the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 09 Sep 2026
3D crustal model of Southern Apennines (Italy) by integrating geological, geophysical and petrophysical constraints
Abstract. The Southern Apennines in Italy represent a geologically complex and seismically active region shaped by the ongoing convergence between the African and Eurasian plates. In this study, we present a high-resolution 3D crustal model of the Southern Apennines, parameterized on a 5 × 5 × 1 km³ grid and constructed through the systematic integration of geological, geophysical, and petrophysical data. The considered dataset includes laboratory velocity measurements (vp) under different pressure and temperature conditions, stratigraphic well logs, seismic tomography (vp), and geophysical models from inversion of gravimetric and magnetic data. Pressure- and temperature-dependent corrections were applied to P-wave velocities derived from regional seismic tomography, enabling us to delineate key intra-crustal boundaries, including the top of the Mesozoic limestones, the base of the Mesozoic sedimentary succession, the onset of the felsic crystalline basement, and the deeper transition to mafic, gabbroic lower-crustal units, within a 3D volume extending down to the Moho discontinuity. This modelling strategy ensures full internal consistency across spatial scales and tectonic domains, and is validated using independent borehole data, receiver functions, and other geophysical interpretations. The resulting 3D volume offers a robust framework for investigating crustal architecture and supports applications in geophysical modeling and seismic hazard assessments. All datasets are openly accessible under FAIR principles, promoting transparency, reproducibility, and interdisciplinary reuse of the results.
- Preprint
(2505 KB) - Metadata XML
-
Supplement
(799 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on essd-2026-513', Anonymous Referee #1, 25 Oct 2026
-
RC2: 'Comment on essd-2026-513', Juan Alcalde, 29 Oct 2026
The manuscript presents a 3D crustal model of the Southern Apennines, integrating geological, petrophysical and geophysical datasets into a common framework and comparing the results with independent observations. Overall, I think this is a very good manuscript. It is well written, concise and clearly explains the modelling process and the datasets used. In fact, several questions that came to mind while reading were answered shortly afterwards, which is always a good sign! The fact that the models are also made available adds considerable scientific value to the work.
I have a few relatively minor comments that I think could help improve the manuscript, but my overall assessment is very positive, and I would like to congratulate the authors on their work. The comments may look rather long, but I think they should be fairly straightforward to address.
- I feel that the abstract does not quite do justice to the work presented in the manuscript. It goes straight into the study without clearly explaining the problem being addressed or the main scientific motivation. I am not sure whether this is due to word limits, but I think the abstract could better reflect the scope and value of the work, particularly the methodology used to integrate the different datasets.
- Related to this, I found it difficult to understand from the abstract and introduction how the gravity and magnetic data contribute to the modelling. It only became clear when reading Section 3 that these datasets are mainly used as independent constraints to validate the model. I think the distinction between the datasets used to construct the model and those used to evaluate it should be clearer from the beginning.
- Another aspect that took me some time to understand was how the laboratory P-wave velocity measurements, and their dependence on pressure and temperature, were incorporated into the modelling. As I understand it, the authors apply pressure and temperature corrections to their petrophysical data based on relationships derived from previous experimental studies. If this is correct, I think the explanation could be slightly refined to make the workflow and the role of these corrections clearer.
- The authors use the Wyllie et al. (1956, 1958) equation to establish a relationship between P-wave velocity and porosity. However, Wyllie et al. (1958) themselves acknowledged the limitations of this relationship when applied to carbonates, particularly those with complex pore structures. More recent studies (e.g. Anselmetti and Eberli, 1993, 1999; Jeppson and Kitajima, 2022) have further explored these limitations and alternative approaches. I think the authors should at least acknowledge these limitations and assess whether the Wyllie equation is the most appropriate choice for the carbonate lithologies considered in this study. It would also be useful to discuss the sensitivity of their results to this choice, as it could partly explain the discrepancies between their model and those of Kelemework et al. (2021) and Vezzani et al. (2010) shown in Figure 11.
- My last general comment concerns the comparison of the model with independent geophysical interpretations. I think the discussion of how uncertainties in the input datasets may explain differences between models is very appropriate. However, adding some brief quantification of these differences (e.g. percentages or ranges) would make the comparison more informative and strengthen the final results.
I have included a few additional minor comments and suggested edits below.
Congratulations again to the authors on this excellent manuscript!
References:
Anselmetti, F. S., & Eberli, G. P. (1993). Controls on sonic velocity in carbonates. Pure and Applied geophysics, 141(2), 287-323.
Anselmetti, F. S., & Eberli, G. P. (1999). The velocity-deviation log: a tool to predict pore type and permeability trends in carbonate drill holes from sonic and porosity or density logs. AAPG bulletin, 83(3), 450-466.
Jeppson, T. N., & Kitajima, H. (2022). Velocity‐porosity relations in carbonate and siliciclastic subduction zone input materials. Geochemistry, Geophysics, Geosystems, 23(1), e2021GC010074.Minor comments
Abstract
- L7: Vp is already defined in the previous line.
Introduction
- L18–26: The first paragraph is rather vague. Several statements lack sufficient explanation: why are reference models useful for potential field modelling (L18)? Why are they crucial for numerical simulations (L22)? Which complementary datasets are being referred to (L24)? What is the purpose of accounting for crustal heterogeneity (L26)? These statements should be refined to make the underlying reasoning clearer.
- L29: Why is it "critical"?
- L38–43: This sentence is rather long. Consider splitting it.
- L53–54: "The only fully 3D and spatially continuous coverage": what about the gravity and magnetic datasets mentioned in the abstract?
Rest of the sections
- Fig. 1: Is "Petrophysical approach" the most appropriate term here? Perhaps "Petrophysical analysis" or "Petrophysical interpretation"?
- L317: It would be useful to briefly discuss potential issues associated with downscaling the data to match the different datasets. The spatial resolution could also be included as an additional column in Table 1.
- L339: Use (5 × 5 × 1) km³. This notation should be consistent throughout the manuscript.
- Fig. 6: The lines within the grey shaded area are barely visible in the printed version. Please consider using a different colour.
- Fig. 7: Please mark the well more clearly and clarify in the caption that the box indicates its extent.
- L490–495: The description of the subpanels is unnecessary here, as it is already included in the figure caption.
- L596: Consider "across a portion of the study area", or similar.
- L797–708: Would it be more appropriate to mention the project in the Acknowledgements rather than in the Conclusions?
---------------------------
Note: I have used AI only to help format and polish these comments. All the scientific observations and opinions are my own.
Citation: https://doi.org/10.5194/essd-2026-513-RC2
Data sets
Source Data of "3D crustal model of Southern Apennines (Italy) by integrating geological, geophysical and petrophysical constraints M. Perrini et al. https://doi.org/10.5281/zenodo.21130189
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 472 | 214 | 71 | 757 | 103 | 83 | 73 |
- HTML: 472
- PDF: 214
- XML: 71
- Total: 757
- Supplement: 103
- BibTeX: 83
- EndNote: 73
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Review of the manuscript ‘3D crustal model of Southern Apennines (Italy) by integrating geological, geophysical and petrophysical constraints’ by Perrini et al. (essd-2026-513)
The authors present a 3D geophysical/geological model of the Earth’s crust for Southern Apennines area, and describe the method of its preparation based by combining information from several datasets. The MS is properly structured, concise and relatively well written, (although, in some parts, explanation of the procedure seemed unclear to me). The authors prepared a crustal model of a tectonically complex area, which can be valuable for other geophysicists involved in crustal studies of this region. However, I have some doubts, primarily about the consistency/correctness of the model itself, and about the model creation method, I discuss it in detail below. The idea of finding layer boundaries in a smooth tomographic model in a proposed way seems to be overly optimistic, and the result may be not reliable.
In my opinion, the text itself could be suitable for publication after moderate/major revision, in particular if my doubts about the presented model (and about the used method of attributing the lithology and detection of discontinuities in a smooth tomographic model) are justified and if, in consequence, the model itself requires corrections.
DETAILED COMMENTS:
3-D model:
The model consists of several 2-D surfaces, representing intra-crustal discontinuities, that separate crustal layers. The layers, if we generalize as in Fig. 8 are, in the order from top to bottom: quaternary deposits, carbonates, metasandstones, felsic (‘gneissic’) upper crust, mafic (‘gabbroic’ ) lower crust, mantle.
The problem – visible in Fig. 8 - is that the shallower discontinuities should lie always above the deeper ones (I guess), while, in some places of the model, they intersect each other, therefore, the discontinuities violate the layer order. For example, the top of carbonates (8b), reaches ~10 km depth in the central part, which is well below (by ~5 km) the top of metasandstones (2c), theoretically underlying the carbonates. So, in fact, these interfaces intersect, and such inconsistency makes the model hardly usable.
This inconsistency is not visible in Fig. 11 (validation of the model), because all 5 profiles used for validation omit the regions where top of carbonated reaches this large depths. This problem could be visible in Fig. 12, because here, the model is shown along CROP-4 profile, which passes through the problematic area, but in Fig 12 only deeper discontinuities are plotted, without the problematic top of carbonates.
If my observations are correct, the authors have to check all the surfaces for intersections, and resulting negative thicknesses of the layers.
Text:
1. Idea of defining the discontinuities based on tomo Vp velocities exceeding the Vp values (Table 2) typical for given lithology (explained in Fig 7):
Generally, proposing meaningful locations of velocity discontinuities based on a smooth tomographic model can be problematic. This strategy may work relatively well at depths where, in the tomo model, Vp changes substantially with depth, eg. 4-8 km and the top of ‘metasandstones’ or ‘gneisses’ . However, at depths of 20-30 km, where the Vp smoothly and increases with a low, constant gradient only finding the top of the lower crust by looking for the depth where Vp exceeds some threshold value (here, arbitrarily set to 6.88 km/s (gabbro from Table 2)) results in huge uncertainty. First, with low Vp gradient and smooth model, using only slightly different value may change a lot the resulting depth of interface. Second, the range of lithologies and Vp in lower crust may be wide – if, instead of gabbro with Vp of 6.88 km/s, we assume LC with granulitic composition and Vp of 6.40 km/s, the depth of the LC top, detected in a way proposed by the authors, will be substantially different. I think the authors should discuss in more detail what will be effect of assuming a Vp thresholds with some (realistically) minimum (6.4 km/s?) and maximum (7.0 km/s? more?) values. How will it change the position of a discontinuity?
The uncertainty of this boundary (and my doubts presented above) are particularly important as in Ch. 6 you draw tectonic inferences based on the depth of this interface:
‘In the western portions of the profiles, the modeled gabbroic unit lies very close to the Moho and falls within the ±4 km depth uncertainty associated with the Moho estimates of Kelemework et al. (2021a). This close correspondence is likely linked to crustal thinning and magmatic underplating associated with localized crustal delamination beneath the Tyrrhenian domain.’
And in Discussion, you present this boundary as one of most important outcomes od the study:
‘The model also identifies a laterally continuous felsic-to-mafic crustal transition represented by the gabbroic interface, providing an additional crustal boundary that may be useful for future geological and geophysical investigations. The vertical relationship between this interface and the Moho discontinuity varies across the study area and constitutes one of the main structural features documented by the released dataset.’
In my opinion, in the light of large non-uniqueness of its depth determination, and strong dependence on assumed LC lithology, such final statements are not well justified.
2) The extent of the input tomographic model is quite wide. Why the authors cut the 3-D model area to be just the land part + 10 km buffer? Why not to make the model similar area, but with rectangular shape, encompassing the same land fragment, but including marine parts? Of course, marine parts may lack some data, and data for validations. But it may be more useful, and limiting model to borders of the land is quite arbitrary.
Minor text corrections:
Lines 59-61 versus 102-106 and 139-140: “Permo-Triassic crystalline basement” conflicts with the distinction between Permo-Triassic siliciclastics and Hercynian basement.
Lines 142-144: repair the incomplete sentence beginning “This resulting reference simplified stratification.”
Table 1, lines 165-167: remove the duplicated opening caption sentence
Lines 190-193: You write here that aggregation was used (which statistic?), while in SN1 you write linear interpolation was used.
Line 227: the reference to Figure 2 for the resampled Moho is incorrect.
Line 295, Fig 5 ‘coefficients of the best-fit linear regressions reported in panels (a) and (b)’ – They are NOT shown
Line 345: fix “Kelemenwork.”
Section numbering: 3.9 appears before 3.8. Renumber and update cross-references.
Lines 413-418: remove the repeated geothermal-gradient statement and use “Brocher (2005),” a single-author work
Figures 6-7: depth-axis units are kilometers, not the meters, as written now.
Lines 463+ :
The petrophysical approach was then extended to the full 3D model domain, assigning each cell to a lithological unit based on the corrected vp ranges. This procedure allowed the reconstruction of intra-crustal surfaces corresponding to major lithological transitions
AND Lines 470+ :
Building on the 1D application, the methodology was extended to the 3D tomographic volume to delineate the main crustal boundaries across the study area and reconstruct the surfaces representing key intra-crustal discontinuities.
-there fragments say basically the same, remove one, or combine them.
Lines 489-490: not all three numbered structures outcrop; [2] is described as buried earlier in the manuscript.
Fig. 3, Fig. 12: ‘Hercynian crystalline basement’ – seems to mean ‘upper crust (UC)’. Then. ‘Crystalline basement bottom’ in Fig. 12 – is it the bottom of the UC, or bottom of the crust (Moho)? From SupplNote 3 ,it seems that it is Moho, so why not to call it just Moho for clarity?