“Lower Carboniferous igneous intrusions within the crystalline basement of the Baltic Basin (SW edge of the East European Craton, Poland) – insight based on seismic data interpretation and seismic forward modelling”.  Krzywiec et al.

Seismological imaging of igneous intrusions within the continental crust is an area of extensive, global, continuing research. Much of this effort globally has used passive seismic methods applied to teleseismic arrivals. The data from deep seismic reflection profiling, much of it legacy, is an under-used resource and so I welcome this contribution by Krzywiec and co-workers. It is an interesting contribution that interprets and models results of deep seismic reflection profiling in Poland (PolandSPAN), focussing on discontinuous mid-crustal reflection packages.  The authors have an opportunity here to make this study far more useful to the global community of earth scientists by drawing out the general application of their methods and making broader comparisons with other examples. It is this recommendation that underpins many of the comments in this review. As written, the paper tends to the parochial – the authors have a great opportunity here to do something more! So while this could be published with only minor modifications, I encourage the authors to reframe the text in this light.

The paper is clearly within the scope of SolidEarth, is generally well-written and illustrated and has high scientific quality.

Given then opportunities noted above, it would be useful to clarify the aims of this contribution. Of course, large igneous intrusions are exposed at outcrop, exhumed from various levels (including deep). The point here is detecting them in the subsurface, potentially linking them to coeval extrusives at outcrop. But why does this have importance for understanding anything other than the evolution of some local continental crust (i.e. beneath Poland)? As noted in the geological setting… the region has igneous rocks at outcrop. Is the point here to establish their extent at depth? If yes, say so! Or is it about establishing the geometry of intrusions? Do sills emplaced in the basement have the same propensity for steps and splays as they do when emplaced in layered sedimentary rocks, nearer the syn-volcanic earth’s surface? The abstract does a better job at setting the scene. Likewise, the general importance of this study, the opportunities provided by deep seismic reflection profiling, the methods used here, and the general interpretations could be developed much more fully in the discussion.

 Analysis of the tuning thickness is used to determine the seismic velocity and this inputs to modelling of synthetic seismic responses, the better to characterise the shape and velocity structure of these subsurface bodies.  These build from extensive discussion of the impact of tuning thickness on imaging. This reads well generally and is well-paced for non-seismologists. This allows the thickness of sills and their geometry to be made, together with deductions made of their velocity structure (and hence composition). The overall seismology and deductions seem convincing (though I’m not an expert!). I do have issues with the justifications and descriptions of methods and results which could be tightened up by better (and possibly more relevant) referencing. These are outlined below.

At various points in the text, the authors’ approach is described as “educated guessing”. While this is very self-effacing, it is of course common to all interpretation – which naturally always carries uncertainty and assumptions. It would be helpful to be explicit about the limitations imposed by the specific assumptions adopted here – a point that could be returned to in the Discussion.

Section 3.1

It’s good to see the information on the acquisition and processing parameters for the PolandSPAN survey. Presumably it is the quality of these that has allowed the detailed work here (as the authors say, these are unparalleled). But if the authors choose to broaden the vision of this paper to promote the greater use of legacy surveys elsewhere – it would be worth comparing the PolandSPAN parameters with other, historical, onshore surveys (ECORS, FIRE, various Iberian experiments etc). Or address this in the discussion.

Section 3.3.

Introductory paragraph (lines 181-187). The arguments for adopting synthetics rather than ray-tracing methods needs referencing to justify these statements – especially as the only citation (Slonka et al) is to a study of sedimentary rocks rather than as a tool to enhance crustal imaging.

3.3.1

This discusses the selection of modelling parameters. It would be helpful to state explicitly the values (and their ranges) as used in this study (velocities, densities). And be specific in the comparisons made with the geology (end sentences of this section).

Section 4

Line 253 – why was 500 ms used as the wavelet extraction window? What are the implications of changing this?

Line 260-2. Please be explicit that the tuning thickness chart is Fig. 6c (it’s in the caption but it makes the text easier to follow if in there too). Maybe note that tuning thickness charts are also provided for Fig 7 and 8.

2-D seismic modelling

It’s good to contrast acid/intermediate composition (granite-granodiorite) with basic-composition igneous rocks. However, it’s curious to use parameters from both basalt and dolerite. Presumably the basalt ones relate explicitly to lavas while dolerite are for intrusive rocks (the difference not being compositional but textural – as basalt may have porosity). Surely only dolerite is relevant here (e.g. rhyolite hasn’t been modelled along with granite).  

But more critically – given the thickness of the interpreted igneous bodies (up to 200m for the wedge approximations) – these are likely to be layered – and could include ultrabasic compositions… (And indeed the basic component is likely gabbro). So over-emphasising specific values on seismic velocity and density is un-necessary – beyond eliminating those for acid/intermediate composition rocks. Simply setting this up as a two-comparison acid/intermediate vs basic (dolerite/gabbro) test – would suffice.

I’d expect to see justification for the host rock parameters (“Precambrian basement”) – as of course it is the contrast with this that is important for investigating the intrusions.

This aside apart – the modelling section reads well. The illustrations are nice (Figs. 10 and 12, 11 not needed if you drop the basalt!).

Section 4.4

(line 430 onwards)

The comparisons here with interpretations of igneous intrusions in the crust revealed by seismic imaging could be enlarged and better-structured – though may be better reserved for the Discussion.

The wedge model and the deductions of the thicknesses of the sill, together with its steps – are great (Figs. 14, 15, 16).  Although the authors have chosen to reserve further comparison with outcrop “analogues”, choosing only to compare with seismic examples of Cartwright et al. – I think again that this is a missed opportunity. Stepping and branch sills have been known and illustrated in the geological record for over 200 years (yes, really – Macculloch 1819) … which are well-known to the igneous geology community… Simply citing work from the 2000s onwards isn’t great!

The Discussion focusses on Poland and its geology. As noted above, this would be far more valuable if these discussions were broadened out. The final paragraph hints at this. More discussions should be made about the underpinning assumptions

Seismic Interpretations

The seismic interpretations (figs 3, 4, 5) should include uninterpreted (clean) images.

As presented – the images have very high vertical exaggerations. Is the deep normal fault in Fig 5 really dipping at just 25 degrees (as implied by this interpretation when at ~v=h…)?

Figure 1 – please change the colours between extrusive and intrusive igneous rocks – they’re too similar at the moment.

Rob Butler

Aberdeen

October 2025

Review of Lower Carboniferous igneous intrusions  by Kryzwiec et al. 

This paper presents some new results on intrabasement features recently revealed by deep reflection profiling in Poland. These results should be of interest across a rather broad range of geosciences disciplines.

In general the paper is well written with only a few grammatical glitches. The discussion of the geological setting is relatively clear and thorough.  The figures are generally of excellent quality. However the figures purporting to show the results of the seismic wavelet analysis generally lacking sufficiently detailed depictions of the actual data that are needed to evaluate the validity of some of the modeling conclusions.

My biggest criticism is that the discussion of the wavelet analysis, intended to reveal primarily the thickness of the interpreted intrusions is presented in a confusing manner, and lacks any clear comparison of the modeled waveforms with the observed waveforms on a  common scale that allows for the assessment of the modeling results. Moreover important details of the analysis are not explained. 

For example, a first step in the analysis as described by the authors as “wavelet extractions” within a window encompassing the deep reflections of interest. What does this mean? Deconvolution? How can the wavelets be “extracted” without making some a priori assumption of their waveform? Since the analysis all about interpreting these waveforms, how were they extracted without altering them. Moreover, to what extent are the wavelets of interest “contaminated” by multiples generate in the overlying sedimentary sections Was deconvolution use, and could such deconvolution have distorted the intrusive reflection wavelets?

A related concern is modeling the waveforms using the Ricker wavelet. The Ricker wavelet is usually associated with impulsive sources such as explosives. However the data shown here is implicitly from Vibroseis sources, although that is never explicitly stated in the paper. Vibroseis correlationresults in a Klauder wavelet. How do the authors justify using the Ricker wavelet for modeling of what are presumably Klauder wavelets in the actual data? Were the data subjected to a Klauder to Ricker transform.

The modeling of the data appears to consist of three parts. The first is essential 1D modeling to determine the thickness of the presumed intrusion layer by matching amplitudes peaks and wavelet peak to trough measurements of real data with synthetics for a simple hgh impedance layer. However there is no  explicitly explaination of  how the apparent peak to drought time measurement of the composition waveform relates to the  impedance model. Only once the wedge modeling is subsequently described is this relation depicted graphically, though still not explicitly explained. Moreover, is not clear but it seems that the authors are trying to use an average waveform for the intrusion response in a large window without considering possible later various in intrusion thickness (although lateral variations are address later in the analysis. The authors should be clear on this point.

In fact the modeling effort would be much clearer if they started with the wedge modeling discussion,  which shows how distinct reflections from the top and bottom of a thin layer “merge” as the layer thins. This graphic could be used to clearly indicate what the various parameters are in the  graphics in the previous modeling section (e.g. actual time travel time difference, observed peak to trough time, maximum amplitude and resonance frequency etc.). 

In none of the discussions do the authors  show  their modeled waveforms next to actual recorded data at the same scale and in the same display format. The “zoomed” versions of the color section in Figures 6-8 simply to not show the observed waves in sufficient detail nor in a comparably format (eg.  Variable area wiggle) to make in useful judgment of how well the modeling matches the observations. 

Also the color scale wavelets appear to exhibit may more Iwiggles (side lobes?) than explained by the modeling. Is this an artifact of displaying the modeling and data with different display modes or gain functions (eg. AGC)?  Does such “additional“ wavelet complexity indicate finer layering with the intrusion, or do they represent of multiples generated in the sedimentary column above. Were such multiples multiples removed  by processing (deconvolution? wavelet extraction?)?

In any case, at least a sample of the data needs to be shown in the same wiggle format and at the same display scale as the modeling. And these more zoomed versions of data should show your picks for peaks and troughs used to deduced thickness.

The second part of their analysis they use full 2D modern of the complete geological column including the sedimentary section for several different proposed lithologies for the intrusion in an attempt to see which seems to match the relative amplitudes of the intrusions compared with the shallower reflections. The result is rather predictable:  the model with the largest impedance contrast the between intrusion and host rock gives the strongest reflection. Using the overall appearance of the seismic section to demonstrate this is rather qualitative. A more quantitative approach would be to to compare amplitude vs travel time curves for traces that sample the intrusive. Also any amplitude analysis begs the question of what attenuation values were used for the geologies traversed by the seismic waves. 

As mention earlier the wedge modeling in the next section of the paper clarifies some of the ambiguities in the discussion of the earlier modeling and should precede the other waveform analyses. The quantification of lateral variations in thickness is one of the most outstanding contributions of this study. However here too the presentation would have been more effective by zooming into the wavelets to make their details clearer. Also, could amplitude processing (eg lateral variations an spherical divergence corrections) have distorted the intrusion election amplidues?

Lastly the authors attribute all of the lateral waveform variations to changes in intrusion thickness, but give no consideration to the a possible alternatives: lateral variations in composition of either intrusions or country rock, or the possibility of finer layering with the intrusion.

However, in summary the authors present some beautiful new seismic data depicting likely intrusions in the sub-sedimentary crust. However the discussion of the waveform analyses could be improved for clarity of technique and results and completeness of interpretational possibilities.