Deep crustal structure of the southern Baltic Sea in the light of seismic and potential field data

Ponikowska, Małgorzata; Stovba, Sergiy Mykolayovych; Malinowski, Michał; Nguyen, Quang; Mazur, Stanisław

doi:https://doi.org/10.5194/egusphere-2025-3107

Preprints

https://doi.org/10.5194/egusphere-2025-3107

Preprints

27 Aug 2025

| 27 Aug 2025

Deep crustal structure of the southern Baltic Sea in the light of seismic and potential field data

Małgorzata Ponikowska, Sergiy Mykolayovych Stovba, Michał Malinowski, Quang Nguyen, and Stanisław Mazur

Abstract. The southern Baltic Sea lies within a critical transitional zone between two major geological provinces of Europe: the Precambrian East European Platform and the Palaeozoic Platform of Western Europe. While the shallow expression of this boundary is generally marked by the Caledonian deformation front, the deeper crustal configuration remains contentious due to thick Phanerozoic cover. This study integrates seismic interpretation with 2-D gravity and magnetic modelling to investigate the deep crustal architecture beneath the southern Baltic Sea. Four new seismic profiles (BGR16-256, BGR16-202, BGR16-257, BGR16-259), acquired during the BalTec (MSM52) expedition, were analysed alongside borehole and legacy seismic data. Seismic imaging reveals that the upper crust was primarily shaped by Permian–Mesozoic extension and Late Cretaceous inversion. Extensional basins such as the Mid-Polish Trough and Rønne Graben accumulated up to 4 km of sediments, later uplifted and folded during inversion, which caused displacements of 1.5–2 km and produced asymmetrical marginal troughs with NE-directed compressional vergence. The gravity and magnetic models, constrained by seismic horizons, enable imaging of deeper crustal levels including the top of the lower crust and the Moho, which lies between 38 and 42 km depth. These data reveal that thick Baltica-type crust extends south-westward beyond the Teisseyre-Tornquist Zone, contradicting interpretations that propose a sharp lithospheric boundary along this zone. A key finding is the identification of a NE–SW-trending crustal lineament, likely inherited from Precambrian lithospheric fabric. Furthermore, evidence of pre-Triassic tilting and erosion of Silurian strata suggests a significant tectonic event, possibly related to early Carboniferous uplift. The combined data provide new insights into the complex tectonic evolution of the region, supporting a model of Baltica crustal affinity beneath the southern Baltic Sea and emphasising the interplay of inherited Precambrian structures, Permian-Mesozoic extension, and Late Cretaceous inversion.

Received: 30 Jun 2025 – Discussion started: 27 Aug 2025

Competing interests: At least one of the (co-)authors is a member of the editorial board of Solid Earth.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Małgorzata Ponikowska, Sergiy Mykolayovych Stovba, Michał Malinowski, Quang Nguyen, and Stanisław Mazur

Status: final response (author comments only)

RC1:
'Comment on egusphere-2025-3107', Anonymous Referee #1, 25 Sep 2025
This is a well-written and, generally, well-executed study of additional shallow seismic lines in the southern Baltic Sea, supplemented by gravity and magnetic modeling. The results support a complex transition zone between the East European Craton and Phanerozoic Europe — a concept that appears to have increasingly gained traction in recent years.
While I generally follow the seismic interpretation and understand the technical aspects of the potential field modeling, I have some concerns regarding the gravity and magnetic models. Some of these concerns may possibly be resolved by adding appropriate clarification or documentation; others may require some more significant changes. The main issue is how the top lower crust and Moho depth are constrained, which is not properly documented. Furthermore, the general modelling approach is unclear (what was the basis for the decision-making of which data to fit and which structures to perturb). One of the four analyzed profiles in this paper is assigned a different density in the lower crust (Why?), and all densities appear to be inconsistent with those published in Ponikowska et al. (2024), which are crossing the profiles. All of these profiles are later used together in the same inversion scheme. As a result, I argue that the Moho depths across the various profiles are not directly comparable due to these inconsistencies. Additionally, I would expect to see better integration with other seismic lines in the region and a more in-depth discussion of why different Precambrian terranes may have responded differently to rifting. It would also be helpful to integrate and discuss how this affects our understanding of the STZ/TTZ — or more broadly, the margin of the EEC from the surface to the base of the lithosphere.
General Comments
Gravity/Magnetic Modeling Approach:
I have several questions about the overall modeling approach. The paper provides a general explanation of how gravity and magnetic modeling were performed and work, but does not describe the specific strategy adopted in this study. It’s also unclear what constraints were applied and how/to what degree — beyond the shallow seismic reflection lines, there should be other, sometimes deep adjacent seismic lines, and how the parameters or parameter ranges were selected.
As far as I can tell, only one profile (PQ2-002) has a more or less direct deep control from a nearby deep seismic line throughout much of the profile length, but this constraint does not appear to be visibly incorporated. The two profiles differ significantly and have inconsistent modeling parameters for gravity and magnetic modelling. The other lines do not have coincident deep seismic coverage, though some deep seismic lines intersect at 1–3 locations. It remains unclear whether these were actually used as constraints or not. If not, why were they excluded, especially given that such intersections could serve as important constraints in the otherwise extremely unconstrained potential field modelling?

This leads to another point: with little constraint on deep crustal structure and the simplifying assumption of homogeneous layers (upper+middle crust, lower crust, upper mantle), it becomes difficult to determine whether gravity anomalies originate from the upper–lower crust interface or the Moho (or density variations in between, which of course are neglected here because of the assumptions made). For instance, in Figure 7, the text explains that some of the gravity misfit is related to “uplifts in the lower crust.” I don’t follow this. How are these uplifts constrained? As far as I can tell, they are purely modelling features, and the gravity signal is still not well-matched, after all. From a gravity perspective, these anomalies could just as well stem from Moho undulations or features in the upper crust. A feature at the upper–lower crust boundary might explain the magnetic data, which would not be possible with a varying Moho — but this rationale isn’t laid out clearly. For each seismic line, the workflow for building and refining the model should be better explained. Which dataset/interface was given priority and why?

I recognize that in a poorly constrained region, modeling requires assumptions. One valid strategy is to apply a minimum number of changes to the starting model. Another approach is to prioritize either gravity or magnetic data and treat the other as secondary. In this study, it seems that fitting the magnetic data was prioritized. This should be clearly stated. In many places, structural adjustments appear to be driven primarily by the magnetic response, resulting in gravity misfits. That’s acceptable, but the fit could be otherwise improved by adjusting Moho depth — a step that doesn’t appear to have been taken.

Finally, the large differences primarily between the density structures in this paper and those in Ponikowska et al. (2024) are concerning. The density differences are often significant, and such discrepancies should lead to overall different Moho depth levels. The authors should again at least explain/discuss this line of thought. To help evaluate this, the lower crust and Moho interfaces from the crossing lines should be shown in the profiles for comparison. The inversion process uses a constant starting Moho density contrast of 0.4 g/cm³, but values within the profiles used range from 0.33–0.4 g/cm³ (as far as I can tell). While these values may be adjusted during inversion, this variation should affect the Moho depth level and thereby introduce biases.

Other Seismic Lines in the Study Area
In discussing and comparing crustal structure, other seismic lines beyond Ponikowska et al. (2024) may be discussed in more detail — potentially PL1-5600, PQ2-91, BASIN9601, the BalTec refraction line, and the BABEL lines. These lines intersect or approach the modeled area and are notably absent from the interpolated models of top basement, Moho depth, and crustal thickness.
EEC Margin, STZ–TTZ, Trans-European Suture Zone
For a complete discussion of the EEC boundary's width and complexity, the authors should include the most recent literature — including work by some co-authors of this paper — and extend the discussion into the sub-crustal domain. Consider incorporating the Sorgenfrei–Tornquist Zone and the so-called “Tornquist Fan,” which represents an equally complex transition. This is critical, as lithosphere-scale processes can reveal structural complexity that’s not visible in crustal data alone.
Other General Comments
Figures 6–9: Please add the Moho depth interpretations, and perhaps also the density values from crossing deep seismic lines (Ponikowska et al. 2024 and others) for direct comparison.

Figure numbering: Starting from section 5.10, figure numbers appear misaligned — perhaps all figures are one number too high? For example, Fig. 12 and Fig. 13 are both labeled “Depth-to-basement” but clearly show something different. Fig. 13 probably shows crystalline crust thickness.

Uncertainty: There is no mention of uncertainty or resolution in the final models. This is especially important in the inversions. Forward models are one thing, but inversions typically provide at least some uncertainty estimates — this is missing here.

Detailed Comments (with possible repetition from general comments):
L22–23: A more detailed discussion of how different Precambrian basement blocks may have reactivated to form this “kink” would be valuable. (not necessarily here)

L34–36: Consider also discussing the mantle lithosphere, which remains poorly understood.

L39–44: Several additional seismic lines (PL1-5600, BASIN9601, BABEL) have been published or re-evaluated recently and should be considered.

L50–51: It would be helpful to mention here how the interpretations were constrained.

L52: “Constrained by potential field models” is vague. What exactly is being constrained, and how? Are the 2D forward models used as constraints? Perhaps to better call these a model input, as these 2D lines are also a product of relatively unconstrained potential field modeling.

L54–56: How was the “improved spatial resolution” assessed? I don’t see any uncertainty/resolution analysis.

L84: Repeating an earlier point — include more recent and ‘diverse’ literature, for example, on the mantle lithosphere (e.g., tomography).

L88: Please clarify why the strike-slip model doesn’t align with the apparent STZ–TTZ connection.

L94: Unclear reference — what does “It” refer to?

L102: The lithospheric thickness change should be discussed for the TTZ as well.

L111: Should this be “The crustal keel has been…” or “Crustal keels have been…”?

L126: The “bend” of the CDF toward the TTZ is hard to locate — please clarify.

L213: Densities are said to be derived from seismic velocities — but which ones/which velocity models? Please specify. Also, why does only BGR16-202 have a different lower crustal density? It intersects BGR16-212 (Ponikowska et al. 2024), which has higher values. This inconsistency is not addressed, and similar discrepancies exist with BGR16-259 and other profiles crossing PQ2-002 and PQ2-004-005.

L217: The Moho and top of lower crust aren’t really primary — they’re in fact the only interfaces, if I understand right. How was the top of the lower crust chosen? It seems as unconstrained as other interfaces. This modeling rationale needs documentation.

L218: Again, which seismic velocities were used for density conversion?

L237: More detail is needed on how previous studies were used to constrain the initial top lower crust and Moho horizons. Were they interpolated? Were intersections with deep lines used? Please include these surfaces in the 2D profiles.

L249: Why weren’t other lines like BABEL, BASIN9601, POLONAISE, PL1-5600 used, even though they cross the area?

L294: Maybe refer to Fig. 2 for the Ustka Fault. It can be traced to PQ2-004-005, but not clearly further east (BGR16-212 and PL1 5600) – why?

Figure 5: “Cr” is used for Cretaceous; should be “K.”

L356: Cretaceous inversion structures are mentioned but not described. There is evidence — please elaborate.

L371: I might get it wrong, but I have the impression that it’s the only profile to the south/west of the fault.

L398: Perhaps rewrite: “in the first 20 km of the profile.” Also, please mark the Kolobrzeg Anticline or refer to the figure where it is shown.

L430: Is top basement not well constrained? Instead, it probably indicates heterogeneity in the (upper) crust. Furthermore, the observed gravity anomaly shows the opposite of what is expected – a low where there should be a high above the shallow basement. The modelled gravity shows the exact opposite behavior.

L432: This statement is confusing. If misfits align with lower crust uplifts, are these independently constrained features or just model features? Could the misfit be diminished by introducing Moho variations instead? The modeling approach seems to prioritize fitting magnetics. That’s fine, but if magnetics allow for lateral susceptibility variation, this could explain some of the misfits, instead of changing the top lower crust too much. Moho adjustments should not affect magnetics but could improve gravity fit. Why were they not attempted?

To summarise: What exactly was the modeling approach? How were the Moho and lower crust interfaces defined and adjusted? Was priority given to gravity or magnetics, and why? Why were Moho depths not perturbed when unconstrained? Many long-wavelength gravity anomalies could be addressed with Moho changes. The approach may be fine, but it needs to be explicitly laid out.
L459–460: Saying “where the model predicts a basement uplift” is misleading — the uplift is part of the modelling, used to fit geophysical data.

L484: “…and density” is unclear — the lower crustal density is fixed.

L505: Figure numbering seems off.

L523: Should this be FA instead of BA? Probably refers to Fig. 3a. Also, what would be expected from the BA or FA here?

L524–525: Please elaborate. Why would later tectonic activity explain this?

L525: The ENE–WSW trend also appears in the FA (Fig. 3b).

L548–549: The “depression west of Bornholm” isn’t clear — possibly meant east? Please also clarify where exactly the Moho uplift is beneath the NW Mid-Polish Trough.

L598: Indicate where this fault lies (I think ~100–105 km). The basement appears to outcrop beneath the Cenozoic, so a fault still seems possible.

L603–604: The phrase “westward increase in thickness of the Caledonian accretionary wedge” is confusing. It seems to thicken eastward — please clarify.

L616: Expand the TTZ and EEC margin discussion using other results covering neighboring areas and incorporating mantle lithosphere structure.

L635: The abrupt 90-degree shift in the thick crust boundary is notable and contrasts with the STZ–TTZ trend. A more detailed discussion of possible controlling factors (e.g., rheology, stress regime, inheritance) would be valuable.
Citation: https://doi.org/10.5194/egusphere-2025-3107-RC1
RC2:
'Comment on egusphere-2025-3107', Anonymous Referee #2, 01 Oct 2025
This study presents four new deep seismic lines from the Baltic Sea and interprets them in the broader tectonic context of the southwestern Baltic Sea. New fault and trends are described. In addition, potential field modelling is carried out along the 2D profiles in addition to gravity inversions to derive regional maps of the top basement and Moho depth.
This paper provides new insights and a valuable contribution to the discussion on the crustal nature and tectonic history of the southwestern Baltic Sea. It is well written and structured and addresses the necessary background information and assumptions. The seismic data presented are of very high quality (although the reader does not get to see them in full resolution) and the figures are overall clear. The methods are well presented and the scientific results sound, though I think that some results need to be more carefully presented and discussed. In particular the derived grids of top basement and Moho thickness need a better evaluation. I have listed 5 major points that I think need to be addressed. In addition, I am attaching an annotated manuscript with several minor comments.
Seismic horizons
It is often not clear how the geological units are assigned to the interpreted seismic reflectors and horizons. I don’t think the interpretations need to be modified, but the reader needs to get a better understanding where these interpretations come from. The few boreholes shown are hardly sufficient to interpret the entire seismic sections. I suggest to either refer to previously interpreted seismic profiles from the region and/or to show close-ups of the seismic characteristics of the geological units to let the reader understand how the interpretation was guided.

A larger, non-interpreted version of the seismic profiles would be much appreciated by the inclined reader. Could these be added as an appendix?

2D Potential field modelling
On profile BGR16-202 (Figure 7), the modelled gravity signal clearly does not fit the observed one. Instead of arguing for too smooth gravity data, I would invoke 3D structures and a perhaps unlucky positioning of the profile. This profile runs roughly parallel to several basement folds (which profile BGR16-256 crosses perpendicular). Hence, basement offsets to the left and right of the profile may be present and would be reflected in the gravity signal but not in the 2D seismic line. Here, a 3D model would be necessary to build an accurate gravity model.

It is not clear how the top and base of the lower crust are defined in the 2D models. There may be very large uncertainties here, I assume. These must be discussed. The magnetic data seems overfitted. Here, numerous crustal blocks are used. Some of the magnetic trends (long-wavelength signal) could perhaps be modelled with the upper crustal thickness. The long-wavelength signal of the gravity signal should be provided by the Moho. Perhaps these could explain some of the large residuals at the model ends?

Moho & Top-basement grids
The process to derive a regional Moho and top-basement grid is described in detail but it leaves some doubts whether one can invert the same dataset once for a Moho horizon and once for a top basement horizon without doubly interpreting some of the gravity anomalies. If I understood it correctly, it was not the residual of the Moho inversion that was used for the top-basement inversion, neither the original gravity anomalies, but something in between. Be that as it may, the paper needs a proper assessment of the Moho (and top basement) uncertainties. This needs to be given (a) for the 2D models and (b) for the regional grids. In addition, there needs to be a discussion on how the new grids differ from previously published grids. Are the tectonic implications derived from the Moho and crustal thickness maps consistent with previously published maps?

Data availability
Will the new data (i.e. the top basement and Moho grids) be available in a data repository?
Citation: https://doi.org/10.5194/egusphere-2025-3107-RC2
EC1: 'Comment on egusphere-2025-3107', Christopher Juhlin, 09 Oct 2025

Dear Authors,
You have received two thorough reviews of your manuscript and both reviewers are positive. However, they also request greater clarity in the presentation of the data and interpretation. Please provide responses to the two reviews on a point by point basis. I agree with reviewer #1 that it would be good with uninterpreted versions of the profiles. These can be included in a supplement.
Best Regards,
Chris Juhlin, Guest editor

Citation: https://doi.org/10.5194/egusphere-2025-3107-EC1

Małgorzata Ponikowska, Sergiy Mykolayovych Stovba, Michał Malinowski, Quang Nguyen, and Stanisław Mazur

Viewed

Total article views: 1,361 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
1,297	46	18	1,361	60	65

HTML: 1,297
PDF: 46
XML: 18
Total: 1,361
BibTeX: 60
EndNote: 65

Views and downloads (calculated since 27 Aug 2025)

Month	HTML	PDF	XML	Total
Aug 2025	187	5	3	195
Sep 2025	1,008	18	5	1,031
Oct 2025	102	23	10	135

Cumulative views and downloads (calculated since 27 Aug 2025)

Month	HTML	PDF	XML	Total
Aug 2025	187	5	3	195
Sep 2025	1,008	18	5	1,031
Oct 2025	102	23	10	135

Viewed (geographical distribution)

Total article views: 1,354 (including HTML, PDF, and XML) Thereof 1,354 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 29 Oct 2025


Total:	0
HTML:	0
PDF:	0
XML:	0