Pith. sign in

REVIEW 3 major objections 6 minor 59 references

Free Sentinel-1 archives and deep learning yield a quarterly map of 3,728 offshore oil and gas platforms in 2025, with more than 2,700 installations and a matching number of removals since 2017 signaling growth of mobile units.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-13 21:55 UTC pith:X22TJRTH

load-bearing objection Solid open quarterly platform inventory for three basins; the stock numbers and regional trends hold, but the headline turnover and short-lifespan story rest on an untested continuous-lifetime rule. the 3 major comments →

arxiv 2603.19801 v1 pith:X22TJRTH submitted 2026-03-20 eess.IV cs.AIcs.CV

Offshore oil and gas platform dynamics in the North Sea, Gulf of Mexico, and Persian Gulf: Exploiting the Sentinel-1 archive

Robin Spanier , Thorsten Hoeser , John Truckenbrodt , Felix Bachofer , Claudia Kuenzer This is my paper
classification eess.IV cs.AIcs.CV
keywords earth observationobject detectiontime seriesoffshore platformsoil and gasSentinel-1deep learningmarine infrastructure
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper shows that freely available Sentinel-1 radar time series plus a deep-learning object detector can produce a consistent quarterly inventory of offshore oil and gas platforms across the North Sea, Gulf of Mexico, and Persian Gulf from 2017 to 2025. In early 2025 the method locates 3,728 platforms (356, 1,641, and 1,731 in the three basins) and attaches size, water depth, coastal distance, national affiliation, and first- and last-appearance dates. While the Persian Gulf kept expanding until 2024, the other two basins declined after 2018–2020 peaks; at the same time more than 2,700 platforms appeared or moved and a comparable number vanished. The rising share of short-lived sites is read as evidence that mobile units such as jack-ups and drillships are becoming more important. The resulting open dataset supplies planners, regulators, and researchers with an independent, scalable record of marine energy infrastructure that proprietary or regionally incomplete sources cannot match.

Core claim

A consistent quarterly inventory derived from Sentinel-1 median composites and a pretrained YOLOv10 detector identifies 3,728 offshore oil and gas platforms in 2025—356 in the North Sea, 1,641 in the Gulf of Mexico, and 1,731 in the Persian Gulf. Between 2017 and 2025 more than 2,700 platforms were newly installed or relocated while a comparable number were decommissioned or relocated; the increasing fraction of short-lifespan locations indicates structural growth of mobile offshore units such as jack-ups and drillships. Platform size, water depth, coastal distance, and exclusive-economic-zone affiliation are also derived and released as an open GeoParquet dataset.

What carries the argument

Spatiotemporal consolidation of quarterly YOLOv10 detections on Sentinel-1 VH median composites: bounding boxes that overlap across quarters (IoU ≥ 0.1) are merged into unique platform IDs whose lifetime is defined as the continuous span from first to last detection, with gaps filled under the premise that fixed hydrocarbon sites operate without repeated short-term cycles. This step converts raw detections into the inventory counts, install/remove totals, and lifespan distributions that carry the paper’s structural-change argument.

Load-bearing premise

Every platform is treated as continuously present between its first and last detection, with quarterly gaps filled, because oil and gas sites are assumed to exploit fixed deposits without repeated short-term installation cycles.

What would settle it

Take a random sample of the short-lifespan (<5 year) platform IDs and check them against AIS tracks, company well reports, or high-resolution optical imagery; if many never existed as stationary platforms or are ships/noise, the reported ~2,700 turnover figures and the mobile-unit structural-change claim fail.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Open quarterly inventories can supplement or replace incomplete AIS and regulatory reports for maritime planning and environmental assessment.
  • Rising short-lifespan fractions quantify a shift from decades-long fixed platforms toward temporary mobile units, most visible in the North Sea.
  • Region-specific trends (Persian Gulf expansion until 2024, North Sea and Gulf of Mexico decline) can be tracked against energy-policy and geopolitical events without proprietary data.
  • The same workflow is claimed to scale to the full Sentinel-1 archive and thus to global platform monitoring.
  • Attributes such as water depth and coastal distance give a ready spatial basis for assessing competition among oil/gas, wind, shipping, and conservation uses.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If most short-lifespan detections are true mobile units rather than coverage or detection dropouts, regulators will need reporting categories for transient MODUs that permanent-platform inventories currently miss.
  • Joining the public timestamps to CO₂-storage or artificial-reef licensing records could measure how quickly decommissioned sites are reused—an analysis the paper flags but does not perform.
  • Running the same detector on additional SAR constellations would close the known acquisition gap in the central Gulf of Mexico and test whether the ~2,700 turnover figure holds under denser sampling.
  • Stable depth and distance distributions despite high turnover imply that new mobile units still occupy the same shallow-shelf niches as older fixed platforms, so spatial conflict patterns may evolve more slowly than raw platform counts.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The manuscript presents an automated, scalable workflow that applies a previously validated YOLOv10 detector to quarterly Sentinel-1 GRD median composites (2017Q1–2025Q1) to inventory offshore oil and gas platforms in the North Sea, Gulf of Mexico, and Persian Gulf. After post-processing (confidence, backscatter, IoU de-duplication, wind-farm filtering) and spatiotemporal consolidation, the authors derive a public Offshore Platform Dataset (OPD v1.0.0) with locations, EEZ affiliation, coastal distance, water depth, bounding-box size, and first/last detection quarters. They report 3,728 platforms in 2025Q1 (356 NS, 1,641 GoM, 1,731 PG), regionally divergent stock trends, and high turnover (~2,763 installs/relocations and ~2,718 decommissions/relocations), interpreting rising short-lifespan sites as evidence of structural growth in mobile units (jack-ups, drillships).

Significance. If the inventory and dynamics hold under reasonable sensitivity checks, the work supplies a rare, open, multi-region quarterly time series of oil/gas platforms with lifecycle and spatial attributes—directly useful for maritime spatial planning, decommissioning policy, and energy-transition analysis. Strengths include geographic hold-out training of the detector (F1 ≈ 0.90 on independent seas), competitive comparison of OPD against Paolo et al., OOGPs, and OSM (macro F1 0.884), explicit limitations (GoM coverage hole, 10 m resolution, no manual clean-up), and a fully open GeoParquet release. The modular Sentinel-1 + YOLO pipeline is in principle globally extensible and complements industry/government inventories that are incomplete or restricted.

major comments (3)
  1. Spatiotemporal consolidation section: lifetime is defined as continuous from first to last detection, with intermediate quarterly gaps filled because “offshore platforms are linked to the continuous exploitation of fixed hydrocarbon deposits,” while single-quarter detections are retained. This rule directly produces the headline turnover (~2,763 / ~2,718) and the short-lifespan pie-chart shares that underwrite the structural-change claim about mobile units. The manuscript notes intermittent non-detections and keeps isolated events, but reports no sensitivity of those figures (or of the <5-year / 5–8-year fractions) to alternative gap rules (e.g., no fill; fill only gaps ≤1 quarter; discard single-quarter detections; require ≥N consecutive quarters). Without that analysis, the dynamics interpretation remains under-supported even though the 2025 stock inventory is better validated.
  2. Results / lifespan dynamics and Discussion: “decommissioned or relocated” and “installed or relocated” are aggregated, and short-lifespan sites are interpreted as jack-ups/drillships. SAR median composites and bounding-box size cannot reliably separate true removal, relocation to a new site, temporary MODU departure/return, residual ships, or coverage/detection dropouts. The structural-change narrative therefore over-reaches the observables. Either (i) restrict claims to “appearance/disappearance of persistent backscatter clusters” and treat mobile-unit growth as a hypothesis, or (ii) add independent corroboration (AIS, industry well/platform databases, optical confirmation for a stratified sample of short-lifespan IDs) and report how many short-lifespan objects survive that filter.
  3. Limitations and Methods (data acquisition): a non-negligible portion of the central GoM deepwater basin lacks consistent Sentinel-1 coverage; very small platforms near the 10 m pixel scale are acknowledged as under-detected. Because GoM contributes ~44% of the 2025 stock and a large share of deepwater/FPSO examples, the paper should quantify the spatial extent of the coverage hole, the fraction of known deepwater assets that fall inside it, and the effect on regional time series and turnover counts (e.g., by masking or by comparison to BOEM/BSEE or PEMEX public lists for the US/Mexican shelves).
minor comments (6)
  1. Figure 4 / dataset comparison: report precision and recall (not only F1) per region and clarify the matching protocol (distance or IoU threshold, one-to-one assignment) used against the 2023 ground-truth set so that differences versus OOGPs and Paolo et al. are reproducible.
  2. Methods (post-processing): confidence threshold 0.4, IoU 0.2 for de-duplication, IoU ≥ 0.1 for multi-quarter clustering, and pixel threshold 150 (≈ −16.5 dB) are free parameters. A short ablation or justification against the independent test set would strengthen reproducibility.
  3. Platform “size” is bounding-box area of the SAR backscatter signature, not physical footprint. The text states this, but figure captions and the abstract-level phrasing should consistently say “detected backscatter extent” to avoid misreading as structural area.
  4. Several passages in the supplied preprint text contain garbled/OCR-like character runs (especially early pages and figure captions). Clean the production PDF and ensure all figure labels (hotspots E–H, depth scales) remain legible at print size.
  5. Table 1 and acquisition-density figure: state explicitly how many quarterly tiles had zero or sparse acquisitions and whether those quarters were dropped or interpolated for the stock time series.
  6. Related work: briefly position OPD against national inventories (e.g., BOEM platform structures, OSPAR, UK NSTA) even if access is restricted, so readers can judge complementarity rather than only open EO products.

Circularity Check

0 steps flagged

No circular derivation: platform counts and dynamics are produced by applying a held-out detector plus fixed post-processing rules, not by redefining the target from its own inputs.

full rationale

The paper’s derivation chain is self-contained and non-circular. Quarterly platform locations are obtained by running a YOLOv10 model (trained exclusively on South China Sea / Caspian / Guinea / Brazil data) on independent Sentinel-1 median composites of the three study regions that were completely withheld from training; reported precision/recall/F1 (0.91/0.89/0.90) and the later external-dataset comparison (F1 0.884 vs Paolo, OOGPs, OSM) are therefore genuine out-of-sample evaluations, not fitted-and-repredicted quantities. Spatiotemporal consolidation (IoU ≥ 0.1 clustering, lifetime = first-to-last detection with gaps filled) is an explicit methodological choice justified by domain knowledge of fixed hydrocarbon fields; it is not a self-definitional loop that forces the short-lifespan or ~2 700 turnover figures by construction. Fixed numerical thresholds (confidence 0.4, pixel value 150, IoU 0.2) are post-processing constants, not parameters fitted to the reported dynamics. Self-citation to the companion detector paper is ordinary method reuse and is not load-bearing for uniqueness or for the numerical claims themselves. Consequently no step reduces the central inventory or structural-change results to their own inputs.

Axiom & Free-Parameter Ledger

5 free parameters · 4 axioms · 0 invented entities

The inventory rests on standard SAR domain knowledge, a previously validated detector, and a handful of fixed post-processing thresholds plus one strong continuity assumption used to turn sparse detections into lifetimes. No new physical entities are postulated; free parameters are algorithmic cut-offs rather than fitted scientific constants.

free parameters (5)
  • detection confidence threshold = 0.4
    Predictions below 0.4 are discarded; value chosen for the multi-region run and directly controls which boxes enter the inventory.
  • duplicate-grouping IoU = 0.2
    Overlapping detections within a quarter are merged at IoU ≥ 0.2; alters platform counts in dense fields.
  • temporal-clustering IoU = 0.1
    Boxes across quarters are linked into one platform ID at IoU ≥ 0.1; defines the unit of lifespan analysis.
  • backscatter pixel threshold = 150
    Detections whose pixels are entirely below 150 (≈ −16.5 dB in the 8-bit composite) are removed as noise.
  • 8-bit backscatter mapping = [-40, 0] dB → [0, 255]
    σ° capped from −40 dB to 0 dB and linearly scaled to 0–255 before export; changes the numeric inputs seen by the detector.
axioms (4)
  • domain assumption Metallic offshore platforms produce stable, bright SAR backscatter clusters (often larger than physical size via layover and multiple scattering) that remain visible in quarterly median composites while ships are suppressed.
    Stated in introduction and methods; underpins both detection and the use of median composites.
  • ad hoc to paper A platform’s operational lifetime may be treated as continuous between its first and last detection, filling any intermediate quarterly gaps.
    Explicitly adopted in the spatiotemporal consolidation section to derive installation/removal dates and short-lifespan statistics.
  • domain assumption The YOLOv10 model trained on South China Sea, Caspian, Gulf of Guinea and Brazilian coast data generalizes to the North Sea, Persian Gulf and Gulf of Mexico at the reported F1.
    Taken from the companion paper [42] and used without re-training for the full 2017–2025 archive.
  • domain assumption Bounding-box area on SAR composites is a consistent relative size proxy across regions and years even though it is not the true physical footprint.
    Stated when size statistics are introduced; used for all regional size comparisons.

pith-pipeline@v1.1.0-grok45 · 22683 in / 3117 out tokens · 42024 ms · 2026-07-13T21:55:02.553426+00:00 · methodology

0 comments
read the original abstract

The increasing use of marine spaces by offshore infrastructure, including oil and gas platforms, underscores the need for consistent, scalable monitoring. Offshore development has economic, environmental, and regulatory implications, yet maritime areas remain difficult to monitor systematically due to their inaccessibility and spatial extent. This study presents an automated approach to the spatiotemporal detection of offshore oil and gas platforms based on freely available Earth observation data. Leveraging Sentinel-1 archive data and deep learning-based object detection, a consistent quarterly time series of platform locations for three major production regions: the North Sea, the Gulf of Mexico, and the Persian Gulf, was created for the period 2017-2025. In addition, platform size, water depth, distance to the coast, national affiliation, and installation and decommissioning dates were derived. 3,728 offshore platforms were identified in 2025, 356 in the North Sea, 1,641 in the Gulf of Mexico, and 1,731 in the Persian Gulf. While expansion was observed in the Persian Gulf until 2024, the Gulf of Mexico and the North Sea saw a decline in platform numbers from 2018-2020. At the same time, a pronounced dynamic was apparent. More than 2,700 platforms were installed or relocated to new sites, while a comparable number were decommissioned or relocated. Furthermore, the increasing number of platforms with short lifespans points to a structural change in the offshore sector associated with the growing importance of mobile offshore units such as jack-ups or drillships. The results highlighted the potential of freely available Earth observation data and deep learning for consistent, long-term monitoring of marine infrastructure. The derived dataset is public and provides a basis for offshore monitoring, maritime planning, and analyses of the transformation of the offshore energy sector.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

59 extracted references · 30 canonical work pages

  1. [1]

    Geometric accuracy of remote sensing images over oceans: The use of global offshore platforms,

    Y. Liu, C. Hu, Y. Dong, B. Xu, W. Zhan, and C. Sun, "Geometric accuracy of remote sensing images over oceans: The use of global offshore platforms," ������� ����������������������, vol. 222, pp. 244–266, 2019, doi: 10.1016/j.rse.2019.01.002

    work page doi:10.1016/j.rse.2019.01.002 2019

  2. [2]

    Current and projected global extent of marine built structures,

    A. B. Bugnot���������"Current and projected global extent of marine built structures," �����������, vol. 4, no. 1, pp. 33–41, 2021, doi: 10.1038/s41893-020-00595-1

    work page doi:10.1038/s41893-020-00595-1 2021

  3. [3]

    Safety assessment of shipping routes in the South China Sea based on the fuzzy analytic hierarchy process,

    J. Wang, M. Li, Y. Liu, H. Zhang, W. Zou, and L. Cheng, "Safety assessment of shipping routes in the South China Sea based on the fuzzy analytic hierarchy process," ��������������, vol. 62, pp. 46–57, 2014, doi: 10.1016/j.ssci.2013.08.002

    work page doi:10.1016/j.ssci.2013.08.002 2014

  4. [4]

    B. J. Williamson, P. Blondel, E. Armstrong, P. S. Bell, C. Hall, and J. J. Waggitt, "A Self-Contained Subsea Platform for Acoustic Monitoring of the Environment Around Marine Renewable Energy Devices–Field Deployments at Wave and Tidal Energy Sites in Orkney, Scotland," ��������������������, vol. 41, no. 1, pp. 67– 81, 2016, doi: 10.1109/JOE.2015.2410851

    work page doi:10.1109/joe.2015.2410851 2016

  5. [5]

    Evaluating differences in marine spatial data resolution and robustness: A North Sea case study,

    P. E. Posen, K. Hyder, M. Teixeira Alves, N. G. Taylor, and C. P. Lynam, "Evaluating differences in marine spatial data resolution and robustness: A North Sea case study," ��������������������������, vol. 192, p. 105206, 2020, doi: 10.1016/j.ocecoaman.2020.105206

    work page doi:10.1016/j.ocecoaman.2020.105206 2020

  6. [6]

    Automatic Geolocation and Measuring of Offshore Energy Infrastructure With Multimodal Satellite Data,

    P. Ma, M. Macdonald, S. Rouse, and J. Ren, "Automatic Geolocation and Measuring of Offshore Energy Infrastructure With Multimodal Satellite Data," �������� ������������, vol. 49, no. 1, pp. 66–79, 2024, doi: 10.1109/JOE.2023.3319741

    work page doi:10.1109/joe.2023.3319741 2024

  7. [7]

    OPDNet: An Offshore Platform Detection Network Based on Bitemporal Bimodal Remote-Sensing Images and a Pseudo-Siamese Structure,

    X. Zhang, M. Qiu, S. Tao, X. Ge, and M. Wang, "OPDNet: An Offshore Platform Detection Network Based on Bitemporal Bimodal Remote-Sensing Images and a Pseudo-Siamese Structure," ������������������ ���������������������������������������, vol. 18, pp. 6409–6421, 2025, doi: 10.1109/JSTARS.2025.3539674

    work page doi:10.1109/jstars.2025.3539674 2025

  8. [8]

    An overview of design, analysis, construction and installation of offshore petroleum platforms suitable for Cyprus oil/gas fields,

    K. Sadeghi, "An overview of design, analysis, construction and installation of offshore petroleum platforms suitable for Cyprus oil/gas fields," ������� ��������������, vol. 2, no. 4, pp. 1–16, 2007

    2007

  9. [9]

    OECD, �������������������������

  10. [10]

    World Oil Outlook 2050,

    OPEC, "World Oil Outlook 2050," Vienna, 2025. [Online]. Available: https://www.opec.org/assets/ assetdb/woo-2025.pdf

    2050

  11. [11]

    Assessing the global ocean science community: understanding international collaboration, concerns and the current state of ocean basin research,

    R. W. K. Potter and B. C. Pearson, "Assessing the global ocean science community: understanding international collaboration, concerns and the current state of ocean basin research," �����������������, vol. 2, no. 1, 2023, doi: 10.1038/s44183-023-00020-y

    work page doi:10.1038/s44183-023-00020-y 2023

  12. [12]

    Decommissioning of offshore oil and gas platforms: A systematic literature review of 15 factors involved in the process,

    P. C. J. Da Vidal���������"Decommissioning of offshore oil and gas platforms: A systematic literature review of 15 factors involved in the process," �����������������, vol. 255, p. 111428, 2022, doi: 10.1016/j.oceaneng.2022.111428

    work page doi:10.1016/j.oceaneng.2022.111428 2022

  13. [13]

    Ecological role of offshore structures,

    A. R. Gates and D. O. B. Jones, "Ecological role of offshore structures," �����������, vol. 7, no. 4, pp. 383– 384, 2024, doi: 10.1038/s41893-024-01316-8

    work page doi:10.1038/s41893-024-01316-8 2024

  14. [14]

    Marine Infrastructure Detection with Satellite Data—A Review,

    R. Spanier and C. Kuenzer, "Marine Infrastructure Detection with Satellite Data—A Review," ������� �������, vol. 16, no. 10, p. 1675, 2024, doi: 10.3390/rs16101675

    work page doi:10.3390/rs16101675 2024

  15. [15]

    Petrobras, ��������������������������������[Online]

    D. Petrobras, ��������������������������������[Online]. Available: https://commons.wikimedia.org/w/index.php? curid=5621984

  16. [16]

    Murray, �������������������������������������������� ��������������: BBC, 2026

    A. Murray, �������������������������������������������� ��������������: BBC, 2026. [Online]. Available: https:// www.bbc.com/news/articles/cq5y7dd284do

    2026

  17. [17]

    Skopljak, ������������������������������������������� ��������������������: Offshore Energy, 2025

    N. Skopljak, ������������������������������������������� ��������������������: Offshore Energy, 2025. [Online]. Available: https://www.offshore-energy.biz/denmark- grants-its-first-ever-approval-for-co2-storage-facility/

    2025

  18. [18]

    Exploring 4D seismic potential for monitoring CO2 injection in depleted North Sea gas fields,

    S. Y. Toh, C. MacBeth, J. Landa, and H. Heidari, "Exploring 4D seismic potential for monitoring CO2 injection in depleted North Sea gas fields," �������������� ���������������������������������, vol. 148, p. 104529, 2025, doi: 10.1016/j.ijggc.2025.104529

    work page doi:10.1016/j.ijggc.2025.104529 2025

  19. [19]

    Monitoring long-term storage of CO2 in a gas and condensate field in the North Sea off the coast of Norway using seismic methods,

    A. P. Singh���������"Monitoring long-term storage of CO2 in a gas and condensate field in the North Sea off the coast of Norway using seismic methods," ����������, vol. 90, no. 4, M167-M180, 2025, doi: 10.1190/geo2024-0715.1

    work page doi:10.1190/geo2024-0715.1 2025

  20. [20]

    OECD Science, Technology and Industry Working Papers,

    OECD, "OECD Science, Technology and Industry Working Papers," 2021

    2021

  21. [21]

    Tracking the global reduction of marine traffic during the COVID-19 pandemic,

    D. March, K. Metcalfe, J. Tintoré, and B. J. Godley, "Tracking the global reduction of marine traffic during the COVID-19 pandemic," ���������������������, vol. 12, no. 1, p. 2415, 2021, doi: 10.1038/s41467-021- 22423-6

    doi:10.1038/s41467-021- 2021

  22. [22]

    European Commission and Directorate-General for Environment, ����������������������������������� ��������������������������������������������������� ��������������: Publications Office of the European Union, 2022

    2022

  23. [23]

    The Ocean 100: Transnational corporations in the ocean economy,

    J. Virdin���������"The Ocean 100: Transnational corporations in the ocean economy," ����������������, vol. 7, no. 3, 2021, doi: 10.1126/sciadv.abc8041

    work page doi:10.1126/sciadv.abc8041 2021

  24. [24]

    Automating offshore infrastructure extractions using synthetic aperture radar & Google Earth Engine,

    B. A. Wong, C. Thomas, and P. Halpin, "Automating offshore infrastructure extractions using synthetic aperture radar & Google Earth Engine," ��������������� ��������������, vol. 233, p. 111412, 2019, doi: 10.1016/j.rse.2019.111412

    doi:10.1016/j.rse.2019.111412 2019

  25. [25]

    Estimating offshore oil production using DMSP-OLS annual composites,

    C. Sun, Y. Liu, S. Zhao, and S. Jin, "Estimating offshore oil production using DMSP-OLS annual composites," ��������������������������������������������������, vol. 165, pp. 152–171, 2020, doi: 10.1016/j.isprsjprs.2020.05.019

    work page doi:10.1016/j.isprsjprs.2020.05.019 2020

  26. [26]

    Automatic extraction of offshore platforms using time-series Landsat-8 Operational Land Imager data,

    Y. Liu, C. Sun, Y. Yang, M. Zhou, W. Zhan, and W. Cheng, "Automatic extraction of offshore platforms using time-series Landsat-8 Operational Land Imager data," �����������������������������, vol. 175, pp. 73– 91, 2016, doi: 10.1016/j.rse.2015.12.047

    doi:10.1016/j.rse.2015.12.047 2016

  27. [27]

    IceLines - A new data set of Antarctic ice shelf front positions,

    C. A. Baumhoer, A. J. Dietz, K. Heidler, and C. Kuenzer, "IceLines - A new data set of Antarctic ice shelf front positions," ���������������, vol. 10, no. 1, p. 138, 2023, doi: 10.1038/s41597-023-02045-x

    work page doi:10.1038/s41597-023-02045-x 2023

  28. [28]

    World Settlement Footprint 3D - A first three-dimensional survey of the global building stock,

    T. Esch���������"World Settlement Footprint 3D - A first three-dimensional survey of the global building stock," �����������������������������, vol. 270, p. 112877, 2022, doi: 10.1016/j.rse.2021.112877

    work page doi:10.1016/j.rse.2021.112877 2022

  29. [29]

    Global dynamics of the offshore wind energy sector monitored with Sentinel-1: Turbine count, installed capacity and site specifications,

    T. Hoeser and C. Kuenzer, "Global dynamics of the offshore wind energy sector monitored with Sentinel-1: Turbine count, installed capacity and site specifications," ������������������������������������������������������� ��������������, vol. 112, p. 102957, 2022, doi: 10.1016/j.jag.2022.102957

    doi:10.1016/j.jag.2022.102957 2022

  30. [30]

    Available: https://unctad.org/ publication/exploring-space-technologies-sustainable- development

    UNCTAD, ��������������������������������������������� �������������[Online]. Available: https://unctad.org/ publication/exploring-space-technologies-sustainable- development

  31. [31]

    The role of satellite remote sensing in climate change studies,

    J. Yang���������"The role of satellite remote sensing in climate change studies," ������������������, vol. 3, no. 10, pp. 875–883, 2013, doi: 10.1038/nclimate1908

    work page doi:10.1038/nclimate1908 2013

  32. [32]

    Invariant triangle-based stationary oil platform detection from multitemporal synthetic aperture radar data,

    L. Cheng, K. Yang, L. Tong, Y. Liu, and M. Li, "Invariant triangle-based stationary oil platform detection from multitemporal synthetic aperture radar data," ��������������������, vol. 7, no. 1, p. 73537, 2013, doi: 10.1117/1.JRS.7.073537

    work page doi:10.1117/1.jrs.7.073537 2013

  33. [33]

    Offshore Metallic Platforms Observation Using Dual-Polarimetric TS-X/TD-X Satellite Imagery: A Case Study in the Gulf of Mexico,

    A. Marino, D. Velotto, and F. Nunziata, "Offshore Metallic Platforms Observation Using Dual-Polarimetric TS-X/TD-X Satellite Imagery: A Case Study in the Gulf of Mexico," ������������������������������������������� ��������������, vol. 10, no. 10, pp. 4376–4386, 2017, doi: 10.1109/JSTARS.2017.2718584

    work page doi:10.1109/jstars.2017.2718584 2017

  34. [34]

    Automatic Extraction of Offshore Platforms in Single SAR Images Based on a Dual-Step-Modified Model,

    J. Zhang, Q. Wang, and F. Su, "Automatic Extraction of Offshore Platforms in Single SAR Images Based on a Dual-Step-Modified Model," ���������������� ������������, vol. 19, no. 2, 2019, doi: 10.3390/s19020231

    work page doi:10.3390/s19020231 2019

  35. [35]

    Proliferation of offshore wind farms in the North Sea and surrounding waters revealed by satellite image time series,

    W. Xu���������"Proliferation of offshore wind farms in the North Sea and surrounding waters revealed by satellite image time series," ��������������������������������� �������, vol. 133, p. 110167, 2020, doi: 10.1016/j.rser.2020.110167

    doi:10.1016/j.rser.2020.110167 2020

  36. [36]

    Detecting Offshore Drilling Rigs with Multitemporal NDWI: A Case Study in the Caspian Sea,

    H. Zhu, G. Jia, Q. Zhang, S. Zhang, X. Lin, and Y. Shuai, "Detecting Offshore Drilling Rigs with Multitemporal NDWI: A Case Study in the Caspian Sea," ��������������, vol. 13, no. 8, p. 1576, 2021, doi: 10.3390/rs13081576

    work page doi:10.3390/rs13081576 2021

  37. [37]

    Synthetic aperture radar: marine user’s manual,

    C. R. Jackson and J. R. Apel, "Synthetic aperture radar: marine user’s manual," 2004

    2004

  38. [38]

    DeepOWT: a global offshore wind turbine data set derived with deep learning from Sentinel-1 data,

    T. Hoeser, S. Feuerstein, and C. Kuenzer, "DeepOWT: a global offshore wind turbine data set derived with deep learning from Sentinel-1 data," ���������������������, vol. 14, no. 9, pp. 4251–4270, 2022, doi: 10.5194/essd- 14-4251-2022

    work page doi:10.5194/essd- 2022

  39. [39]

    Satellite mapping reveals extensive industrial activity at sea,

    F. S. Paolo���������"Satellite mapping reveals extensive industrial activity at sea," ������, vol. 625, no. 7993, pp. 85–91, 2024, doi: 10.1038/s41586-023-06825-8

    doi:10.1038/s41586-023-06825-8 2024

  40. [40]

    Global offshore wind turbine dataset,

    T. Zhang, B. Tian, D. Sengupta, L. Zhang, and Y. Si, "Global offshore wind turbine dataset," ���������������, 16 vol. 8, no. 1, p. 191, 2021, doi: 10.1038/s41597-021- 00982-z

    work page doi:10.1038/s41597-021- 2021

  41. [41]

    SyntEO: Synthetic dataset generation for earth observation and deep learning – Demonstrated for offshore wind farm detection,

    T. Hoeser and C. Kuenzer, "SyntEO: Synthetic dataset generation for earth observation and deep learning – Demonstrated for offshore wind farm detection," ������ ��������������������������������������������, vol. 189, pp. 163–184, 2022, doi: 10.1016/j.isprsjprs.2022.04.029

    doi:10.1016/j.isprsjprs.2022.04.029 2022

  42. [42]

    Deep learning- based object detection of offshore platforms on Sentinel- 1 imagery and the impact of synthetic training data,

    R. Spanier, T. Hoeser, and C. Kuenzer, "Deep learning- based object detection of offshore platforms on Sentinel- 1 imagery and the impact of synthetic training data," ���������������������������������������, vol. 47, no. 5, pp. 1–25, 2026, doi: 10.1080/01431161.2026.2612908

    work page doi:10.1080/01431161.2026.2612908 2026

  43. [43]

    IHO Sea Areas, version 3,

    Flanders Marine Institute, "IHO Sea Areas, version 3," 2018

    2018

  44. [44]

    The North Sea,

    J.-P. Ducrotoy, M. Elliott, and V. N. de Jonge, "The North Sea," �������������������������, vol. 41, 1-6, pp. 5–23, 2000, doi: 10.1016/S0025-326X(00)00099-0

    work page doi:10.1016/s0025-326x(00)00099-0 2000

  45. [45]

    A brief analysis of North Sea physics,

    J. Sündermann and T. Pohlmann, "A brief analysis of North Sea physics," �����������, vol. 53, no. 3, pp. 663–689, 2011, doi: 10.5697/oc.53-3.663

    work page doi:10.5697/oc.53-3.663 2011

  46. [46]

    Oil 2025: Analysis and forecast to 2030,

    IEA, "Oil 2025: Analysis and forecast to 2030," 2025. [Online]. Available: https://iea.blob.core.windows.net/ assets/c0087308-f434-4284-b5bb-bfaf745c81c3/ Oil2025.pdf

    2025

  47. [47]

    The Persian Gulf,

    V. Prescott and C. Schofield, "The Persian Gulf," in ���� ������������������������������������������, V. Prescott and C. Schofield, Eds.: Brill Nijhoff, 2004, pp. 497–518

    2004

  48. [48]

    The circulation of the Persian Gulf: a numerical study,

    J. Kämpf and M. Sadrinasab, "The circulation of the Persian Gulf: a numerical study," ����������, vol. 2, no. 1, pp. 27–41, 2006, doi: 10.5194/os-2-27-2006

    work page doi:10.5194/os-2-27-2006 2006

  49. [49]

    Available: https://www.eia.gov/todayinenergy/detail.php?id=65504

    EIA, ��������������������������������������������� ���������������������������������[Online]. Available: https://www.eia.gov/todayinenergy/detail.php?id=65504

  50. [50]

    The Gulf of Mexico: An Overview,

    L. McKinney���������"The Gulf of Mexico: An Overview," ������� , vol. 34, no. 1, pp. 30–43, 2021, doi: 10.5670/oceanog.2021.115

    work page doi:10.5670/oceanog.2021.115 2021

  51. [51]

    Target detection in synthetic aperture radar imagery: a state-of-the-art survey,

    K. El-Darymli, P. McGuire, D. Power, and C. Moloney, "Target detection in synthetic aperture radar imagery: a state-of-the-art survey," ��������������������, vol. 7, no. 1, p. 71598, 2013, doi: 10.1117/1.JRS.7.071598

    work page doi:10.1117/1.jrs.7.071598 2013

  52. [52]

    E. S. A. ESA, ������������[Online]. Available: https:// sentiwiki.copernicus.eu/web/sentinel-1

  53. [53]

    Available: https:// developers.google.com/earth-engine/datasets/catalog/ COPERNICUS_S1_GRD

    GEE and ESA, ������������������������������������������������������������ �������������� ��������[Online]. Available: https:// developers.google.com/earth-engine/datasets/catalog/ COPERNICUS_S1_GRD

  54. [54]

    YOLOv10: Real-Time End-to-End Object Detection,

    A. Wang���������"YOLOv10: Real-Time End-to-End Object Detection," 2024

    2024

  55. [55]

    Available: https:// emodnet.ec.europa.eu/geonetwork/srv/eng/ catalog.search#/metadata/8201070b-4b0b-4d54-8910- abcea5dce57f

    EMODnet, ��������������������������������������� �����������������[Online]. Available: https:// emodnet.ec.europa.eu/geonetwork/srv/eng/ catalog.search#/metadata/8201070b-4b0b-4d54-8910- abcea5dce57f

    arXiv

  56. [56]

    Available: https://download.gebco.net/#

    GEBCO, �������������������������������[Online]. Available: https://download.gebco.net/#

  57. [57]

    The Offshore Oil and Gas Platforms (OOGPs) dataset based on satellite data spanning 2017 to 2023,

    L. Si, S. Zhou, I. Irakulis-Loitxate, J. Roger, and L. Guanter, "The Offshore Oil and Gas Platforms (OOGPs) dataset based on satellite data spanning 2017 to 2023," 2026

    2017

  58. [58]

    Country Analysis Brief: Iran,

    EIA, "Country Analysis Brief: Iran," Washington DC,

  59. [59]

    Available: https://www.eia.gov/ international/content/analysis/countries_long/Iran/pdf/ Iran%20CAB%202024.pdf

    [Online]. Available: https://www.eia.gov/ international/content/analysis/countries_long/Iran/pdf/ Iran%20CAB%202024.pdf