the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Jan 2024
Interseismic and long-term deformation of southeastern Sicily driven by the Ionian slab roll-back
Abstract. New satellite geodetic data challenge our knowledge of the deformation mechanisms driving the active deformations affecting Southeastern Sicily. The PS-InSAR measurements evidence a generalized subsidence and an eastward tilting of the Hyblean Plateau combined with a local relative uplift along its eastern coast. In order to find a mechanical explanation for the present-day strain field, we investigate short and large-scale surface-to-crustal deformation processes. Geological and geophysical data suggest that the southward migration of the Calabrian subduction could be the causative geodynamic process. We evaluate this hypothesis using flexural modeling and show that the overloading of the Calabrian accretionary prism, combined with the downward pull force induced by the Ionian slab roll-back, are capable of flexuring the adjacent Hyblean continental crust, explaining the measured large-scale subsidence and eastward bending of the Hyblean Plateau. To explain the short-scale relative uplift evidenced along the eastern coast, we perform elastic modeling on identified or inferred onshore and offshore normal faults. We also investigate the potential effects of other deformation processes including upwelling mantle flow, volcanic deflation, and hydrologic loading. Our results enable us to propose an original seismic cycle model for Southeastern Sicily, linking the current interseismic strain field and the available long-term deformation data. This model is mainly driven by the southward migration of the Ionian slab roll-back which induces a downward force capable to flexure the Hyblean crust.
- Preprint
(21350 KB) - Metadata XML
-
Supplement
(10935 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2024-85', Mara Monica Tiberti, 27 Feb 2024
General comments
The topic is interesting and within the scope of SE; the title and abstracts are adequate.
The paper presents a deformation model of continental lithosphere adjacent to a retreating subduction zone (the Calabrian subduction in the Mediterranean Sea), trying to explain the observed complex pattern of uplift and subsidence at different wavelengths. The input data are already published; the original part is the model itself.
While the long wavelength behaviour can be easily explained with the combined action of the slab pull and the load of the accretionary wedge, the short wavelength pattern requires more discussion and cannot be completely fitted with the available constraints. The conclusions are, in general, supported by the data, except for one excessively speculative topic that, however, is quite well discussed in the appropriate section (see “Specific comments” below).
The manuscript is clearly written, except for the introduction, which suffers from a few misunderstandings about the cited papers (see “Specific comments” below).
Specific comments
The introduction needs a consistent revision. The authors often seem to have misunderstood the contents of the cited papers, and the resulting geological framework is unclear. This may confuse the readers instead of orienting them. See the attached file for suggestions and explanations.
The profile along with the calculations are carried out is not accurate. Its location should be shifted towards E. Otherwise, it crosses an area of transition between continental and oceanic crust where thicknesses (especially the accretionary wedge thickness) do not correspond to what is represented in Figure 5 and to what is used in the calculations.
In addition, the authors use a model of the slab (Hayes et al., 2018) that is good for a global survey but does not take into account local complexities and constraints. This results in a lack of accuracy of the depth of the top of the oceanic crust beneath the accretionary wedge. A 3D model of the subduction interface specifically built for the Calabrian subduction is recommended. See the attached file for suggestions and explanations.
Finally, I should point out that the faults that justify the short wavelength pattern of uplift and subsidence are very speculative (see the attached file for further comments). This is well stated in the discussion section, but should be better addressed within the results section and in the conclusions as well (that those faults are speculative).
Technical comments
See the attached file.
-
AC1: 'Reply on RC1', Amélie Viger, 01 Mar 2024
First of all, we would like to thank you for taking the time to review our manuscript. We appreciate your comments and interest in our study.Before answering your comments and questions, we would like to underline that our study should be seen as a first attempt to explain the current surface deformation and associated seismic cycle of SE Sicily. Our study is, therefore, essentially based on published geological and geophysical data, that we considered of sufficient quality and resolution (even if in some cases just acceptable) to investigate successfully first-order processes. We are aware that further investigations will be needed to go into greater detail and validate some of our interpretations. That is particularly true for the extrado deformation hypothesis (and potentially associated surface faulting/folding). We want to be very careful on this point, and as you suggested, we will be adding dedicated sentences in the Results and Conclusion sections.One of the main points of your review is that we should use the Maesano et al. (2017) dataset rather than the Hayes et al., 2018 dataset and move eastward the CD profile. You are right, we made a mistake in Figure 5; the CD profile should be located further East. Note that looking at the shapes of the depth iso-contours, won't change the geometry of the top of the Ionian crust. As we are only dealing here with a very long wavelength signal (>~100km), this also has slight consequences on the AB flexural profile. However, we agree on using the Maesano et al. (2017) dataset and have just started comparing it with the Hayes et al. (2018) one.Finally, with regard to the introduction of the manuscript, we agree with you that some sentences are unclear or may confuse the readers. This is not because we have misunderstood the cited articles but mainly because we tried to shorten as much as possible a manuscript that is already quite long. We, therefore, recognize that we have sometimes made some unfortunate simplifications. Of course, we will take into account all your specific comments included in your annotated version of our manuscript.Citation: https://doi.org/
10.5194/egusphere-2024-85-AC1
-
AC1: 'Reply on RC1', Amélie Viger, 01 Mar 2024
-
RC2: 'Comment on egusphere-2024-85', Andrea Argnani, 01 Mar 2024
Publisher’s note: this comment was edited on 1 March 2024. The following text is not identical to the original comment, but the adjustments were minor without effect on the scientific meaning.
Comments to Authors
The authors use an elegant approach to attempt explaining the subsidence rates obtained from InSAR data on the Hyblean Plateau, following on a previously published paper (Henriquet et al 2022). However, this approach suffers in places of too many assumptions and ends up to be too generic and unrelated to the existing geological data. In the following the main sections of the paper are commented, with also notes on minor points. Some parts need substantial improvement, particularly the section where the authors discuss the relationships between the combined short-term and long-term tectonics and the seismic cycle, which is rather confused and disappointing, and fails to meet the expectations of the title.
lines 40-41: too many papers cited, and none is properly seismological. The reference Azzaro and Barbano 2000 is sufficient to give a picture of historical seismicity, without citing papers that are not focussed on this issue: Billi et al argued that the 1693 tsunami was caused by a submarine landslide, without implication on the location of the earthquake, and Gutscher et al suggested that the 1693 earthquake originated in the subduction interface, away from the Hyblean Plateau.
lines 45-50: The complexity of the Mediterranean tectonic evolution has been oversimplified: Alpine Tethys subducted under the Apulia-Adria microcontinent (therefore dipping southward). Since the Oligocene the same Alpine Tethys subduction has experienced slab roll-back and the opening of back-arc basins (therefore dipping northward). To avoid unnecessary explanations of how this change of subduction polarity occurred, it would be better to just describe the Oligocene to present evolution.
In the captions of Fig 2 (panel a) and Fig 7 (panel b), the so called "turbiditic vally faults" are indicated as identified by Gutscher et al 2016. This fault system however, is the same previously mapped by Argnani and Bonazzi 2005. Gutscher et al 2016, mapped the morphological unit called "turbidite valley", but did not name the faults, seen on just one profile, within this domain. Previous works should be acknowledged properly.
lines 91-94: the paper of Argnani and Bonazzi 2005 is a more appropriate reference because they mapped the active fault system after interpreting ca. 2500 km of seismic profiles, purposely acquired over the Malta Escarpment. It is true that the acquisition of new data helped improving the comprehension of active tectonics along the Malta Escarpment; however, strictly speaking, Gambino et al 2021 did not contribute to any acquisition of geophysical data, but used previously published data.
lines 198-199: Bianca et al offer an incomplete picture of the faults because of limited data coverage, and the fault system in Gambino et al is the same as Argnani and Bonazzi 2005 that should be cited.
2.3 The synthetic structural profile (Fig. 4)
lines 263-266: not necessarily several faults in the Turbidite Valley; for other authors (Argnani and Bonazzi, 2005) there is just one fault with a splay. The three "major parallel faults" of Gambino et al 2021 are too close to be independent faults at crustal scale (e.g., Argnani 2021 Frontiers Earth Sci.).
lines 268-269: "potentially related to recent re-activation of the shallow propagation of the inferred Mesozoic tilted blocks" It is difficult to grasp the meaning of this statement, which appears very speculative.
The crustal-scale cross section has many critical points that deserve some attention.
i) The choice of using the well Palma1 does not make much sense: the well is ca. 40 km away from the cross section, whereas other wells, much closer, are available in the VIDEPI repository, like Plinio Sud1. By the way, Palma1 reaches the Triassic, unlike what is represented in Fig. 4.
ii) The refraction profile DT-P3 is located at the northern end of the Malta Escarpment. The authors have arbitrarily drawn similar velocity contours on their crustal cross section (Fig. 4), which is located ca. 20 km farther south. It should be noted that in doing so the the values at the crossing with DY-P4 have not been respected. The uncontrolled isovelocty contours have then been used to draw two extensional faults that reach the base of the crust. In the authors interpretation these two faults are intended to correspond to the Malta Escarpment fault and the "turbiditic valley" fault. But the line of reasoning is too speculative and has no supporting evidence.
iii) Inconsistent match with Dellong profile DY-P4: in profile AB the thickness of the CAP (interval 4.9 to 5.1 km/s) is by far too thick at the crossing. It should be noted that Dellong et al 2020 (G3-Rep), in their reply to Argnani 2020 (G3), interpret the unit bounded by the isovelocities 4.9-5.1 km/s as non-sedimentary, and belonging to the Calabrian basement. However, in the profile in panel (b) the subduction decollement is located between the lower and upper crust, and the CAP is indicated above the upper crust; this creates some ambiguity: is the subduction decollement different from the base of the CAP? And if so, where is the base of the CAP? Dellong et al 2020 (G3-rep) (Fig 1) mark the lower and upper oceanic crust, however, at the crossing with profile AB part of the upper-plate crust should be present as well. The authors should better describe the interpretation of the structural relationships in this sector of the profile.
iv) The crossing with the Ionian Fault (IF) is marked on the profile, but its position cannot be right: according to Dellong et al. 2020a the IF is located east of profile DY-P4, likely outside the AB cross section. In some literature the Alfeo Fault System has no expression on the cross section, although it is considered a lithospheric fault. None of these two fault systems are mentioned in the text. A mechanical connection between the Hyblean and Ionian lithosphere is inferred. Perhaps the authors should comment on how this assumption fits with the occurrence of the Alfeo lithospheric fault?
v) The Triassic units are represented as ca. 8 km thick and the lower crust is indicated as Palaeozoic. Is there any evidence for these attributions? These details, which are not relevant to the adopted modelling, should be omitted, unless they are based on some evidence.
vi) A up to 10 km-thick succession of Jurassic-Cretaceous and Eocene-Quaternary is represented in the cross section (130-170 km). These units are connected to the Mesozoic units of the Hyblean Plateau and are overlying the CAP. These relationships do not make much sense: the CAP originated in Cenozoic time so it cannot be covered by Jurassic-Cretaceous sediments. Note that Dellong et al 2020rep consider this lower velocity domain as a "basin" filled by "accretionary wedge sediments". Though I do not agree with their definition of basin, the evidence from seismic reflection profiles (Argnani and Bonazzi 2005) supports the presence of an accretionary prism, as in Argnani 2020 (G3). Also note that Triassic rocks have been dredged at the base of the Malta Escarpment (Scandone et al 1981), just to undermine the stratigraphic correlation between the Hyblean and the Ionian regions presented in the cross section of Fig 4.
Summing up, some of the domains drawn in the cross section, particularly on the Malta Escarpment-Ionian sector do not respect the existing data, whereas others are based on unproven or unlikely assumptions.
- Mechanical model hypotheses
lines 296-296: as above, the offshore active fault system have been first described by Argnani and Bonazzi 2005
3.1 Lithospheric flexure along a NNW-SSE profile
lines 308-310: a Permo-Triassic age is attributed to the Ionian oceanic lithosphere. This issue is debated; however, whereas Speranza et al. agree with this age, for Catalano et al. the oceanic spreading is Jurassic, following a Permo-Triassic rifting. It should be noted that a Permo-Triassic age of the Ionian ocean contrasts with what stated in lines 259-261, where a Triassic-Jurassic rifting is assumed.
line 350: a density of 2800 kg/m3 is used for the sediments, with reference to Dellong et al 2020, which, however, attribute this density to the Ionian oceanic crust.
3.3 Interseismic loading and aseismic creep
This section is highly speculative and is penalized by many assumptions and lack of constraints
The faults addressed by Viger et al are shown in Fig 2 and consist of the Malta Escarpment and the Turbidite Valley faults, both located offshore. Whereas the latter corresponds to the fault mapped by Argnani and Bonazzi 2005 and subsequently by Gambino et al 2021, the Malta Escarpment fault has not been documented. It was drawn by Gutscher et al from a morpho-bathymetry, marking the base of a scarp; however, on seismic profiles this scarp appears just a morphological feature passively onlapped by sedimentary strata (Argnani and Bonazzi 2005 and Gambino et al., 2021). The Augusta-Syracusa Fault and the Onshore Fault are also poorly constrained, and seem too short to account for the short-wavelength subsidence anomaly observed in the entire eastern Hyblean plateau. Summing up, out of the four fault systems considered by the authors only the easternmost one seems properly documented (Argnani and Bonazzi 2005). This undermines the use of the Augusta-Syracusa and Onshore faults in the interseismic elastic modelling of Fig. 7.
Fault modelling has been carried out without constraints on fault geometry, particularly for the two faults that can explain the short-wavelength surface deformation, i.e., the Onshore and the Augusta-Syracusa Faults. The fault parameters have been chosen ad hoc in order to fit the data, but the lack of geological constraints on the faults, and even on the occurrence on these faults, undermines this part of the modelling.
3.4 Alternative hypotheses
The authors explore three hypotheses that could account for the InSAR observation. However, two of them are not realistic: in the Hyblean region published modelling of mantle flow upwelling shows no effects, and no recent volcanic activity has been reported. The hydrologic loading could possible contribute to the subsidence, but the authors reject this hypothesis saying that the data required to test it are out of the scope of the paper..... which means hypothesis not tested.
Considering that the pattern of InSAR-derived subsidence is fairly consistent over the entire Hyblean Plateau, the effect of glacial isostatic adjustment (GIA) should perhaps be taken into account. The modelling of the GIA in the Mediterranean region predicts some subsidence in the Hyblean region (Spada and Melini 2022); the order of magnitude of the subsidence rate is comparable to the subsidence observed by InSAR.
line 538: Faccenna et al. 2005 instead of Faccenna 2005
- Discussion
4.1 Short-term and long-term model limits
lines 581-585: Similar mechanical properties are assumed from the Hyblean continental to the Ionian oceanic lithosphere: how is this compatible with the occurrence of the Alfeo lithospheric fault crossing the AB profile?
lines 631-633: the N-S topographic step is not obvious and should be indicated in the figure. It is suggested (lines 647-650) that the step could be the expression of a creeping fault, although the Onshore fault, used for modelling the InSAR velocities, is located farther eastward. Moreover, GPS velocities at the northern border of the Hyblean Plateau suggest a contractional regime (Mastrolembo Ventura et al., 2014) which could be responsible for the relative uplift recorded by the geodetic transect. The point in this section is not clear, and the authors should explain it further.
4.2 Combined long-term tectonics and seismic cycle model
line 661: perhaps "Augusta-Syracusa fault and onshore fault" instead of "Augusta-Syracusa fault and offshore fault"
InSAR data show that the eastern Hyblean Plateau is undergoing subsidence, though with variable rates; the pattern describes an overall, long-wavelength eastward increase of subsidence with a slower subsidence near the coast, with a shorter wavelength. The two signals coexist in what the authors consider an interseismic period. The lithospheric flexure, caused mainly by slab pull, can explain, according to the authors the eastward increasing subsidence, whereas the short wavelength relative uplift is considered as due to two creeping faults located onshore. However, Late Quaternary marine terraces testify an overall slow uplift of the Hyblean Plateau and this is explained as produced during the coseismic-postseismic period, when the offshore faults (more likely the only one which is documented as active) rupture. However, Meschis et al have shown that the footwall uplift of the offshore fault is much less of what required to account for the uplift of marine terraces; this invalidates the logic of the seismic cycle proposed by the authors. The possible causes of additional uplift proposed by the authors are extremely speculative. Even more doubtful is their connection with earthquakes in the coseismic period. The effect of a lithospheric tear decoupling the Ionian from the Hyblean lithosphere would likely be to switch off the slab pull, and the resulting subsidence. The onshore creeping faults are considered as extrados faults, though the extremely bland arching shown in the cross section without vertical exaggeration (Fig 4b) is not really supporting this interpretation.
Figures
Fig 1: the figure is taken, with only minor changes, from Henriquet et al 2020. It is fine to cite the source of the focal mechanisms and GNSS data, but the rest of the geological citations could be avoided. These citations repeat the caption of Henriquet et al where the figure covered a larger area and are not really relevant for the present study. For instance, the study area of Lymer et al does not overlap the map in Fig 1. The map of Corti et al was not original but was taken from Tricart et al 1994. Chamote-Rooke et al. is only relevant to the part east of 18°E, and Rabaude and Chamote-Rooke is just an extended abstract focussed on the Algerian margin and, only marginally, on the northern Sicily margin.
Fig 2: the panels (a) and (b) are not mentioned in the caption.
Fig 4: the map in panel (c) and the text (lower right) are too small to read. The map needs to be enlarged, whereas a larger version of panels (a) and (b), with the legend of the horizons, should go in the supplementary materials .
Fig 5d: perhaps in panel (d) it should be dashed lines Te = 27 km and continuous lines Te = 30 km.
Fig 9: It could be useful to show on the shaded relief map the marine terraces mentioned in the text. In panel (b) the potential fault/fold scarps (thin, dashed black line?) should be better indicated.
Citation: https://doi.org/10.5194/egusphere-2024-85-RC2 -
AC2: 'Reply on RC2', Amélie Viger, 01 Mar 2024
Thank you very much for your availability and your detailed review. We have taken note of your comments and suggestions for corrections.In particular concerning the need to improve the citations, to better take into account available geological data and to improve the presentation of some of our less constrained hypotheses.We will also clarify the relevance of our final seismic cycle model, which we consider the most original part of the manuscript, since the process evoked has never been proposed (as far as we know) as a source for elastic loading on major active faults.Citation: https://doi.org/
10.5194/egusphere-2024-85-AC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 234 | 76 | 22 | 332 | 44 | 10 | 8 |
- HTML: 234
- PDF: 76
- XML: 22
- Total: 332
- Supplement: 44
- BibTeX: 10
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1