the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 10 Jun 2025
Petrophysics-guided reprocessing of legacy seismic data to improve mineral exploration targeting in the Irish Zn-Pb Orefield
Abstract. The Limerick Syncline, part of the Irish Zn-Pb Orefield in southwest Ireland, represents a geologically complex and relatively underexplored region, despite hosting several significant mineral deposits, e.g., Stonepark and Pallas Green. The mineralogy and stratigraphic setting of the sulphide mineralisation in the Limerick Syncline are generally similar to other Irish-type deposits. However, a thick volcanic sequence overlies and interfingers with the carbonate host rocks, mineralisation and alteration. This setting has posed significant challenges to seismic imaging in the region, despite being a powerful technique for mineral exploration. As a result, the overall structural setting of the area has remained poorly understood. This study presents an optimised seismic processing workflow tailored to these geological complexities and applied to the legacy 2D seismic reflection profile LK-11-02. The workflow integrates newly acquired downhole and laboratory P-wave velocity data with first-arrival travel-time tomography to enhance the accuracy of the velocity model used for post-stack migration. This resulted in better signal recovery and enhanced reflector coherence, in particular, reflection continuity. As a result, imaging of key stratigraphic boundaries and carbonate-volcanic interplay geometries was enhanced, along with clearer reflection amplitude changes associated with known stratigraphic units. Acoustic impedance analysis using laboratory density data enabled a better understanding of the origins of seismic reflectivity and a more confident geological interpretation of the laterally variable lithologies. The migrated section reveals previously unrecognised major structural features in the Limerick Syncline, enhancing our geological understanding of the region. It highlights the potential of applying this workflow to seismic reflection methods, improving its use as a mineral exploration tool in the Irish Zn-Pb Orefield and similar volcanic-influenced regions.
- Preprint
(2452 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-2512', Samuel Zappalá, 13 Jun 2025
Dear authors,
the proposed manuscript is well written and the proposed idea interesting and useful, but I have some questions and concerns regarding the innovation of the work and the evaluation of some of the proposed solutions that need to be addressed. You can find them in the following lines together with some smaller suggestions to increase the clarity of the text and of the figures. I hope my comments will help you to improve the clarity of your work.
The abstract, introduction and geology are very well written and to the point. The rest of the text will benefit of some shortening, especially the velocity analysis and the discussion chapters are too long and present details that are not useful with some repetition. The velocity/lithologies comparison appears too detailed considering that is not showing any strong evidence and it is not used on the interpretation or the results, I would remove most of its description (comprised figure 5) and talk about the useful results in just 1-2 phrases.
From the manuscript appears that the main change you applied in the seismic processing is the velocity analysis that follow this scheme: the petrophysical velocity analysis from the cores is validated using the two available sonic logs. The starting model for the tomography is based on the petrophysical velocity analysis from the cores. The stacking velocity model is computed with semblance analysis and corrected according to the tomography model. If I understood correctly, a structured summary of these steps will help the reader to better understand it.
Related questions:
(1) the petrophysical velocity analysis from the cores matches only partly the sonic logs with a big number of outliers (see figure 4b after 300 m of depth or the low measurements coherency in figure 4a), you justify this behavior in the text with understandable reasons, but how do you handle the problem? Or you just ignore it?
(2) have you tested different starting model for the tomography (simpler or more complex)? Did you remove the petrophysical constrains to evaluate if they are really influencing the final tomography results? did you evaluate how much it is sensible to the inaccuracies in the velocities from the core samples?
(3) why did you not use a smoothed version of the tomography model for stacking the data (at least for the shallow depth) instead than computing another velocity model with semblance analysis?
(4) did you try to use constant velocity stack analysis instead than semblance? It is usually effective in hardrock domain but I am not sure what will be better in this mixed case.
(5) it is not clear what velocity model you are using for the migration, is it exactly the same one that you use for the stacking?
You should compare the migrated section resulting from your proposed approach with the one from the newest dataset (dataset 2) and not with the old one (dataset 1) for 2 main reasons: (1) Dataset 2 supposedly shows better results since it has a more recent processing and (2) dataset 2 is the starting data you used for your sections. It would be interesting and useful to compare also the velocity model used from the previous study that you use for comparison.
What you are comparing now (Figures 9a and b) just shows that two different processing returns two different results, nothing new in seismic. The differences between the two processing flows are too many to be able to state that they are due to the proposed velocity approach. They could as well be due to the pre-stack applied processing before your work, or to the different applied migration method and not related at all to the velocity analysis that you performed. So, this point (that I would say is the core of the manuscript) needs to be strengthened.
In the title you say “petrophysics-guided reprocessing” but only the velocity analysis is partially guided from the petrophysical data, everything else is a basic seismic processing flow.
Figure 6 needs some work to be clearer. In general, the figure is too small and is not possible to see what you describe in the text. (1) Boreholes are too thin and they have no frame making difficult to identify them when they overlap with the tomography section. (2) Some of them is plotted behind the tomography and overlap is not visible at all making impossible the comparison. (3) Do you need to plot the whole depth of the boreholes’ velocities? There is nothing to compare them below the tomography model and they do not give any useful information. (4) get rid of the single rays propagating deeper in the model, they are very confusing and not trustable. Once you clean up the figure you can add some zoom in window to highlight the details you are describing in the text. Use this figure to show the quality of your velocity model and how it benefitted from the petrophysics data.
In chapter 4.2 It is not clear how you measure the acoustic impedance, you talk in the text about acoustic impedance related to the different formations but how do you calculate it? Is it computed considering the interfaces for each borehole? Is it an average of all density/velocity values for those formations? Is it between each measured point? Explain it more in detail.
Technical corrections
Line 43 – It should be (Eaton et al., 2010).
Line 48 – It should be (Wilkinson et al., 2015).
Line 52 – what do you mean by “1000 line-km”?
Figure 1 – Very nice figure, just some comment. a) and b) should be bigger and bold, but you can actually remove them and just refer to Ireland map as an inset. You may also add a thin black frame to the all figure, but this is just an aesthetical thing. Modify the caption as “…showing the seismic lines, the drill holes…”. Highlight the seismic line that is used in the manuscript (i.e. write in red LK 11-02 on the figure).
Line 86 – It should be (Chew et al., 2009).
Figure 2 – It is not clear what is the transparent pink that cut through the other formations. Write it in the caption.
Along the manuscript use or the full names for the geological formations or their abbreviations and be consistent both on the text and in the figures/tables, now you are switching between them and this makes it difficult to follow it. If you want decide to use the abbreviations keep their full name only the first time (in the geological chapter).
Instead than table 1, a figure showing the boreholes lithological sections will add useful information and it will be easier to follow.
Line 145 – move the sonic log information on chapter 3.1
Line 158 – “Robertson Geo Mining & Minerals” should be a reference? I cannot see it in the reference list.
Line 177 – seismic data – give more information regarding the acquisition. How heavy the seismic vibrators are? What kind of sweep are they generating? What frequency range is used? What sampling rate is used? What maximum and average offset is recorded? How is the raw data quality?
Line 196 and table 2 - What type of velocity analysis has been performed in dataset 1 and 2?
Figure 4 – make the figure bigger or increase the font size of the text in the figure, it is too small to read. You are again using different names for the stratigraphic units, I would suggest to use the abbreviations.
Line 257 – it should be (Lees and Miller, 1995).
Line 276 – It should be (Sleeman et al., 1999).
Line 355 – I do not see these reflections observed between the CDPs 720 and 1130 and between 1130 and 1335, around 300 to 500 ms that you refer to. There is some discontinue reflection but hard to define them.
Figure 8 – Is it migrated or unmigrated? I cannot see many differences from the one in figure 9b.
Line 356 – I am not so confident about the R3 reflection on the north. It is clearly visible the northern part but it is not easy to follow it after CDP 250. You should move the second arrow closer.
Line 368 – You cannot refer to figure 9c before you refer to figure 9a (or to figure 9 in general).
Figure 9 – In the legend you write the name of the formations to indicate what is a horizon. Write if it is Top or Bottom of the formation.
Line 404 – I do not see any light green horizon in the figure, maybe the light blue.
Line 408 – There is any information about these faults in literature? If yes it will be useful to reference them.
Line 439 – Which dominant frequency are you considering in your calculations?
State clearly the uncertainty of your interpretation, especially for the reflections that are not intersected by the boreholes and for the faults.
Line 466 – it should be (Fusciardi et al., 2003).
Line 572 – de Morton 2015 (1) is not in the text.
Line 646 – You should distinguish the two Slezak 2023 references according to the authors guidelines.
Citation: https://doi.org/10.5194/egusphere-2025-2512-RC1 -
RC2: 'Comment on egusphere-2025-2512', Anonymous Referee #2, 11 Jul 2025
The submitted article presents the results of a 2D seismic survey ~10.9 km long acquired over the Limerick syncline in southwestern Ireland for mineral exploration purposes. The author(s) revised the seismic processing workflow, incorporating petrophysical knowledge (both laboratory and downhole wireline logging) from nearby boreholes and first-arrival traveltime tomography (FATT, P-wave). It provided enhanced subsurface images compared to the two previous attempts made by Velseis Processing Ltd. and Earth Seismic Processing Ltd (ESP). This helped the author(s) in better geological interpretation of the key lithological/stratigraphy units prone to mineralisation. A better understanding of the interfingering between the volcanic and carbonate rocks is also obtained. Two fault systems were delineated and interpreted as potential pathways for the mineral system, which contained metal-bearing hydrothermal fluids.
However, I have some major reservations about the results presented in the article. I hope the author(s) find my comments valuable and will help them further improve the article. I am providing the main points below. Minor comments are provided at the end.
- My first comment is related to the term “legacy data”. I have failed to understand the reasoning behind. To me, it is simply a 2D seismic data acquired in 2011 (so relatively recent), which the author(s) have used to obtain a better subsurface image. Please provide appropriate justification; otherwise, I strongly advise removing this term from the article to avoid giving a wrong impression to the readers. Also, it is quite strange to read terms like legacy sonic log, legacy first-break pics, and so on throughout the article [see Row 115-116]. I advise removing such terminology once the dataset is properly introduced [section 3].
- The second comment is related to the two datasets described in section 3. I have failed to understand how the first dataset 1 processed by Velseis Processing Ltd is relevant to this study as the authors have chosen dataset 2 for their purpose. It is natural to obtain two different results if two significantly different processing workflows are followed [Table 2]. I advise limiting the data description to dataset 2 only as this was used as input for further improvement. Regarding the use of first breaks from dataset 1, it can be simply mentioned that they were available and further corrected, and tomography was performed to obtain a P-wave velocity model.
- This is related to the choice of results used for the comparison. As dataset 2 is our input dataset, I will expect to see all the comparisons concerning to it [Row 195-196]. Instead, following the workflow shown in Figure 3, and the results shown in Figure 9 – PoSTM result from dataset 1, (a) is compared with its improved version in (b). Shouldn’t it be that the Pre-STM result available from ESP [Row 182-183] is to be compared with the new processed results? Or if the reason given in Row 213-215 is followed, PoSTM results can be easily produced for dataset 2 and can also be compared. But given the complex geological setting of the Limerick syncline, PreSTM is a more natural choice and should produce a better image – probably that is the reason ESP performed PreSTM instead of PoSTM by Velseis.
- Combining the two comments above, it is my understanding that the semblance analysis was performed on the CDP gathers of dataset 2 because it has a better signal-to-noise ratio as compared to dataset 1 [Row 190], and PoSTM has produced better imaging (dataset 1) than PreSTM (dataset 2). If this is the case, it is unclear in the article; kindly make this distinction clearer in the article on the datasets and the obtained results. But again, to my reservation, following Table 2, I have an impression that dataset 2 should have provided better results given the use of refraction analysis, two passes of velocity analysis, surface-consistent statics, residual scaling, etc. I suggest adding a figure showing a comparison of raw data with the previous processing by Velseis/ESP, or both at various stages. And then compare it with the new results. I understand that not everything can be shown, but at least the main steps can be shown as this is the core of the article i.e. seismic processing.
- I also have an unwise impression that the stack section shown in Fig. 8 is the same as Kirchhoff PoSTM in Fig. 9b. We expect the events to move up-dip after the migration; kindly cross-verify. The same is for Fig. 9a, looking at the migration smileys – it appears to be from PreSTM although it was not performed on dataset 1. Kindly cross-verify if there is no mix-up during the production of the figures.
- I did not understand the reason behind the time-converted drill holes. How was it done? Shouldn’t the interpretation of the final sections be converted to depth? Without it, how do we compare it with the information provided in section 2, in particular, to the thickness of the different formations shown in Fig. 2?
- Right now, it is extremely difficult to compare the seismic results (time-sections) with the direct information, i.e. drillholes and traveltime tomography results (both are in depth-domain), also not defined from a common depth/elevation reference point. There are starting and ending depths provided for the boreholes in Table 1 but there is no information from which elevation/depth level they are defined. They can be either be defined w.r.t. the mean sea level or the seismic reference datum. I advise synchronising all the results with a chosen reference level such that uniformity can be achieved for all results.
- Table 1 provides little information on the drill holes. I request that lithological sections be added for all boreholes defined from a common reference level. If this reference level is chosen differently from the seismic reference datum, it is to be clearly mentioned.
- Currently, it is unclear how the reprocessing is petrophysics-guided. Using petrophysics data for qualitative analysis of the obtained velocity model (from FATT in our case) is pretty standard, or doing seismic-to-well-tie per se. Unless a joint inversion of petrophysical data with FATT is performed or it is used as a constraint, the term “petrophysics-guided” is redundant and I advise removing it from the title, unless a proper justification is provided regarding it.
- As FATT is one of the most important components of the reprocessing, there is hardly any technical detail provided in section 3.3 such as offset information, grid samplings for forward modelling & inversion, picking uncertainty, etc. This section needs to be significantly expanded. Also, how the petrophysical data was used in building the velocity model is not clear.
- At this stage, it is difficult to estimate if the improvement in the seismic imaging is mainly due to the velocity model or data processing. Please provide clear evidence for it as this is the core theme of the article.
- The bibliography section needs to be revised as per the Solid Earth’s guidelines.
Minor comments:
- Row 14-15: Add some information on what Irish-type deposits are like.
- Row 15: Add a sentence or two describing the general geological setting and then jump to interfinering (specific to the current study). Currently, there is a lack of flow between the two sentences.
- Row 19: Remove “LK-11-02”.
- Row 41: “…mineral exploration tool.” (Add citation)
- Row 46: change to ‘Irish-type’
- Row 51: change to ‘seismic reflection survey’
- Row 54-55: Locations of Stonepark and Pallas deposits are not marked in Figure 1
- Row 55: Remove ‘the study area of this contribution’
- Row 80-81: Remove ‘Using the resulting improved image’
- Row 116, 118, 119: Remove the term ‘legacy’ here and in all the forthcoming sentences
- Row 123: Remove ‘associated’
- Row 127: Add reason here for why only 2?
- Row 158: Add citation or remove the reference. It is pretty common knowledge.
- Row 166: Define LSQR
- Row 177: Keep only ‘Seismic Data’
- Row 179-184: Is there an open-access processing report available for these datasets? Please cite here?
- Row 186: I suggest removing the term ‘standard’, or else define what standard means in this context.
- Row 189: Remove ‘historic’ and ‘slightly’ – it is hugely different
- Row 195-196: Change CDP to CMP for common abbreviation. You can put the sequence number for ease to know exactly at which step the data is taken, at 16 from dataset 2?
- Row 204-205: ‘guided by’ – this sentence is very loosely defined. At this stage, it is not very clear how exactly petrophysics and tomography were combined. This part needs a clear description. Showcase clearly how your approach is different from other regular methods.
- Row 231-232: “downhole petrophysical is rarely available in mineral exploration” – I disagree with this sentence given the fact that still until now, most of the seismic studies for mineral exploration are brownfield studies where the study of the physical rock properties is the first step towards it. Contrary, it is rare to see that petrophysical properties are not available.
- Row 222-317: This section is extremely long and very hard to follow. I recommend rewriting with a focus on the main scope of the paper. Extreme details had been provided related to the geology which are indeed very important but do not hold significant importance for this article. It can be part of another paper dedicated to the interpretation of the Limerick Syncline based on the new results.
- Row 285: “far offsets were included” – please provide some details on this here or in section 3.3
- Row 307: “shows interfingered low and moderate velocity zones” – I recommend using arrows or similar objects when describing any events for the ease and do not leave the reader to its imagination. Please consider a similar approach while describing other events in the article.
- Row 319-320: It is unclear whether the data shown is produved from a single borehole or all nearby boreholes. Kindly mention it!
- Row 347: Replace “informed” with “provided”
- Row 355: Doesn’t “CDP 720 and 1130 and between 1130 and 1335” mean between 720 to 1335? “300-500 ms” – please use the arrow to indicate it, I don’t see it.
- Row 368: “Figure 9c…” It is quite weird to read directly for 9c, is it the same as 9b? I recommend moving section 5.2 here first.
- Row: 370: “unknown faults F1 and F2”. I have the impression that these are not unknown faults when looking at Fig. 1. If I roughly extend the marked faults F1 and F2 from Fig. 9c up to the surface, they more or less coincide with the fault in the southern end of the profile LK-11-02 and around the cross-cutting profile LK-11-01 (Fig. 1).
- Row 383: What does “geometries” mean here?
- Row 421: I do not see any R1 in Fig. 9? Is it the same as shown in Fig. 8?
- Row 434-439: What is the dominant frequency at the target depth?
- Row 453-455: “faults (F1, F2) are now clearly identified” – but earlier in Row 370, it is mentioned they were unknown.
- Row 463: What do you mean by “gentle rotation of the reflectors”?
- Row 464: “domed reflections” – indicate it by arrows.
- Row 477: This section feels redundant given the fact that a very long section 4.1 with the same analogy already exists. I recommend separating them in terms of results and interpretation.
- Row 485: “Coincidentally with Stonepark” – it is not marked in Fig. 1.
- Row 502-503: “We interpreted a south-dipping fault zone, which can be part of the conduits that provided the mineral system with metal-bearing hydrothermal fluids.” I think this part has not been properly discussed in the article. The interpretation only comes in the last line in Rows 492-494, where it is from a north-dipping fault, but here it is from a south-dipping fault? Is it because of the syncline structure? Please make sure that this is properly introduced before in the introduction and the geology section.
- Rows 506-508: Very difficult to follow. Kindly expand it!
Figures
- Figure 1: Remove (a) and (b) and just mention that the inset shows a map of Ireland. Remove the other seismic lines other than LK-11-02 as they are not part of this study. The location of Stonepark and Pallas Green is not marked. Increase the font size of drill holes for better visibility. Consider showing a zoom of the northern part of the 2D seismic profile LK-11-02.
- Figure 2: Define what those features are in light pink crosscutting different formations?, Red colour – mineralisation?
- Figure 4: Text size on the axes and legends is too small. Consider replacing the light grey colour with something bright as it is very hard to see.
- Figure 5: I am not sure if it is required to show the plot for individual lithologies. At this stage, it was very difficult for me to follow the interpretation of results w.r.t stratigraphic units and individual lithologies, so leave it up to the author(s) to decide, particularly when these individual lithologies are not mapped by the seismic method.
- Figure 6: I advise enlarging this figure. Remove the single rays going deeper and focus on the upper ~200-300 m where most of the interpretation is done. Use zoom sections on the areas which are discussed in the main text. Add a similar stratigraphic bar on the top as shown in Figure 9c. I also recommend adding a figure with the lithological sections for all the drill holes alongside Table 1. Remember to mark the depths of the boreholes from a common reference datum for a better depth perspective.
- Figure 7: In (b), I recommend reversing the Y-axis to follow the same depth description as the geological sequence shown in Fig. 2. For (a), similar to Fig. 5, I leave it to the author(s) to decide.
- Figure 8: It is very difficult to compare the stack section image with the earlier results. The X-axis is defined in terms of CDP points; for Fig. 6, it is in UTM coordinates which is aagin different from Fig. 1. To attain uniformity, I recommend defining the CDP points on the profile in Fig. 1 and use the same CDP point description for Fig. 6. Although the profile length is marked in the bottom axis but then this is not marked in Fig. 1 which again makes it difficult to correlate these results with the geology of the area.
- Figure 9: Move the bar shown in Fig. 9c to the top of the figure. Show amplitude scales for each section. Is 9c the same as 9b with different amplitude scaling? Change the colour for Herberstown Limestone Fm., it is too light. Dolomitization/Brecciation, Argillaceous, Dykes are either not marked in the figure or are not visible.
Tables
- Table 2: Please provide the geometry information for both datasets. For dataset 1 at 17 – a dash between 18 and 140.
- Table 3: Add the first line mentioning at which step of dataset 2 from Table 2 is used as input data. At stage 5 here, why is the lower end of the frequency chosen as 5 Hz when BP was applied at 12/17 a step before? Please explain.
Bibliography: Please follow the Solid Earth’s recommended citation Style. Please leave a gap between citations.
Citation: https://doi.org/10.5194/egusphere-2025-2512-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 282 | 66 | 12 | 360 | 9 | 21 |
- HTML: 282
- PDF: 66
- XML: 12
- Total: 360
- BibTeX: 9
- EndNote: 21
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1