| 18 Jun 2026
Insights into stress-loading mechanisms in the western Peloponnese, Greece from a high-resolution earthquake catalog
Abstract. The western Peloponnese exhibits a continuum of earthquake faulting styles that reside in a complex stress field near the western termination of the Hellenic Subduction System. Here we present a detailed study of recent seismicity in western Peloponnese, including the spatiotemporal distribution of earthquake sequences and their clustering and migration properties to infer possible driving mechanisms. We build a detailed earthquake catalog from December 2023 – September 2024 with a magnitude-of-completeness Mc ~1 and a location precision on the order of 100 s of meters by combining new data from temporary seismic station deployments with publicly available data from permanent stations. Catalog statistical and clustering analysis shows increased background seismicity rates and seismic moment release in the northern part of the study area that is consistent with larger strain rates and a higher stressing rate reported from geodetic data. The seismicity distribution and focal mechanism solutions suggest that the predominantly north-south extensional regime in the north changes to east-west extension in the central-western Peloponnese near the town of Zaharo. Nearest-neighbor cluster analysis reveals mainshock-aftershock-type sequences in the northwest near the town of Vartholomio that are consistent with tectonic loading. Conversely, clustering properties in the central-western Peloponnese near Zaharo are consistent with swarm-like sequences driven by external forcing, such as pore-fluid pressure changes, and potential aseismic slip. Independent studies of slab dewatering suggest fluid-driven pore-pressure gradients that might be responsible for migrating seismicity and swarm-like behavior in the upper plate.
Review for the Manuscript by Essing et al. ‘Insights into stress-loading mechanisms in the western Peloponnese, Greece from a high-resolution earthquake catalog’.
Essing et al. in their study build an enhanced earthquake catalogue of about 15000 events, mostly microearthquakes, that occurred over a time-period of 10 months (in 2023-2024) in western Peloponnese, and explore the emerging spatiotemporal patterns. Their focus is on three seismicity clusters, the analysis of which suggests two independent ‘loading’ mechanisms on the faults involved: a direct tectonic loading with a typical mainshock-aftershock type sequence and one that relates to fluid circulation and likely aseismic slip. They also compare the emerging seismicity patterns to deformation patterns that derive from existing long-term (geomorphological) and short-term (geodetic, other seismological) datasets and discuss similarities and differences. By doing so, they derive generic conclusions about subduction processes in the forearc of the Hellenic subduction system.
The manuscript is of regional interest, well-structured and the analysis is done carefully. I find the identified seismicity clusters and the two diverse loading mechanisms proposed, convincing and well argued. The relation of the clusters to the overall kinematics of Peloponnese appears, however, to be incomplete and less convincing – mainly because the authors rely on out-dated datasets and kinematic information for their analysis. Thus, the interpretations and the generic conclusions drawn from this analysis require some revisions (see below).
Main comments:
My main concern comes from the attempt by the authors to assess (and amalgamate) their results without using an up-to-date understanding of the kinematics of the region, which is overall well studied. As a result, they extract conclusions without the support of currently available datasets (including the authors’ own data). The active faults, for example, included in figures 1-2 & 5-6 and which are used by the authors, are sparse and outdated (most of known active faults are either missing or overwhelmingly simplified). This becomes a problem when the study tries to link these seismicity clusters to likely candidate structures that generated them and, subsequently, use this comparison to extract generic conclusions on the regional kinematics. Similarly, the study would benefit from the use of the most up to date geodesy maps and information (including evidence for aseismic slip nearby) to compare their results against but also from offshore fault/earthquake information (see references below).
These deficiencies concerns the ‘interpretation’ part of the manuscript, not the analytical, and can be fixed relatively easily by 1) updating all relevant fault maps to incorporate the existing upper-plate active and likely active faults in Greece as likely sources for the recorded swarms; and 2) discuss the swarm characteristics with respect to the faults’ kinematics (but also the state-of-the-art kinematics of the broader region). The most inclusive database of active faults in Greece, with each fault trace being represented on the landscape on a 1:25,000 scale, is the AFG by Begg et al. (2025) while the state-of-the-art geodesy analysis is presented by Chousianitis et al. (2024).
Below, I explain why these deficiencies have a prime effect on the interpretative part of the study.
From a quick scan of the AFG database (Begg et al. 2025), and contrary to the stated opinion of the authors, it appears that all three identified clusters plot within (very) close proximity to existing normal faults. This is a finding that (independently) validates the authors’ results on the clusters, and which the authors should acknowledge (and welcome). Note that this fault is also partly mapped in the 1:100,000 scale of EPPOS (1991). But this also requires that some of their discussion/interpretations - and conclusions - should be modified accordingly. First, the clusters north of Zacharo appear to spatially, geometrically & kinematically correlate with AFG’s Afios Fault, that is a dip-slip fault on a westward dipping plane. In this context, Afios fault appears mostly as a transfer structure linking E-W striking normal faults north and south of its extent. This observation, which is also supported by the authors’ identified clusters, stress tensor analysis (Konstantinou et al. 2017; Kapetanidis and Kassaras, 2019) and geodetic datasets (i.e. Chousianitis et al. 2015, 2024), is at odds with the authors’ interpretation that this fault results from a ‘change in the stress field’ in this part of Peloponnese associated with NW-SE extension. Hence, by not accounting for the overall faulting patterns around Zacharo, the authors have, in my view, misinterpreted the strike of Afios Fault (and of their ‘Zacharo’ clusters) as ‘diagnostic’ of a major change in the regional kinematics between Zacharo and Vartholomio (when it is likely a transfer structure between large scale south-dipping normal faults; see Begg et al. 2025 and references therein). Similar ‘relaying’ patterns occur commonly in Greece at various spatial scales (e.g. Poulimenos, 2000; Delogkos et al. 2017; Mouslopoulou et al. 2022).
Similarly, the cluster in ‘Vartholomio’ appears to partly overlap with the mapped N-S striking and E-dipping Vranas Fault, located west of Vartholomio (Begg et al. 2025). While the cluster is associated with some strike-slip earthquakes, and the authors characterize it as strike-slip, it has a long-term signature on the landscape that requires some dip-slip (e.g. it forms a clear morphological step ‘down to the east’) and could be interpreted as an oblique-slip fault (i.e. this would also justify the mixed nature of recorded moment tensors). Offshore studies, in very close proximity to the Vartholomio cluster, support also co-existence of dip-slip and strike-slip faulting (and even their coseismic rupture) (see Kokalas et al., 2013; Wardell et al., 2014; Mouslopoulou et al. 2020, 2025; Sachpazi et al. 2020).
The above have implications in the discussion of ‘loading’ mechanisms. The way the manuscript argues, it is like there is a spatial separation in the fault/earthquake dynamics and the majority of faults proximal to the ‘Zacharo’ cluster will favour a ‘fluid-driven’ swarm-like precursors and the majority of faults near Vartholomio will be associated with a ‘tectonic’ loading. I disagree with this interpretation (and I explain above why). Further, exactly beneath the Vartholomio area but also further offshore, Mouslopoulou et al. 2020 have shown that there is swarm-like earthquake sequences associated with aseismic slip (preceded and followed), contradicting the author’s ‘conclusion’. The seismic-aseismic interplay in the region perhaps should also be commented with respect to the new locking maps presented by Chousianitis et al. (2026) for upper-plate faults. Hence, in my view, although the findings are robust, they cannot be used to indicate or diagnose the generic conclusions the authors present. Instead, I would advise the authors to carefully describe their findings with respect to an up-to-date view of the kinematics and limit far reaching interpretations which cannot be supported by the presented dataset.
Another issue relates to the major kinematic change reported by the authors between the regions of Zacharo and Vartholomio. I see indeed a significant change in the extension direction which appears to rotate by about 90 degrees from a roughly E-W (i.e. many faults around the area of Vartholomio and Kylini strike N-S) to a roughly N-S (many faults around Pyrgos, Zacharo strike E-W). I disagree with the authors’ stated opinion that there is an overall NE-SW extension direction near the township of Zacharo. If that was the case, it would have been systematically recorded by landscape features (which is not the case) and also recorded by other short-term datasets (i.e. geodesy, moment tensors, etc). Instead, the fault that extends parallel to the 'Zacharo' clusters (the NW-SE striking Afios Fault, downfaulted to the southwest), appears to act as a relay structure, linking E-W striking faults (Begg et al. 2025).
Further, the Introduction would benefit from some generic questions being laid out early on to help readers understand the context of the study and (perhaps) its international significance. It needs some restructuring.
Below I provide some non-exhaustive line specific comments which I hope the authors will find useful. They must also scan the manuscript to identify similar problems and also improve the clarity of some sentences.
Non exhaustive line specific comments:
Lines 25-26:
Line 36: Not entirely true. Nikolopoulou et al. (2025) presents an excellent representation of the Movri fault through relocated post-2008 seismicity.
Lines 35, 43, 48 and elsewhere: As more GNSS stations became available during the last decade onshore Greece, Chousianitis et al. 2024 presents the most up to date regional GNSS dataset and should be used here instead of (or in addition to) Chousianitis et al. 2015.
Lines 45-48: This change in extension direction onshore Peloponnese is also nicely reflected by the patterns of active normal faulting southward across Peloponnese (see Begg et al. 2025). The greatest kinematic change appears across central-south Peloponnese.
Lines 53-54: There is another recent study, that of Chousianitis et al. 2026, that discusses locking distribution across Peloponnese. Perhaps you should also refer to that?
Lines 63-65: The closing statement in the introduction should be improved. The specific key questions that this study aims to address should be introduced, perhaps with a sentence summarizing the main findings. The methodological part is of less interest to the reader.
Section 4.1.1: This section does not really describe the Seismicity in the Continental Crust or any Results - it is rather a methodological description of the analysis performed. Perhaps modify the title and relocate to the respective section?
Line 278: can you translate seconds to minutes (from 1 to xx minutes) so that it is more meaningful while reading the manuscript?
Lines 395-397: More than that, I would say that this is how the landscape forms (basins and ranges) due to accrued slip on faults due to repeated earthquakes. A classic reference is that of King et al. (1988).
Lines 410-418: Differences and similarities between the two sequences examined, should be stated clearly, perhaps in the opening paragraph here. The structure of the sentences used here, is confusing. Please rewrite. From the references provided and the maps illustrated, it is unclear how this comparison is justified. The fault maps presented are inadequate, as many active faults are missing. Begg et al. (2025) have mapped numerous active faults in the region, including faults that align with the Zacharo and Vartholomio clusters. The Afios Fault, for example, extends 10-20 km NNW from Zacharo, and appears to align well with the recorded clustered seismicity – but it would be best if the authors superimposed them for proper comparison. The long-term evidence on the landscape provided by active fault traces should be the primary criterion upon which the comparison should be based.
Lines 419-420: I am not sure that I fully agree with that statement. The majority of faults near the township of Zacharo have E-W strikes (see Begg et al. 2025). The major change in the strike of normal faulting occurs further south and east (at central-south Peloponnese) where most faults switch to NNW-SSE strike (reflecting roughly E-W extension). This switch is supported by the mapped landscape features (see strike in the AFG faults in central-south Peloponnese), geodesy and seismology (Konstantinou et al. 2017; Kapetanidis & Kassaras, 2019; Chousianitis et al. 2024).
Lines 422-423: Where is the evidence for NE-SW extension near Zacharo? I can see numerous large E-W striking faults, consistent with N-S extension and corresponding patterns in geodetic strain rates (Chousianitis et al. 2024) (see vectors near Pyrgos, Katakolon, Zacharo in Chousianitis et al. 2024). Instead, I believe that your clusters capture unrest on a transfer NE-SW striking structure that links large E-W striking normal faults – this fault may be the Afios Fault mapped by Begg et al. 2025 (note that both the cluster and the fault appear to dip to the SW). Hence, this part of Peloponnese appears to respond predominantly to N-S extension, which is accommodated by numerous large E-W trending normal faults.
Lines 425-426: Indeed, I see a significant change occurring between Zacharo and Vartholomio: the kinematics appear to transition from roughly E-W extension (i.e. many faults around the area of Vartholomio and Kylini strike N-S) to N-S extension (many faults around Pyrgos, Zacharo strike E-W). I disagree with the opinion that there is NE-SW extension near Zacharo. If that was the case, it would have been recorded on the landscape by corresponding faulting. The fault that roughly parallels the 'Zacharo' clusters (NW-SE), is the Afios Fault and appears to be a transfer structure (see main comments above).
Line 431: Not entirely true. See Mouslopoulou et al. 2020 - there is a swarm like activity nearby (offshore western Peloponnese), that partly overlaps with the study area. Further, Line 435: This has been shown very close to the study area, where SSEs preceded and followed earthquake swarms before the 2018 Zakynthos earthquake (Mouslopoulou et al. 2020 & Saltogianni et al. 2021).
Lines 469-471: Neither the geodetic/seismological data (i.e. Konstantinou et al. 2017; Kapetanidis & Kassaras, 2019; Chousianitis et al. 2024) are uniformly resolving NE-SW extension nor the mapped faults support this statement. A N-S striking fault (down to the east) lies very close to this cluster and indicates, at least partly, dip-slip (Vranas Fault in Begg et al. 2025). This may be an oblique-slip fault (i.e. accommodating both strike-slip and dip-slip).
Lines 480-490: I am not convinced that this paragraph is fully supported by the data. The cluster has a roughly N-S strike and has hosted a strike-slip and a normal fault earthquake within about a day and, later, another strike-slip event. The first two appear to be spatially separated while the second strike-slip event occurs on a similar part of the plane as the first strike-slip event. The region is populated with normal and strike-slip faults (mostly offshore). Very close to this cluster, there is a series of N-S trending normal faults mapped, with the nearest being down to the east (consistent with the uplifted topography at its west). Thus, a more detailed discussion is required here, relating the specific cluster to nearby features and not to the >30 km inferred fault that ruptured in 2008.
Conclusions: should be revised to accommodate the proposed review comments.
Figure 1: How about Halpaap et al. 2018, 2019? I think they have resolved using seismic tomography the top of the slab beneath western and central Peloponnese. Why not using their model? The active faults in Peloponnese appear to be minimum. Please include the known mapped faults and compare your results against them. There are a number of databases dedicated to Greece and Peloponnese, with the most complete (1: 25,000 scale mapping) being the recent work of Begg et al. (2025). Plotting the real number and geometry of active faults is vital for this study. This comment applies for figures 1-2 and 5-6.
Caption of Figure 1: Clarify ‘long term’ and ‘new’ seismic stations. Does 'new' mean 'temporary'? Also, how can we visually separate 'new' from 'long-term' stations?
Figure 2a: you need to explain what are the colour coded boxes and on which basis you limit their size. Also, why don’t you plot the location and kinematics of the two large (M 5-6) 1988 Vartholomio Earthquakes (Papazachos and Papazachou, 2003)?
Figure 2b: The way this graph is presented the earthquakes do not appear to be spatially linked - they are only linked temporally. The two highlighted areas include events in the proximity of Vartholomio and Zaharo, as well as the remaining area of interest. The presentation of this sequence of graphs appears to show that the increase in the seismicity rate occurs only due to earthquakes included in the colour coded boxes - which is not the case. This should be clarified and, if needed, corrected.
Figure 3a: For two of the indicated time periods (around March 2024), I don’t see that they correspond to daily averages below 10-3. Can you confirm?
Figure 3b: Timelines for the Vartholomio and Zaharo events in Figure 3b should be restricted at their corresponding latitudes (37.5 and 37.8). The way it is presented is confusing.
Caption of Figure 4: Explain what these acronyms mean (Rij, etc).
Figures 5 & 6: update active faults and add scale.
Figure 6b shows that the activity at the northern end of the cluster is at shallower depths than the activity at the southern end of the cluster (near the coastline). Figure 6c: You should reverse the vertical scale to match the figure (A' should be in the north).
References used in this review:
Begg et al. (2025) - Scientific Data: https://doi.org/10.1038/s41597-025-06283-z
Chousianitis et al. (2015) – Journal of Geophysical Research: https://doi.org/10.1002/2014JB011762
Chousianitis et al. (2024) – Journal of Geophysical Research: https://doi.org/10.1029/2023JB028004.
Chousianitis et al. 2026 - Journal of Geophysical Research: https://doi.org/10.1029/2025JB033218
Delogkos et al (2017) – Geol. Soc. London, Sp. Publ. – https://doi.org/10.1144/SP439.19
EPPOS (Earthquake Planning & Protection Organisation), 1991. Neotectonic Map of Greece, scale 1:100,000: "Pyrgos" sheet.
Halpaap et al. (2019) – Science Advances: https://doi.org/10.1126/sciadv.aav7369
Kapetanidis & Kassaras (2019). Journal of Geodynamics: 10.1016/j.jog.2018.11.004
King et al. (1988) J. Geophys. Res. doi:10.1029/JB093iB11p13307.
Kokkalas et al. (2013) – Tectonophysics: https://doi.org/10.1016/j.tecto.2012.08.004
Konstantinou et al. (2017) - Journal of Geophysical Research: 10.1002/2016JB013272.
Mouslopoulou et al. (2020) - G-cubed: doi:10.1029/2020GC009243
Mouslopoulou et al. 2022– Tectonics: https://doi.org/10.1029/2022TC007453
Mouslopoulou et al. (2025) – Tectonics: 10.1029/2025TC008943
Papazachos B.C., Papazachou C., 2003. The earthquakes of Greece. Ziti publications, Thessaloniki, Greece, 286 pp.
Perouse et al. 2017 - International Journal of Earth Sciences: https://doi.org/10.1007/s00531-016-1345-9
Poulimenos (2000) – Journal of Structural Geology – https://doi.org/10.1016/S0191-8141(99)00152-2
Nikolopoulou et al. (2025) – Tectonophysics: 10.1016/j.tecto.2025.230889
Sachpazi et al. (2020) – Tectonophysics – https://doi.org/10.1016/j.tecto.2020.228643
Saltogianni et al. 2021 – GRL - https://doi.org/10.1029/2021GC010090
Wardell et al. (2014) – https://doi.org/10.4430/bgta0087