PyBacktrack 1.5: A community tool for reconstructing paleobathymetry of drill sites, geohistory and global paleobathymetry

Müller, R. Dietmar; Cannon, John; Williams, Simon; Dutkiewicz, Adriana; Wright, Nicky

doi:10.5194/egusphere-2026-3680

Preprints

https://doi.org/10.5194/egusphere-2026-3680

Preprints

04 Sep 2026

| 04 Sep 2026

PyBacktrack 1.5: A community tool for reconstructing paleobathymetry of drill sites, geohistory and global paleobathymetry

R. Dietmar Müller, John Cannon, Simon Williams, Adriana Dutkiewicz, and Nicky Wright

Abstract. PyBacktrack is a Python community software package that reconstructs the paleo-water depth (paleobathymetry) of ocean drill sites and industry wells through time by combining a model of tectonic subsidence with sediment decompaction and isostatic correction. Since the release of pyBacktrack 1.0, the software has undergone substantial development that culminates in pyBacktrack 1.5. The package now (i) runs on Python 3 and is distributed via the Python Package Index, conda-forge and Docker; (ii) includes a new paleobathymetry gridding module that reconstructs and backtracks all submerged present-day crust through geological time to produce time-dependent paleobathymetry grids; (iii) uses an updated default oceanic age-to-depth model, a globally complete set of rift start/end age grids derived from a deforming plate model, an updated total sediment thickness grid and an updated present-day age grid; (iv) supports both mantle and paleomagnetic reference frames and allows the gridded paleobathymetry to be optionally merged with externally produced paleobathymetry on synthetic, subducted crust; and (v) ships with expanded sea-level model and dynamic topography model libraries. We illustrate pyBacktrack 1.5 with three applications: (1) paleobathymetry reconstructions since the Early Cretaceous of the North Atlantic and the Southern Ocean; (2) one-dimensional backstripping and geohistory analysis at a long-record industry well on the Australian North West Shelf in three configurations (including or omitting eustatic sea level, and including a joint sea level and dynamic topography correction); and (3) a margin-scale gridded tectonic-subsidence-rate analysis of 109 Australian Northwest Shelf industry wells, quantifying how much of the inferred tectonic subsidence rate is dependent on eustatic and dynamic-topographic corrections in space and time. Together, these developments make pyBacktrack 1.5 a more accessible, extensible and globally applicable framework for reconstructing paleobathymetry and testing the tectonic, sedimentary, sea-level and dynamic-topographic controls on basin evolution.

Received: 23 Jun 2026 – Discussion started: 04 Sep 2026

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

R. Dietmar Müller, John Cannon, Simon Williams, Adriana Dutkiewicz, and Nicky Wright

Status: final response (author comments only)

RC1:
'Comment on egusphere-2026-3680', Valentin Zuchuat, 25 Sep 2026

Dear Editor(s),
Please see the attached document justifying my recommendation for minor revision.
Best regards,
Dr Valentin Zuchuat

Citation: https://doi.org/10.5194/egusphere-2026-3680-RC1
- AC1: 'Reply on RC1', Dietmar Müller, 03 Nov 2026
  
  We thank Valentin Zuchuat for the positive assessment and the detailed comments in the letter and the annotated PDF. We have revised the text, redrawn all figures except Fig. 1 and updated the companion figure scripts. The letter's five main points are answered first, followed by the comments in the annotated PDF. Line numbers refer to the discussion paper; section numbers in our replies refer to the revised manuscript, which is provided with and without tracked changes.
  MAIN POINTS
  Comment 1: "No Wheeler diagram and no need to rename the known concept of burial history diagrams. [...] I would recommend deleted any mention of Wheeler diagrams from the manuscript as non of the presented material are, in fact, Wheeler diagrams. The curves presented, notably in Fig. 8 of the paper, are classic burial history curves (e.g., Guidish et al., 1985), with a twist of adding the tectonic subsidence and sea-level curves on top of it."
  Reply: Agreed. Every mention of Wheeler diagrams has been removed. Fig. 8 and Sect. 4.8 (now titled "Stratigraphic log and burial history of the featured well") refer to burial-history diagrams and cite Guidish et al. (1985), and the corresponding script output is now named fig08_burial_history_<well>.
  Comment 2: "Broaden the concept of bathymetry, subsidence, and their impact on sediment routing in the introduction. [...] palaeobathymetry refers to any ancient, submerged surface, i.e., also including lacustrine and fluvial setting. [...] It omits that mechanical subsidence driven by fault movement/tectonic activity, especially in extensional settings, kicks in prior to thermal subsidence. Finally, there is no mention on the impact of palaeobathymetry (and the change of thereof) on sediment routing."
  Reply: The introduction now defines paleobathymetry as the past water depth of any submerged surface, from the deep ocean to epicontinental seas and lakes, and states that pyBacktrack addresses the marine part of this range, on oceanic crust and rifted continental margins. It names fault-controlled mechanical subsidence during rifting ahead of thermal subsidence, and adds the control of paleobathymetry on carbonate production and on where sediment is eroded, transported and deposited along source-to-sink systems (Helland-Hansen et al., 2016), with the deposited sediments forming the archive used in paleoclimate and burial-history analyses.
  Comment 3a: "The authors use a lot of geological time names (e.g., Early Cretaceous, Oligocene, etc.). However, none of these names are plotted anywhere. Therefore, please add the names of geological epochs and ages to Fig. 4-6-7-8-9-11."
  Reply: Figs. 4, 6, 7, 8, 9 and 11 now carry a timescale bar beneath the age axis with the stages, epochs and periods of the International Chronostratigraphic Chart (v2024/12; Cohen et al., 2013). Stage names are abbreviated, and names that do not fit their interval are dropped automatically. The bar is defined in the Fig. 4 caption.
  Comment 3b: "The log of the Asteras well should be added to the paper, either next to Fig. 5, or 7 or 8."
  Reply: Fig. 8 is now a two-panel figure. Panel (a) shows the stratigraphic log of Asteras: the lithology fractions of the twenty units as a hatched composition bar against present-day depth, a colour strip keyed to the units in the burial history of panel (b), and the bottom age of each unit.
  Comment 3c: "Fig. 5 needs the following additional updates: add all the geographical names of the area called in the paper. Add the extent of Fig. 10 on the map."
  Reply: Fig. 5 has been enlarged to include the Exmouth, Barrow and Dampier sub-basins and the Gascoyne and Cuvier abyssal plains. It labels every basin, sub-basin, abyssal plain and feature named in the text, outlines the area of Figs. 10 and 12 with a dashed box, and shows basin and sub-basin outlines and basin-bounding faults from the Australian Geological Provinces database (Geoscience Australia, 2013).
  Comment 3d: "The second parenthesis on the red y-axis is missing on Fig. 6."
  Reply: Fixed.
  Comment 3e: "Finally, on Fig. 10, would it be possible to add the major faults associated with the rifting episodes to the map, the check how the modelled subsidence fit the structural framework of the area?"
  Reply: The basin-bounding faults of Geoscience Australia (2013) are now drawn on all twelve panels of Fig. 10, so the rate fields can be read against the structural framework.
  Comment 4: "I would also like to understand better is how does PyBackTrack subsidence maps (Fig. 10) compare to published subsidence maps for the areas (e.g., Makuluni et al., 2024). Given that Makuluni et al. (2024) used 258 wells for the North Carnarvon Basin, how does the modelled subsidence from fewer, further spaced wells compare to published reconstructions."
  Reply: Sect. 4.11 now includes this comparison. Makuluni et al. (2024a) backstripped 258 wells in the Northern Carnarvon Basin to reconstruct its Jurassic to Early Cretaceous rift kinematics and found that rifting slowed at ~155 Ma, re-accelerated in the Tithonian at ~145 Ma and declined markedly after ~120 Ma. Our grids start at 150 Ma, the limit of the D10_gmcm9 model, but configuration A shows elevated rates in the latest Tithonian and Berriasian, and its Cretaceous minimum of 4–5 m/Myr falls between 125 and 105 Ma, after the decline they describe. We use the 109 wells archived in the reproducibility package (Müller et al., 2026b) because they span the whole margin from the Carnarvon to the Bonaparte Basin, have openly available stratigraphy and allow every result in the paper to be reproduced from the companion repository. The example is designed to demonstrate the workflow, not to replace a regional subsidence analysis; the same scripts accept any set of well files.
  Comment 5: "I would suggest reorganising chapter 5.1 from the oldest to the youngest, building the geological history from the start rather than the other way around."
  Reply: Done. The interpretation is now Sect. 4.11 (following Reviewer 2) and discusses the four intervals from oldest to youngest: 130–135, 100–105, 20–25 and 5–10 Ma.
  COMMENTS IN THE ANNOTATED PDF
  L34: "That is a bit of a narrow definition of palaeobathymetry, which refers to any submerged surface, i.e., also in lacustrine and fluvial setting."
  
  Reply: Revised; see main point 2.
  L38: "and tectonic/mechanic subsidence"
  
  Reply: Added; see main point 2.
  L50, L54: "carbonate production as well as the position of [...] where sediments are being eroded, transported, and deposited,"
  
  Reply: Added; see main point 2.
  L137: "While it makes total sense to still use Jarvis and McKenzie 1980 sets of equations, could we expect PyBackTrack updates to also integrate depth-dependent models such as Huismans, R., & Beaumont, C. (2011)."
  
  Reply: Yes. Sect. 5.2 now explains that the uniform-stretching model does not capture depth-dependent extension, which changes both the syn-rift and the thermal subsidence and has been invoked for the final stage of breakup on the NW Shelf (Makuluni et al., 2024b). Depth-dependent and multi-phase rift models are listed as a future development in Sect. 5.3.
  L255, L498, L513: "I would recommend deleting all your usages of "Wheeler" diagrams [...] They are burial history diagrams."
  
  Reply: Done; see main point 1.
  L405, L463, L477, L506, L545: "Add geological time boundaries on the x-axis since you refer to them as name (e.g., Jurassic-Cretaceaous)."
  
  Reply: Done; see main point 3a.
  L447, L448: "It would help to have the Asteras log in plotted in one of the figure [...] Would it be possible to show the extent of Fig. 10 grid on this figure? [...] Add all geographical names used in the paper that refer to this area on the map. [...] Is there a particular reason why not more wells were used to grid the subsidence in Fig. 10?"
  
  Reply: See main points 3b, 3c and 4.
  L458: "typo?" (on "Haq2024")
  
  Reply: Haq2024 is our abbreviation for the composite long-term sea-level curve bundled with pyBacktrack (Haq2024_Hybrid_SealevelCurve_Longterm), assembled from Haq and Ogg (2024), Haq (2014) and Haq (2018). The sentence has been rewritten so that the name and its sources are stated explicitly.
  L462: "The second parenthesis on the red y-axis is missing"
  
  Reply: Fixed.
  L478, L507: "Capital A at the start of Asteras"
  
  Reply: Corrected throughout.
  L517: "Does it? How? And what facies change?" (on the role of relative sea level in driving facies changes)
  
  Reply: The paragraph has been rewritten. It now names the three accumulation phases recorded at Asteras (Early Cretaceous sand-rich and reefal units, rapid Campanian to early Eocene shale–limestone and sandy accumulation, and Neogene carbonate-platform sediments after a near-hiatus between 34 and 24 Ma) and relates the facies to the paleo-water depth in the burial history: the sand-dominated Paleocene unit and the Neogene carbonates were deposited in 0–50 and 0–100 m of water, the Late Cretaceous shale–limestone units at outer-shelf depths of 50–200 m.
  L544: "Would it be possible to add a sentence explaining why the curve A (no SL no DT) sometimes have a similar trend as the two others [...] VS when the trend is opposite?"
  
  Reply: Sect. 4.9 now explains this. Curves A and B differ only by C, the rate of change of the combined sea-level and dynamic-topography correction evaluated between the dated horizons of each well. A rising sea level or a falling dynamic topography supplies part of the deepening that configuration A attributes to tectonics and makes C negative; a falling sea level or rising dynamic topography makes it positive. Where both corrections change slowly, as through most of the Late Cretaceous and Paleogene, A and B move together; they separate in the Valanginian–Hauterivian, the latest Albian, at the Oligocene–Miocene boundary and in the late Miocene to early Pliocene. While revising the figures we found that our companion scripts had applied the dynamic-topography correction with the wrong sign (pyBacktrack itself was not affected). With the sign corrected, the pattern differs from the submitted Fig. 9, and the four intervals discussed in Sect. 4.11 were re-derived from the corrected results.
  L552: "How do these results compare to the results of Makuluni et al., 2024 [...] Would it be possible to add the major faults associated with the rifting activity in the area on these maps?"
  
  Reply: See main points 4 and 3e.
  L606: "Having read the chapter 5.1, I think it would make more sense to tell the story in the order it happened, that is from the older to the youngest."
  
  Reply: Done; see main point 5.
  L639, L676, L681: "All these names should appear on the maps. [...] Also need to be on maps" (Exmouth, Barrow and Dampier; Gascoyne–Cuvier margin)
  
  Reply: All place names used in the text, including those in Sect. 4.11, now appear on Fig. 5.
  
  Citation: https://doi.org/10.5194/egusphere-2026-3680-AC1
RC2:
'Comment on egusphere-2026-3680', Eivind Straume, 08 Oct 2026

This paper presents several key updates to the PyBacktrack palebathymtric reconstruction software package. The software itself is a great contribution to scientific communities working with paleobathymethry related topics, the paper is clear and well written, and I thank the authors for making this software available to the community.
My comments are minor, but there are few aspects that could be better addressed. Below are a few general comments (mainly related to the discussion section), otherwise more specific in-line comments to the text and figures:
Line 122: Which crustal thickness grid? You cite the other input grids elsewhere, but cannot find reference to this
167: How do you decide how much to remove? Do you have assumptions for sedimentation rates? Do you have sediment age information embedded globally in this module?
325: Western interior seaway is not present on the maps
327: Closure of the Tethys is not visible on the maps
579: I see the double dash several places in the text, I suppose it is some conversion issue, and should be an em dash.
634: IQR. Spell out interquartile range first time it is used in the text
Discussion:
I do not see why 5.1 is in the discussion. This seems more like an example of the applications of the software package and should be in section 4. Also, as this is a model development paper, I think 5.2. and 5.3 could be expanded. See some suggestions below
The following could be in 5.2., a new discussion section, or added in other appropriate places in the manuscript:
I am wondering if you could add some constraints or guidelines, maybe a short section in the discussion, to how far back in time pybacktrack can/should be used? I suppose for margins there are no strict time limits, but for the seafloor there will be few remaining pieces beyond e.g. 150 Ma (?), so at which point it will mostly be a forward model?
Would also be good to have some more specific guidelines to the input data, especially dynamic topography. You mention there are considerable uncertainties, but not much on how this propagates back in time. For example a global dynamic topography model based backward advection I would think the uncertainties are too great to use beyond the Cenozoic and for upwelling regions linked to plumes such as Hawaii and Iceland probably much shorter.
Regarding dynamic topography, is the input grids converted to the reference frame one chooses to use? For both continental and oceanic domains? For example to a paleomagnetic reference frame if one reconstructs bathymetry for paleoclimate purposes. If not there should be a warning not to use dynamic topography with a paleomagnetic reference frame.
5.3
As I understand it, the pyBacktrack does not take into account processes such as e.g. volcanism in the oceanic domain that are younger than the age of the seafloor they occur upon. It would be challenging to account for all smaller features, but for prominent volcanic chains (Hawaii etc) and LIPs, one could use already available datasets to remove the effects of increased crustal thickness before the time of emplacement. I think this would be a nice addition for the future.
Figures:
In my view the visual presentation of the figures does not match the overall quality of this paper, this can be fixed by minor edits such as adding more appropriate color maps. Please see comments below.
Figure 2: I suggest adjusting the map to include the Eastern parts of the Tethys seaway. The seaway is mentioned in the text but partly missing in the figure . e.g. the seaway looks open at 20 Ma. Likely closed in the Eastern Tethys at this time. Also the Western interior seaway is mentioned in the text but not present in the maps.
Caption: Would change «reconstructed coastline» to «reconstructed present day coastlines», that is at least how it looks now on the map. And change «plate boundaries» to «subduction zones» as they are the only boundaries shown.
Figures 2 & 3: I suggest changing the colormap, this one is not perceptually uniform. Has a touch of green/cyan colors (-2200 m to -1500 m).
Figure 9: Avoid red-green color combination.
Figure 10: The subsidence rate color bar is not perceptually uniform, and also not color vision deficiency friendly. Please change to a more appropriate colormap.
Figure 12: Same issue as Fig 10. Figure caption is also misleading as the ~ -120 m 0 m range are «warm» colors. 0 m to 10 m are cool. Also same or very similar colors occur on both side of 0.
Overall I think this paper is of high quality, the software is a great contribution to related fields, and I recommend only minor revisions.

Best regards,
E. O. Straume

Citation: https://doi.org/10.5194/egusphere-2026-3680-RC2
- AC2: 'Reply on RC2', Dietmar Müller, 03 Nov 2026
  
  We thank Eivind O. Straume for the positive assessment and the suggestions for the discussion and the figures, which have shaped the new Sect. 5.1 and the revised colour scales. We have revised the text, redrawn all figures except Fig. 1 and updated the companion figure scripts. Line numbers refer to the discussion paper; section numbers in our replies refer to the revised manuscript, which is provided with and without tracked changes.
  Comment (line 122): "Which crustal thickness grid? You cite the other input grids elsewhere, but cannot find reference to this"
  
  Reply: CRUST1.0 (Laske et al., 2013), now stated and cited.
  Comment (line 167): "How do you decide how much to remove? Do you have assumptions for sedimentation rates? Do you have sediment age information embedded globally in this module?"
  
  Reply: Sect. 3.2 now explains this. The present-day sediment column at each grid point is treated as a single unit deposited at a constant rate of grain (solid) volume from the start of sedimentation to the present; sedimentation starts at the crustal age on oceanic crust and at the rift start age on continental crust. At each reconstruction time the grains deposited since then are removed and the remaining column is decompacted with the porosity–depth relationship of the chosen lithology. No sediment-age information beyond the crustal age and the rift start age is used. Where seismic stratigraphy or a sediment-age model is available, the one-dimensional backtrack module with a full stratigraphic column is the appropriate tool, and the text says so.
  Comment (line 325): "Western interior seaway is not present on the maps"
  
  Reply: PyBacktrack backtracks crust that is submerged today, so an epicontinental seaway over crust that is now emergent, such as the Western Interior Seaway, lies outside the gridded domain. Sect. 4.2 now states this and no longer describes the seaway; reconstructing such seaways requires paleogeographic models of continental flooding.
  Comment (line 327): "Closure of the Tethys is not visible on the maps"
  
  Reply: Fig. 2 has been redrawn; see our reply to the Figure 2 comment below.
  Comment (line 579): "I see the double dash several places in the text, I suppose it is some conversion issue, and should be an em dash."
  
  Reply: Replaced by en or em dashes throughout.
  Comment (line 634): "IQR. Spell out interquartile range first time it is used in the text"
  
  Reply: Done; the interquartile range is defined at its first use in Sect. 4.11 and in the Fig. 11 caption.
  Comment (Discussion): "I do not see why 5.1 is in the discussion. This seems more like an example of the applications of the software package and should be in section 4. Also, as this is a model development paper, I think 5.2. and 5.3 could be expanded."
  
  Reply: Agreed. The old Sect. 5.1 is now Sect. 4.11, where the interpretation of the NW Shelf subsidence-rate intervals runs from the oldest to the youngest interval. The discussion has been restructured and expanded: Sect. 5.1 ("Temporal reach and choice of input models") is new and answers the three questions below, Sect. 5.2 covers the limitations of the subsidence models and Sect. 5.3 the current state and future development.
  Comment (Discussion, temporal reach): "I am wondering if you could add some constraints or guidelines, maybe a short section in the discussion, to how far back in time pybacktrack can/should be used? I suppose for margins there are no strict time limits, but for the seafloor there will be few remaining pieces beyond e.g. 150 Ma (?), so at which point it will mostly be a forward model?"
  
  Reply: Sect. 5.1 now provides this information. In the bundled age grid (Seton et al., 2020), 24 % of today's oceanic crust is older than 100 Ma, 13 % is older than 120 Ma and only 4.5 % is older than 150 Ma. The backtracked part of a global reconstruction therefore covers most of the ocean floor in the Cenozoic but a rapidly shrinking fraction of it before ~120 Ma, and by the Late Jurassic a merged global grid is essentially a forward model. On rifted margins there is no such limit, and a one-dimensional analysis can extend as far back as the stratigraphic record of the site; there the practical limits are set by the time span of the dynamic-topography model (0–150 Ma for D10_gmcm9) and of the sea-level curve (0–205 Ma for Haq2024).
  Comment (Discussion, input data and dynamic topography): "Would also be good to have some more specific guidelines to the input data, especially dynamic topography. You mention there are considerable uncertainties, but not much on how this propagates back in time. For example a global dynamic topography model based backward advection I would think the uncertainties are too great to use beyond the Cenozoic and for upwelling regions linked to plumes such as Hawaii and Iceland probably much shorter."
  
  Reply: Sect. 5.1 now distinguishes the two families of bundled models. Backward-advection models (Müller et al., 2008b) capture regional detail in the recent past, but their errors grow with age (Conrad and Gurnis, 2003; Flament et al., 2013), and we recommend restricting them to the past ~70 Myr. Forward models with assimilated plate histories reach into the Mesozoic, but their predictions diverge with age. None of the bundled models resolves short-wavelength plume support such as that beneath Iceland or Hawaii, and near active or recent plumes the correction should be regarded as unconstrained. We recommend computing results with at least two dynamic-topography models, reporting the range as an additional uncertainty and restricting the analysis to the time span over which each model is defined. We also corrected the number of bundled models (fifteen, from six studies).
  Comment (Discussion, reference frames): "Regarding dynamic topography, is the input grids converted to the reference frame one chooses to use? For both continental and oceanic domains? For example to a paleomagnetic reference frame if one reconstructs bathymetry for paleoclimate purposes. If not there should be a warning not to use dynamic topography with a paleomagnetic reference frame."
  
  Reply: No conversion is needed. Each dynamic-topography model is bundled with the rotation file and static polygons of the geodynamic model that produced it, and pyBacktrack reconstructs every point with these before sampling the grid in that model's mantle reference frame. The correction is therefore independent of the reference frame chosen for the output and is applied on both oceanic and continental crust, including when paleobathymetry is generated in the paleomagnetic frame. This is now explained in Sect. 3.4 and summarised in Sect. 5.1.
  Comment (Section 5.3): "As I understand it, the pyBacktrack does not take into account processes such as e.g. volcanism in the oceanic domain that are younger than the age of the seafloor they occur upon. [...] for prominent volcanic chains (Hawaii etc) and LIPs, one could use already available datasets to remove the effects of increased crustal thickness before the time of emplacement. I think this would be a nice addition for the future."
  
  Reply: Correct. Sect. 5.2 now explains that the present-day residual of such features relative to the age–depth model is carried back unchanged to the time of crustal accretion, so edifices such as the Hawaiian–Emperor chain or the Kerguelen Plateau are reconstructed too shallow before their emplacement. Sect. 5.3 lists their removal, using compilations of large igneous provinces and seamounts and their ages (Coffin and Eldholm, 1994; Kim and Wessel, 2011; Johansson et al., 2018), as a future development.
  Comment (Figure 2): "I suggest adjusting the map to include the Eastern parts of the Tethys seaway. The seaway is mentioned in the text but partly missing in the figure. e.g. the seaway looks open at 20 Ma. Likely closed in the Eastern Tethys at this time. Also the Western interior seaway is mentioned in the text but not present in the maps. Caption: Would change «reconstructed coastline» to «reconstructed present day coastlines», that is at least how it looks now on the map. And change «plate boundaries» to «subduction zones» as they are the only boundaries shown."
  
  Reply: Fig. 2 has been redrawn on a Lambert azimuthal equal-area projection centred at 15° E, 35° N with a 62° cap, so that it spans the Tethys from the Caribbean gateway to the Persian Gulf, and now shows 120, 80, 40 and 20 Ma. At 20 Ma the grids show the Tethys reduced to the Mediterranean and a narrow eastern gateway between Arabia and Eurasia, and the text describes it in these terms. The Western Interior Seaway is addressed above (line 325). The caption now reads "reconstructed present-day coastlines" and "subduction zones".
  Comment (Figures 2 and 3): "I suggest changing the colormap, this one is not perceptually uniform. Has a touch of green/cyan colors (-2200 m to -1500 m)."
  
  Reply: Both figures, and the companion animations, now use cmocean deep (Thyng et al., 2016), which is perceptually uniform and monotonic in lightness.
  Comment (Figure 9): "Avoid red-green color combination."
  
  Reply: Figs. 9 and 11 now use blue, orange and reddish purple from the colour-vision-deficiency-safe Okabe–Ito set for configurations A, B and C.
  Comment (Figure 10): "The subsidence rate color bar is not perceptually uniform, and also not color vision deficiency friendly. Please change to a more appropriate colormap."
  
  Reply: The rate columns now use batlow (Crameri et al., 2020), with a dashed white zero contour separating subsidence from uplift. The difference column uses the diverging vik palette.
  Comment (Figure 12): "Same issue as Fig 10. Figure caption is also misleading as the ~ -120 m 0 m range are «warm» colors. 0 m to 10 m are cool. Also same or very similar colors occur on both side of 0."
  Reply: Fig. 12 now uses vik symmetric about 0 m, so that each colour corresponds to one sign and magnitude: blue denotes locations lower than today and brown locations higher than today. In revising the figure we also changed the field it shows to DT(t) − DT(0), sampled at the positions of present-day locations reconstructed to time t with the dynamic-topography model's own rotations, which is the quantity that enters the backstripping correction. The caption and Sect. 4.10 describe it accordingly.
  
  Citation: https://doi.org/10.5194/egusphere-2026-3680-AC2

R. Dietmar Müller, John Cannon, Simon Williams, Adriana Dutkiewicz, and Nicky Wright

Viewed

Total article views: 638 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
388	186	64	638	70	68

HTML: 388
PDF: 186
XML: 64
Total: 638
BibTeX: 70
EndNote: 68

Views and downloads (calculated since 04 Sep 2026)

Month	HTML	PDF	XML	Total
Sep 2026	223	74	38	335
Oct 2026	101	76	17	194
Nov 2026	64	36	9	109

Cumulative views and downloads (calculated since 04 Sep 2026)

Month	HTML	PDF	XML	Total
Sep 2026	223	74	38	335
Oct 2026	101	76	17	194
Nov 2026	64	36	9	109

Viewed (geographical distribution)

Total article views: 638 (including HTML, PDF, and XML) Thereof 638 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 11 Nov 2026

Short summary

PyBacktrack 1.5 is free software that helps researchers reconstruct how deep oceans and continental margins were in the past. It combines information on seafloor age, sediments and Earth movements, and can create maps or analyse individual drill sites and wells. The update makes the tool easier to install and use, supporting studies of how ocean basins and landscapes changed through time.


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