriented approximately NNW–SSE, while the minimum principal stress axis is oriented approximately SW–NE. The value indicates that 2 is closer in magnitude to 1. Together with the principal-axis orientations, this stress configuration is consistent with left-lateral strike-slip deformation along the Dead Sea Transform.
Overall, the stress field evolves from extension in the western Atlantic spreading domain to strike-slip and transpressional regimes along the Azores–Gibraltar system, compression and transpression across the Maghreb, renewed extension in the Gulf of Suez and Red Sea domains, and strike-slip deformation along the Dead Sea Transform. This evolution is expressed through systematic changes in both the orientations of the principal stress axes and the stress shape ratio R. The results demonstrate that R provides useful information on the relative magnitude of the intermediate principal stress, while the tectonic regime must be determined from its combined interpretation with the three-dimensional orientations of 1, 2, and 3.
5 Discussion
The seismic moment tensor summation results reveal a clear west-to-east variation in seismic deformation rates along the Azores-Gibraltar fault system, accompanied by a progressive change in deformation style from extension in the Atlantic spreading domain to strike-slip and transpressional deformation farther east, and ultimately to compression across much of the Maghreb. This spatial evolution reflects the transition from an oceanic spreading and transform environment in the Atlantic to a broad and mechanically heterogeneous convergent plate boundary in the western Mediterranean (Figure 3b).
Before interpreting the regional patterns, it is important to emphasize that the seismic moment tensor summation method quantifies only the seismic component of crustal deformation released within a finite deforming volume during the analysed catalogue period. The resulting values should therefore not be interpreted as long-term plate-motion rates or as direct equivalents of GNSS-derived velocities. Geodetic observations measure present-day surface motion and include both seismic and aseismic contributions, whereas seismic moment tensor summation integrates the cumulative seismic moment released by earthquakes. The estimated values consequently represent zone-averaged seismic displacement rates whose magnitudes depend on the adopted deforming volume, catalogue duration, rigidity coefficient, earthquake recurrence, and temporal concentration of moderate-to-large events.
The seismic displacement vectors nevertheless define a coherent large-scale kinematic pattern along the plate boundary. In the Gulf of Aqaba, the vectors are directed towards the NNW (N342°), consistent with Arabian and African relative motion along the Dead Sea Transform system (El-Isa, 2013). Westward across northern Africa, the vectors progressively rotate towards NW and WNW directions, particularly within the Maghreb, reflecting the first-order kinematic expression of Africa–Eurasia convergence in the western Mediterranean. In the Atlantic domain, they become westward (N260°) to north-westward directed (N332°), consistent with the transition towards Mid-Atlantic Ridge spreading. This progressive rotation reflects the interaction among the African, Eurasian and Arabian plates.
The regional orientations of the seismic displacement vectors agree well with independent GNSS observations and plate-kinematic models (Serpelloni et al., 2007; Vernant et al., 2010; ; ). This agreement indicates that seismic moment tensor summation captures the first-order spatial pattern of regional deformation. However, the comparison is most meaningful in terms of vector orientation and tectonic segmentation rather than absolute rate, because the seismic and geodetic approaches sample different components of deformation over different spatial and temporal scales.
Across most segments of the northern African plate boundary, seismic displacement rates remain below 1 mm/yr. These low values indicate that only a limited amount of seismic deformation was released during the analysed catalogue period. They may reflect long earthquake recurrence intervals, distributed crustal deformation, incomplete representation of low-magnitude seismicity, or non-seismic deformation processes. Because the present study does not directly quantify aseismic deformation, the relative contributions of seismic and aseismic processes cannot be determined from these results alone.
The method also has intrinsic limitations. Seismic moment tensor summation estimates do not include aseismic creep, post-seismic deformation, distributed ductile strain, or deformation associated with earthquakes not represented in the focal mechanism catalogue. The Mw ≥ 4.0 threshold also favours moderate-to-large events and may underrepresent diffuse deformation released by smaller earthquakes. Localised seismic displacement rates may therefore exceed regional geodetic convergence rates without implying anomalously rapid plate motion, particularly where seismic moment release is concentrated within a restricted volume or dominated by a few large earthquakes.
Several zones exhibit comparatively high seismic displacement rates associated with major tectonic structures. The most prominent examples occur along the Gloria Fault and the Gulf of Cadiz, where the rates approach 20 mm/yr. This sector marks the transition between the oceanic transform domain of the Azores–Gibraltar fracture system and the compressional tectonics of the Gibraltar Arc. The concentration of seismic deformation is consistent with the occurrence of large historical earthquakes, including the 1755 Lisbon Mw 8.5 earthquake (Zitellini et al., 2001) and the 1975 Ms 7.9 earthquake (Lynnes and Ruff, 1985), which contribute substantially to the cumulative seismic moment released during the analysed period.
Similar behaviour was reported by for the Azores–Gibraltar plate boundary, where most zones yielded seismic slip velocities comparable in magnitude to plate-kinematic estimates, while one zone affected by several large earthquakes produced a substantially higher value. This comparison reinforces the interpretation that the elevated rates obtained for the Gloria Fault and Gulf of Cadiz reflect localised seismic moment release within finite deforming volumes rather than long-term Africa–Eurasia relative plate motion. Their magnitudes should therefore be evaluated in relation to catalogue duration, volume normalisation, rigidity coefficient, and the temporal distribution of major earthquakes.
Another important concentration of seismic deformation occurs along the Tell Atlas of northern Algeria, where the rate reaches 5.05 mm/yr. This region is among the most seismically active parts of the Africa–Eurasia plate boundary in North Africa and has generated several destructive earthquakes, including the 1954 Orléansville Mw 6.5 earthquake (Espinoza and Lopez-Arroyo, 1984), the 1980 El-Asnam Ms 7.3 earthquake (Ouyed et al., 1981), and the 2003 Zemmouri Mw 6.8 earthquake (). The elevated value reflects the concentration of seismic moment release along active reverse and transpressional fault systems that accommodate crustal shortening along the northern Algerian margin (e.g., Yelles-Chaouche et al., 2006). It should nevertheless be interpreted as a seismic displacement rate integrated over the catalogue period rather than as a direct measure of regional plate convergence.
Farther east, the Tripoli Basin displays a moderate seismic displacement rate of 0.73 mm/yr. Although lower than the values obtained for the Gulf of Cadiz and the Algerian Tell, it indicates a localised concentration of seismic deformation within northwestern Libya. This pattern may reflect the interaction between inherited structures of the northern African margin and the regional tectonic influence of the central Mediterranean.
The Gulf of Aqaba exhibits the highest seismic displacement rate in the eastern part of the study area, exceeding 17 mm/yr. This zone corresponds to the southernmost segment of the Dead Sea Transform, where the left-lateral strike-slip system accommodates Africa–Arabia relative motion. The high value indicates concentrated seismic moment release along this active transform during the investigated period. The 1995 (Mw) 7.2 Gulf of Aqaba earthquake (Hofstetter, 2003) illustrates the significant seismic potential of this fault system. As in the Gloria Fault and Gulf of Cadiz sectors, the estimated value should not be interpreted as a direct long-term plate velocity.
A major strength of the study is the broad consistency between the deformation regimes inferred from seismic moment tensor summation and the stress regimes obtained from focal mechanisms inversion. The stress field evolves from extension in the Atlantic spreading domain to strike-slip and transpressional conditions along the Azores–Gibraltar system, compression and transpression across the Maghreb, renewed extension towards the Gulf of Suez and Red Sea domains, and strike-slip deformation along the Dead Sea Transform.
The tectonic regime is determined from the combined interpretation of the orientations of σ1, σ2, and σ3, together with the stress shape ratio R. The ratio (R) constrains the relative magnitude of the intermediate principal stress but does not independently define whether the stress regime is extensional, strike-slip, or compressional. In the Atlantic domain, the extensional deformation inferred for Zones 1 and 2 is supported by normal to transtensional stress regimes, consistent with active oceanic spreading. Along the Gloria Fault and the Gulf of Aqaba, both methods indicate predominantly strike-slip deformation, supporting their interpretation as major transform boundaries.
Across the Maghreb, compressional and transpressional stress regimes inferred from focal-mechanism inversion broadly agree with the shortening-dominated deformation obtained from eigenvalue analysis. The maximum principal stress is generally horizontal and oriented approximately NW–SE, consistent with Africa–Eurasia convergence. This stress configuration explains the prevalence of reverse and strike-slip focal mechanisms observed along the Tell Atlas and confirms the role of the northern African margin as a major domain of strain accommodation.
Farther east, the stress field becomes more heterogeneous across Libya and Egypt, reflecting the transition from Africa–Eurasia convergence to the increasing influence of Africa–Arabia relative motion and Red Sea rifting. In the Gulf of Aqaba, the stress tensor indicates a predominantly strike-slip regime, consistent with the kinematics of the Dead Sea Transform. The agreement between the kinematic and mechanical analyses supports the proposed tectonic segmentation and indicates that the seismic deformation accumulated during the catalogue period is broadly consistent with the present-day regional stress field.
The tectonic regimes inferred from the two methods are coherent in 16 of the 17 investigated zones. The main discrepancy occurs in Zone 3, corresponding to the Terceira Ridge. Seismic moment tensor summation suggests a strike-slip-dominated regime, with shortening and extensional rates of 1.86 and 1.02 mm/yr, respectively, whereas stress tensor inversion indicates a triaxial normal-faulting regime.
The jackknife analysis shows that the deformation tensor in Zone 3 is particularly sensitive to the limited number and distribution of available earthquakes, with one or a few focal mechanisms exerting a disproportionate influence on the inferred deformation pattern. This interpretation is consistent with the statistical sensitivity of the seismic moment tensor summation method when applied to sparse earthquake catalogues, for which individual events may disproportionately influence the resulting strain-rate tensor and the inferred tectonic regime. Seismic moment summation is also sensitive to the temporal distribution of seismic moment release and to the possible underrepresentation of particular mechanism types during the catalogue period. Stress tensor inversion is likewise dependent on the quality and representativeness of the focal-mechanism dataset, but in this case the extensional stress regime is more consistent with the established geodynamic setting of the Terceira Ridge as an active spreading structure within the Azores domain.
The comparison also shows that zones with low seismic displacement rates may nevertheless display well-constrained stress regimes (Table 2). This is particularly evident in part of Libya and the central Maghreb, where seismic displacement rates remain below 1 mm/yr, but stress inversion identifies coherent transpressional, strike-slip, or extensional regimes. This distinction reflects the fact that stress orientation may remain regionally organised even where seismic moment release is limited during the observation period. Such zones may be characterised by long recurrence intervals, low seismic coupling, distributed deformation, or incomplete sampling of their seismic cycle.
Taken together, these results demonstrate that seismic moment tensor summation and stress tensor inversion provide complementary descriptions of active deformation along the Africa–Eurasia–Arabia plate-boundary system (Figure 5). The former quantifies the spatial distribution and orientation of the seismic component of deformation accumulated during the catalogue period, whereas the latter constrains the orientation and relative magnitudes of the regional stress field. Their combined application reveals a coherent tectonic framework extending from the Mid-Atlantic Ridge to the Gulf of Aqaba, despite pronounced variations in seismic activity, deforming-volume geometry, earthquake recurrence, and lithospheric structure.
FIGURE 5
This integrated framework provides robust constraints on plate-boundary segmentation, seismic strain localisation, and regional stress organisation. It also offers a useful basis for future investigations of tectonic deformation and seismic hazard, provided that the derived seismic displacement rates are interpreted within the spatial, temporal, and mechanical assumptions of the seismic moment tensor summation method.
6 Conclusion
This study presents a comprehensive assessment of the present-day seismic deformation and stress field along the northern boundary of the African plate, extending from the Azores Triple Junction to the Gulf of Aqaba, through the combined application of seismic moment tensor summation and stress tensor inversion. The integration of these complementary approaches provides new quantitative constraints on the spatial distribution of the seismic component of crustal deformation, the regional stress field, and the kinematic segmentation of the Africa–Eurasia and Africa–Arabia plate boundaries.
The results reveal pronounced spatial variability in both the seismic deformation rate and the style of deformation along the plate boundary. The orientation of the seismic displacement vectors indicates a progressive rotation from NNW in the Gulf of Aqaba to NW across the Maghreb region and westward in the Atlantic domain, consistent with the first-order kinematic framework of the Africa–Arabia–Eurasia plate boundary. The close agreement between the orientation of the seismic displacement vectors and independent GNSS-derived velocity fields demonstrates that seismic moment tensor summation successfully captures the first-order regional kinematic framework expressed by the seismic component of crustal deformation.
Overall, the seismic deformation rates remain relatively low across most segments of the northern African plate boundary, generally below 1 mm/yr, indicating that only a limited amount of seismic deformation was released during the analysed catalogue period. These low values may reflect long earthquake recurrence intervals, distributed crustal deformation, incomplete sampling of seismicity, or non-seismic deformation processes. Because the present study does not directly quantify aseismic deformation, the relative contributions of seismic and aseismic processes cannot be determined from the present analysis alone. Nevertheless, several localised tectonic domains exhibit substantially higher seismic displacement rates, reflecting concentrated seismic moment release associated with major active fault systems rather than anomalously rapid plate motion. Indeed, high deformation rates are observed in the Gulf of Cadiz region, where the interaction between the Azores–Gibraltar fracture system and the western Mediterranean convergent domain leads to significant strain concentration. Another major zone of deformation is identified in the Tell Atlas of northern Algeria, where active compressional structures accommodate the convergence between Africa and Eurasia and have generated several destructive earthquakes. Elevated deformation rates are also observed in the Tripoli Basin and particularly in the Gulf of Aqaba, where the Dead Sea Transform accommodates the relative motion between the African and Arabian plates. These localised high values should therefore be interpreted within the framework of seismic moment tensor summation, whose estimates depend on the adopted deforming volume, catalogue duration, rigidity coefficient, and temporal distribution of seismic moment release. Accordingly, these localised seismic displacement rates should be interpreted as estimates of cumulative seismic deformation rather than direct measurements of long-term relative plate motion.
The stress tensor inversion results further confirm the strong segmentation of the plate boundary. An extensional stress regime dominates in the Atlantic domain near the Mid-Atlantic Ridge, evolving toward transtensional conditions along the Azores–Gibraltar fracture zone and strike-slip regimes along major transform faults such as the Gloria Fault and the Dead Sea Transform. In contrast, the Maghreb region is characterised by compressional to transpressional stress regimes, consistent with the Africa–Eurasia convergence. The overall agreement between the deformation regimes inferred from seismic moment tensor summation and those derived from stress tensor inversion confirms the mechanical consistency of the proposed tectonic segmentation across nearly all investigated seismogenic zones.
The jackknife sensitivity analysis further demonstrates the robustness of the estimated deformation tensors for most seismogenic zones while highlighting the greater uncertainty associated with regions characterised by sparse earthquake catalogues. In particular, the jackknife analysis identifies Zone 3 (Terceira Ridge) as the only seismogenic zone where the inferred deformation regime remains sensitive to the limited number, temporal distribution, and relative contribution of the available focal mechanisms. These results emphasise the importance of earthquake catalogue completeness and event distribution when interpreting seismic strain rates derived from cumulative seismic moment release.
Beyond the regional tectonic implications, this study also highlights both the strengths and the limitations of the seismic moment tensor summation approach. The method provides a robust and internally consistent framework for quantifying the spatial distribution of the seismic component of crustal deformation and identifying the principal domains of tectonic strain localisation along complex plate boundaries. The close agreement between the deformation regimes inferred from seismic moment tensor summation, stress tensor inversion, and the orientation of independent GNSS velocity fields demonstrates its capability to reproduce the first-order regional pattern of seismic deformation and tectonic segmentation observed independently from geodetic and stress-field analyses. Nevertheless, the derived seismic displacement rates should not be interpreted as direct measures of long-term plate motion. Instead, they represent zone-averaged seismic deformation rates integrated over finite deforming volumes and the observation period, and are therefore influenced by earthquake catalogue completeness, the temporal distribution of seismic moment release, the adopted deforming volume, and the rigidity coefficient. Consequently, the principal strength of the seismic moment tensor summation approach lies not in reproducing absolute plate velocities but in providing a physically consistent quantification of the seismic component of crustal deformation and its spatial variability, thereby allowing identification of the principal tectonic domains and strain localisation along complex plate boundaries.
Taken together, these results demonstrate that seismic moment tensor summation and stress tensor inversion provide complementary and internally consistent descriptions of the present-day tectonic behaviour of the northern African plate boundary. While the former quantifies the spatial distribution of the seismic component of deformation accumulated during the observation period, the latter constrains the orientation and relative magnitude of the regional stress field. Their combined application reveals a coherent tectonic framework extending from the Mid-Atlantic Ridge to the Gulf of Aqaba, despite the pronounced spatial variability in seismic activity, earthquake recurrence, and lithospheric structure. Beyond improving our understanding of the present-day geodynamics of the Africa–Eurasia and Africa–Arabia plate boundaries, this integrated methodology provides a robust framework for future investigations of plate-boundary segmentation, crustal deformation, seismic strain localisation, and regional seismic hazard. The methodology can also be readily applied to other diffuse and mechanically heterogeneous plate-boundary systems worldwide.
The combination of seismic moment tensor summation and stress tensor inversion therefore represents a powerful and complementary approach for investigating the relationships among seismic deformation, stress, and plate-boundary kinematics in complex active tectonic regions.
Statements
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.
Author contributions
AI: Data curation, Conceptualization, Validation, Methodology, Writing – review and editing, Writing – original draft, Investigation, Formal Analysis, Software, Visualization. FO: Investigation, Methodology, Writing – review and editing, Data curation, Validation, Formal Analysis, Visualization, Conceptualization. SA: Visualization, Validation, Methodology, Formal Analysis, Writing – review and editing, Investigation, Software. MB: Project administration, Writing – review and editing, Methodology, Visualization, Formal Analysis, Supervision, Validation, Investigation, Funding acquisition, Conceptualization, Resources. AA: Investigation, Writing – review and editing, Visualization, Formal Analysis, Validation, Supervision, Conceptualization, Methodology.
Funding
The author(s) declared that financial support was not received for this work and/or its publication.
Acknowledgments
The authors sincerely thank the two reviewers for their careful assessment of the manuscript and for their valuable, constructive, and insightful comments. Their suggestions have greatly contributed to improving the clarity of the presentation, strengthening the methodological discussion, and refining the interpretation of the results. The authors are grateful for the time and expertise devoted to the review process, which have significantly enhanced the quality of this work.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The author MB declared that they were an editorial board member of Frontiers at the time of submission. This had no impact on the peer review process and the final decision.
Generative AI statement
The author(s) declared that generative AI was used in the creation of this manuscript. The authors used ChatGPT (OpenAI) solely to improve the manuscript’s language and readability. No scientific content, results, interpretations, conclusions, or references were generated using artificial intelligence. The authors reviewed and validated the final text and take full responsibility for its content.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/feart.2026.1875255/full#supplementary-material
References
1
AkiK.RichardsP. G. (1980). Quantitative Seismology: Theory and Methods, 859. San Francisco: Freeman, 296–297.
2
AndersonE. M. (1951). The Dynamics of Faulting and Dyke Formation with Applications to Britain. 2nd ed.Edinburgh: Oliver & Boyd.
3
AngelierJ. (1990). Inversion of field data in fault tectonics to obtain the regional stress—I. Single-phase fault populations: a new method of computing the stress tensor. Geophys. J. Int.103, 363–376.
4
ArsenikosS.Frizon de LamotteD.Chamot-RookeN.MohnG.BonneauM. C.BlanpiedC. (2013). Mechanism and timing of tectonic inversion in Cyrenaica (Libya): integration in the geodynamics of the East Mediterranean. Tectonophysics608, 319–329. 10.1016/j.tecto.2013.09.025
5
AyadiA.DorbathC.OusadouF.MaoucheS.ChikhM.BounifM. A.et al (2008). Zemmouri earthquake rupture zone (Mw 6.8, Algeria): aftershock sequence relocation and 3-D velocity model. J. Geophys. Res. Solid Earth113, B09301. 10.1029/2007JB005257
6
BahrouniN.BouazizS.SoumayaA.Ben AyedN.AttafiK.El GhaliA.et al (2014). Neotectonic and seismotectonic investigation of seismically active regions in Tunisia: a multidisciplinary approach. J. Seismol.18, 235–256. 10.1007/s10950-013-9395-y
7
BahrouniN.MassonF.MeghraouiM.SalehM.MaamriR.DhahaF.et al (2020). Active tectonics and GPS data analysis of the Maghrebian thrust belt and Africa–Eurasia plate convergence in Tunisia. Tectonophysics785, 228440. 10.1016/j.tecto.2020.228440
8
BatistaL.HübscherC.TerrinhaP.MatiasL.AfilhadoA.LüdmannT. (2017). Crustal structure of the Eurasia–Africa plate boundary across the Gloria Fault, North Atlantic Ocean. Geophys. J. Int.209 (2), 713–729. 10.1093/gji/ggx050
9
BenouarD.LaradiN. (1996). A reappraisal of the seismicity of the Maghreb countries—Algeria, Morocco, Tunisia. Nat. Hazards13, 275–296. 10.1007/BF00215819
10
BernardiF.BraunmillerJ.KradolferU.GiardiniD. (2004). Automatic regional moment tensor inversion in the European-Mediterranean region. Geophys. J. Int.157 (2), 703–716. 10.1111/j.1365-246X.2004.02215.x
11
BezzeghoudM.AdamC.BufornE.BorgesJ. F.CaldeiraB. (2014). Seismicity along the Azores-Gibraltar region and global plate kinematics. J. Seismol.18, 205–220. 10.1007/s10950-013-9416-x
12
BilliA.CuffaroM.OrecchioB.PalanoM.PrestiD.TotaroC. (2023). Retracing the Africa–Eurasia nascent convergent boundary in the western Mediterranean based on earthquake and GNSS data. Earth Planet. Sci. Lett.601, 117906. 10.1016/j.epsl.2022.117906
13
BosworthW.El-HawatA. S.HelgesonD. E.BurkeK. (2008). Cyrenaican shock absorber and associated inversion strain shadow in the collision zone of Northeast Africa. Geology36, 695–698. 10.1130/G24909A.1
14
BougrineA.Yelles-ChaoucheA. K.CalaisE. (2019). Active deformation in Algeria from continuous GPS measurements. Geophys. J. Int.217, 572–588. 10.1093/gji/ggz035
15
BrocherT. M. (2005). Empirical relations between elastic wavespeeds and density in the Earth's crust. Bull. Seismological Soc. Am.95 (6), 2081–2092. 10.1785/0120050077
16
BufornE.UdíasA.ColombásM. A. (1988). Seismicity, source mechanisms and tectonics of the Azores–Gibraltar plate boundary. Tectonophysics152 (1–2), 89–118. 10.1016/0040-1951(88)90031-5
17
BufornE.BezzeghoudM.UdíasA.ProC. (2004). Seismic sources on the Iberia–African plate boundary and their tectonic implications. Pure Appl. Geophys.161 (3), 623–646. 10.1007/s00024-003-2466-1
18
DeMetsC.GordonR. G.ArgusD. F.SteinS. (1990). Current plate motions. Geophys J. Int.101, 425–478. 10.1111/j.1365-246X.1990.tb06579.x
19
DeMetsC.GordonR. G.ArgusD. F.SteinS. (1994). Effect of recent revisions to the geomagnetic reversal time scale on estimates of current plate motions. Geophys Res. Lett.21, 2191–2194. 10.1029/94GL02118
20
DoglioniC.MerliniS.CantarellaG. (1999). Foredeep geometries at the front of the Apennines in the Ionian Sea (central Mediterranean). Earth Planet. Sci. Lett.168, 243–254. 10.1016/S0012-821X(99)00059-X
21
EkströmG.EnglandP. (1989). Seismic strain rates in regions of distributed continental deformation. J. Geophys. Res. Solid Earth94 (B8), 10231–10257. 10.1029/JB094iB08p10231
22
EkströmG.NettlesM.DziewońskiA. M. (2012). The global CMT project 2004–2010: centroid-moment tensors for 13,017 earthquakes. Phys. Earth Planet. Interiors200, 1–9. 10.1016/j.pepi.2012.04.002
23
El-IsaZ. H. (2013). Seismicity and seismotectonics of the Gulf of Aqaba region. Arab. J. Geosci.6, 3437–3449. 10.1007/s12517-012-0604-8
24
EspinozaA. F.Lopez-ArroyoA. (1984). Inconsistencies in the modified Mercalli intensity scales, El Asnam, Algeria, Earthquake. Proc. Conf. Int. Sur la macrozonation sismique, Ech Cheliff, Algérie10-12, 51–65.
25
Frizon de LamotteD.Saint-BezarB.BraceneR.MercierE. (2000). The two main steps of the Atlas building and geodynamics of the western Mediterranean. Tectonics19, 740–761. 10.1029/2000TC900003
26
GrandinR.BorgesJ. F.BezzeghoudM.CaldeiraB.CarrilhoF. (2007a). Simulations of strong ground motion in SW Iberia for the 1969 February 28 (Ms = 8.0) and the 1755 November 1 (M∼8.5) earthquakes—I. Velocity model. Geophys. J. Int.171 (3), 1144–1161. 10.1111/j.1365-246X.2007.03570.x
27
GrandinR.BorgesJ. F.BezzeghoudM.CaldeiraB.CarrilhoF. (2007b). Simulations of strong ground motion in SW Iberia for the 1969 February 28 (Ms = 8.0) and the 1755 November 1 (M∼8.5) earthquakes—II. Strong ground motion simulations. Geophys. J. Int.171 (2), 807–822. 10.1111/j.1365-246X.2007.03571.x
28
HalletD. (2002). Petroleum Geology of Libya. Amsterdam: Elsevier.
29
HardebeckJ. L.MichaelA. J. (2004). Stress orientations at intermediate angles to the San Andreas Fault, California. J. Geophys. Res.: Solid Earth109 (B11), B11303. 10.1029/2004JB003239
30
HardebeckJ. L.MichaelA. J. (2006). Damped regional‐scale stress inversions: methodology and examples for southern California and the Coalinga aftershock sequence. J. Geophys. Res. Solid Earth111 (B11). 10.1029/2005JB004144
31
HeidbachO.ReineckerJ.TingayM.MüllerB.SpernerB.FuchsK.et al (2007). Plate boundary forces are not enough: second- and third-order stress patterns highlighted in the World Stress Map database. Tect.26 (6), TC6014. 10.1029/2007TC002133
32
HeidbachO.TingayM.BarthA.ReineckerJ.KurfeßD.MüllerB. (2010). Global crustal stress pattern based on the World Stress Map database release 2008. Tect.482, 3–15. 10.1016/j.tecto.2009.07.023
33
HeidbachO.RajabiM.CuiX.FuchsK.MüllerB.ReineckerJ.et al (2018). The world stress map database release 2016: crustal stress pattern across scales. Tectonophysics744, 484–498. 10.1016/j.tecto.2018.07.007
34
HofstetterA. (2003). Seismic observations of the 22/11/1995 Gulf of Aqaba earthquake sequence. Tectonophysics369 (1-2), 21–36. 10.1016/S0040-1951(03)00129-X
35
HusseinH. M.Abou EleneanK. M.MarzoukI. A.KorratI. M.El-NaderI. A.GhazalaH.et al (2013). Present-day tectonic stress regime in Egypt and surrounding areas based on inversion of earthquake focal mechanisms. J. Afr. Earth Sci.81, 1–15. 10.1016/j.jafrearsci.2012.12.002
36
JacksonJ.McKenzieD. (1988). The relationship between plate motions and seismic moment tensors, and the rates of active deformation in the Mediterranean and Middle East. Geophys. J. Int.93, 45–73. 10.1111/j.1365-246X.1988.tb01387.x
37
KiratziA. A.PapazachosC. B. (1995). Active crustal deformation from the Azores Triple Junction to the Middle East. Tectonophysics243, 1–24. 10.1016/0040-1951(94)00188-F
38
KostrovV. V. (1974). Seismic moment and energy of earthquakes and seismic flow of rocks. Izvestiya, Acad. Sci. USSR, Phys. Solid Earth1, 23–40.
39
LammaliK.BezzeghoudM.OussadouF.DimitrovD.BenhallouH. (1997). Postseismic deformation at El Asnam (Algeria) in the seismotectonic context of northwestern Algeria, Geophys. J. Int.129, 3, 597–612. 10.1111/j.1365-246X.1997.tb04496.x
40
LynnesC. S.RuffL. J. (1985). Source process and tectonic implications of the great 1975 North Atlantic earthquake. Geophys. J. Int.82 (3), 497–510. 10.1111/j.1365-246X.1985.tb05148.x
41
Madarieta-TxurrukaA.PrietoJ. F.EscayoJ.PietrolungoF.PeláezJ. A.Galindo-ZaldívarJ.et al (2025). New insights on active geodynamics of Iberia and Northwestern Africa from seismic stress and geodetic strain-rate fields. Gondwana Res.149, 314–336. 10.1016/j.gr.2025.08.020
42
MarquesF. O.BalázsA.GeryaT. V.HildenbrandA. (2025). Dynamics and evolution of the Azores Triple Junction and its relation to pre-existing major tectonic structures. Commun. Earth & Environ.6 (1), 303. 10.1038/s43247-025-02255-8
43
McCluskyS.ReilingerR.MahmoudS.Ben SariD.TealebA. (2003). GPS constraints on Africa (Nubia) and Arabia plate motions. Geophys. J. Int.155, 126–138. 10.1046/j.1365-246X.2003.02023.x
44
MeghraouiM.PondrelliS. (2012). Active faulting and transpression tectonics along the plate boundary in North Africa. Ann. Geophys.55 (5). 10.4401/4970
45
MichaelA. J. (1984). Determination of stress from slip data: faults and folds. J. Geophys. Res. Solid Earth89 (B13), 11517–11526. 10.1029/JB089iB13p11517
46
MichaelA. J. (1987). Use of focal mechanisms to determine stress: a control study. J. Geophys. Res. Solid Earth92 (B1), 357–368. 10.1029/JB092iB01p00357
47
MooneyW. D.RomanowiczB.DziewonskiA. (2010). 1.11 crust and lithospheric structure–global crustal structure. Seismol. Struct. Earth Treatise Geophys.1, 361. 10.1016/B978-044452748-6.00011-0
48
MoustafaS. S.AbdalzaherM. S.AbdelhafiezH. E. (2022). Seismo-lineaments in Egypt: analysis and implications for active tectonic structures and earthquake magnitudes. Remote Sens.14, 6151. 10.3390/rs14236151
49
NocquetJ. M. (2012). Present-day kinematics of the Mediterranean: a comprehensive overview of GPS results. Tectonophysics579, 220–242. 10.1016/j.tecto.2012.03.037
50
NocquetJ. M.CalaisE. (2004). Geodetic measurements of crustal deformation in the western Mediterranean and Europe. Pure Appl. Geophys.161, 661–681. 10.1007/s00024-003-2468-z
51
OusadouF.DorbathL.AyadiA.DorbathC.GharbiS. (2014). Stress field variations along the Maghreb region derived from inversion of major seismic crisis fault plane solutions. Tectonophysics632, 261–280. 10.1016/j.tecto.2014.06.017
52
OusadouF.AyadiA.BezzeghoudM. (2024). Catalogue of source mechanisms and overview of present-day stress fields in the western region of the Africa–Eurasia plate boundary. Front. Earth Sci.12, 1366156. 10.3389/feart.2024.1366156
53
OuyedM.MeghraouiM.CisternasA.DeschampsA.DorelJ.FréchetJ.et al (1981). Seismotectonics of the El Asnam earthquake. Nature292, 26–31. 10.1038/292026a0
54
PerrodonA. (1957). Étude géologique des bassins sublittoraux de l’Algérie occidentale. Publ. Serv. Carte Géol Algérie, NS Bull.12, 328.
55
PondrelliS.SalimbeniS.MorelliA.EkströmG.PostpischlL.VannucciG.et al (2011). European–Mediterranean regional centroid moment tensor catalog: solutions for 2005–2008. Phys. Earth Planet. Interiors185, 74–81. 10.1016/j.pepi.2011.01.007
56
ReilingerR.McCluskyS.VernantP.LawrenceS.ErgintavS.CakmakR.et al (2006). GPS constraints on continental deformation in the Africa–Arabia–Eurasia collision zone and implications for the dynamics of plate interactions. J. Geophys. Res. Solid Earth111 (B5), B05411. 10.1029/2005JB004051
57
Rodríguez CollantesD.Blanco HernándezA.de Lacy Pérez de los CobosM. C.Galindo-ZaldivarJ.GilA. J.Sánchez PiedraM. Á.et al (2025). A new Earth crustal velocity field estimation from ROA cGNSS station networks in the south of Spain and North Africa. Remote Sens.17 (4), 704. 10.3390/rs17040704
58
SalamonA.HofstetterA.GarfunkelZ.RonH. (1996). Seismicity of the eastern Mediterranean region: perspective from the Sinai subplate. Tectonophysics263, 293–305. 10.1016/S0040-1951(96)00030-3
59
SalemM. J.El-ArnautiA.El-Sogher SalehA. (2008). The geology of East Libya (Vol. 3). Earth Science Society of Libya.
60
SellaG. F.DixonT. H.MaoA. (2002). REVEL: a model for recent plate velocities from space geodesy. J. Geophys Res. Solid Earth107, B4. 10.1029/2000JB000033
61
SerpelloniE.VannucciG.PondrelliS.ArgnaniA.CasulaG.AnzideiM.et al (2007). Kinematics of the western Africa–Eurasia plate boundary from focal mechanisms and GPS data. Geophys. J. Int.169, 1180–1200. 10.1111/j.1365-246X.2007.03367.x
62
SïdiA.ChardonD.BabyP.OualiJ. (2011). Active oblique ramp faulting in the southern Tunisian Atlas. Tectonophysics (1), 178–189. 10.1016/j.tecto.2011.01.010
63
SoumayaA.Ben AyedN.DelvauxD.GhanmiM. (2015). Spatial variation of present-day stress field and tectonic regime in Tunisia and surroundings from formal inversion of focal mechanisms: geodynamic implications for the central Mediterranean. Tectonics34, 1154–1180. 10.1002/2015TC003895
64
StichD.MancillaF.BaumontD.MoralesJ. (2005). Source analysis of the Mw 6.3 2004 Al Hoceima earthquake (Morocco) using regional apparent source time functions. J. Geophys. Res. Solid Earth110, B06306. 10.1029/2004JB003366
65
StorchakD. A.Di GiacomoD.BondárI.EngdahlE. R.HarrisJ.LeeW. H.et al (2013). Public release of the ISC–GEM global instrumental earthquake catalogue (1900–2009). Seismol. Res. Lett.84 (5), 810–815. 10.1785/0220130034
66
UdíasA.BufornE. (1991). Regional stresses along the Eurasia–Africa plate boundary derived from focal mechanisms of large earthquakes. Pure Appl. Geophys.136, 433–448. 10.1007/BF00878580
67
Van der WoerdJ.DorbathC.OusadouF.DorbathL.DelouisB.JacquesE.et al (2014). The Al Hoceima (M_w) 6.4 earthquake of 24 February 2004 and its aftershocks sequence. J. Geodyn.77, 89–109. 10.1016/j.jog.2013.12.004
68
VernantP.FadilA.MourabitT.OuazarD.KoulaliA.DavilaJ. M.et al (2010). Geodetic constraints on active tectonics of the western Mediterranean: implications for the kinematics and dynamics of the Nubia–Eurasia plate boundary zone. J. Geodyn.49, 123–129. 10.1016/j.jog.2009.10.007
69
ViltresR.DoubreC.DoinM. P.MassonF. (2025). Aseismic creep and strain partitioning accommodating the Nubia-Eurasia oblique convergence in northern Africa from InSAR analysis. Geology53 (4), 349–354. 10.1130/G53117.1
70
WeißB. J.HübscherC.LüdmannT. (2015). The tectonic evolution of the southeastern Terceira Rift/São Miguel region (Azores). Tectonophysics654, 75–95. 10.1016/j.tecto.2015.04.018
71
Yelles-ChaoucheA.BoudiafA.DjellitH.BraceneR. (2006). La tectonique active de la région nord-algérienne. Comptes Rendus Geosci.338, 126–139. 10.1016/j.crte.2005.11.002
72
ZitelliniN.MendesL. A.CordobaD.DanobeitiaJ.NicolichR.PellisG.et al (2001). Source of 1755 Lisbon earthquake and tsunami investigated. Eos, Trans. Am. Geophys. Union82 (26), 285–291. 10.1029/EO082i026p00285-01
Summary
Keywords
Africa–Eurasia plate boundary, focal mechanism inversion, moment tensor summation, seismic deformation, stress tensor
Citation
Issaadi A, Ousadou F, Aguemoune S, Bezzeghoud M and Ayadi A (2026) Spatial variation of the present-day strain-rate and stress regime, from mid-Atlantic triple junction through north Africa to Aqaba transform zone. Front. Earth Sci. 14:1875255. doi: 10.3389/feart.2026.1875255
Received
07 May 2026
Revised
25 July 2026
Accepted
29 July 2026
Published
17 September 2026
Volume
14 - 2026
Edited by
Nicola Alessandro Pino, University of Camerino, Italy
Reviewed by
Angelo De Santis, National Institute of Geophysics and Volcanology (INGV), Italy
Davide Zaccagnino, Southern University of Science and Technology, China
Updates
Copyright
© 2026 Issaadi, Ousadou, Aguemoune, Bezzeghoud and Ayadi.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Mourad Bezzeghoud, mourad@uevora.pt
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.