VISIR-2: ship weather routing in Python

Mannarini, Gianandrea; Salinas, Mario Leonardo; Carelli, Lorenzo; Petacco, Nicola; Orović, Josip

doi:https://doi.org/10.5194/egusphere-2023-2060

Preprints

https://doi.org/10.5194/egusphere-2023-2060

Preprints

16 Nov 2023

| 16 Nov 2023

VISIR-2: ship weather routing in Python

Gianandrea Mannarini, Mario Leonardo Salinas, Lorenzo Carelli, Nicola Petacco, and Josip Orović

Abstract. Ship weather routing, which involves suggesting low-emission routes, holds potential for contributing to the decarbonisation of maritime transport. However, its quantitative impact has been explored only to a limited extent, also for a lack of readily deployable open-source and open-language computational models.

As a response, the VISIR model has been refactored in Python, incorporating new features. The velocity composition with currents has been refined, now encompassing leeway as well. For motor vessels, the angle of attack of waves has been considered, while sailboats now account for the combined effects of wind and sea currents. A least-CO₂ algorithm in presence of dynamic graph edge weights has been implemented and validated, proving a quasi-linear computational performance which outperforms VISIR-1. The software suite’s modularity has been significantly improved, alongside a thorough validation against arious benchmarks.

The resulting VISIR-2 model has been employed in numerical experiments within the Mediterranean Sea for the entire 2022, utilising meteo-oceanographic analysis fields. For a 125-meter-long ferry, the distribution of carbon dioxide savings follows a bi-exponential distribution. Two-digit CO₂ savings were possible for more than ten days in a year. Largest savings were achieved in avoiding upwind sailing and using the lowest engine load. In the case of an 11-meter sailboat, time savings increase with the extent of path elongation, particularly during upwind sailing. The sailboat’s routes were approximately 3 % shorter thanks to optimisation, and there was potential for additional savings when favourable currents were in play. The impact of leeway was minor, but disregarding it would result in a systematic underestimation of route durations.

VISIR-2 is a collaborative model with the capacity to harness knowledge from oceanography, ocean engineering, and computer science, to contribute to the decarbonisation efforts in the shipping industry.

Received: 08 Sep 2023 – Discussion started: 16 Nov 2023

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 preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Preprint (PDF, 32299 KB)

Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
Preprint (32299 KB)

Supplement (9051 KB)

Download & links

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Journal article(s) based on this preprint

24 May 2024

VISIR-2: ship weather routing in Python

Gianandrea Mannarini, Mario Leonardo Salinas, Lorenzo Carelli, Nicola Petacco, and Josip Orović

Geosci. Model Dev., 17, 4355–4382, https://doi.org/10.5194/gmd-17-4355-2024,https://doi.org/10.5194/gmd-17-4355-2024, 2024

Short summary

Gianandrea Mannarini, Mario Leonardo Salinas, Lorenzo Carelli, Nicola Petacco, and Josip Orović

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2023-2060', Anonymous Referee #1, 30 Nov 2023
I recommend that this manuscript is accepted for publication with minor revisions.
The submitted manuscript describes new updates to the VISIR-2 weather routing software. The model has been developed from Matlab into Python and is available as an open-source model with no license. A brief literature review is provided, alongside an in-depth methodology, providing a detailed explanation of each component of the model. The paper finishes by demonstrating VISIR-2’s ability to minimise CO2 emissions in two case studies – one for a motor ship and one for a sailing ship.
The paper excels in its scientific reproducibility and I would like the congratulate the authors on the well-organised manuscript with impressive levels of additional detail provided, including the user manual. I believe that the modular implementation, released under a freely available GNU General Public License, has strong uses for the scientific community and the wider community. The paper excels in its presentation quality, in particular the presentation of the very clear figures. VISIR-2 benefits greatly from its validation and provides very useful and detailed insight into the model’s computation time.
That said, I believe the paper could improve in the following aspects.
I believe the manuscript would benefit from a clearer description of the inherent novelty of the model within the method section. The novel aspects of the model are not highlighted sufficiently in the method, and I finished reading the section wondering which parts of VISIR-2 form the novel features. I would also like to see this same novelty more clearly described in the introduction – here, the authors describe additional features well, but the section fails to comment on the novelty of these features in the context of the body of literature in this field.
The paper heavily focuses its novelty on the fact that their model has been developed as open source, modular and free to use (which VISIR-2 largely does with exceptional care and quality). While this is indeed useful, I believe the manuscript could benefit from a discussion on the novel quantitative implications of their CO2-saving results. (In fact, the authors also touch upon this in the first paragraph of their abstract: “…its quantitative impact has been explored only to a limited extent…”.) I don’t currently fully understand what novel academic question they try to answer with their analysis, or whether it is just used to showcase VISIR-2 (which it does a great job of). I believe that these quantitative novelties are present in the paper, but more work is needed to outline them in the results/discussion section. While outside the area of model development, the results of this paper are great work and I believe it would be of use to advance the field more generally. I would ideally like to see an explanation of how their quantitative results/discussion contribute to new science. This could also be brought out briefly in the abstract.
On a similar note, the results/discussion does a great job of highlighting the potential of weather routing as a CO2 reduction measure. However, there is almost no discussion on how their results compare to other studies in the literature. Where do their CO2 savings fit in the literature? Do they agree/disagree? What relative contribution is the paper making to this body of literature? I would like to see more discussion on this. That said, the level of detail in the results/discussion is brilliant and very commendable, a great job on that.
Some of my above suggestions have been completed to some extent in the conclusion - but should be strengthened in the other sections. No new information should be presented in the conclusion.
Alongside this, I have the following minor comments:
I believe a figure at the start of the method section that provides an outline of all steps of the model would help readers to understand the model structure more generally.

h-hat is mentioned at the end of page 5 but I cannot see this variable in Figure 1.

Removing collinear edges is a great idea, one which I will test myself!

I found Section 2.3.2 Space interpolation confusing. I would consider rewording. A better description would be useful. Same with Figure 5. Please explain the meaning of Head and Tail and give a more clear description of the two interpolation methods. Indeed, please state the difference and why the two are necessary.

The least squares fit for the blue lines in Figure 9 doesn’t seem to work. I’m not sure if an alternative is possible, but if my interpretation is correct, the solid blue line (total computation time) should not fall below the dashed blue line (Dijkstra component of computation time). In fact, this is the same for the red and green lines also.

Section 5.1 Environmental fields – should the URL next to “a lower resolution (0.4, URL)” go as a footnote?

Figure 10 b – super clear and very interesting. An engaging plot.
Citation: https://doi.org/10.5194/egusphere-2023-2060-RC1
- AC1: 'Reply on RC1', Gianandrea Mannarini, 19 Feb 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2060/egusphere-2023-2060-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2060-AC1
RC2:
'Comment on egusphere-2023-2060', Anonymous Referee #2, 09 Jan 2024

General Comments
Overall the manuscript is of high quality and provides a very thorough analysis on a proposed navigational approach for both motorboats and environmentally-driven/impacted surface vessels. Currents and leeways are compared for two separate boat models and meticulously analyzed seasonally over a scoped domain, offering high quality conclusions and results discussion, while also offering rich model bases for the open literature. Another open-source weather routing software is invaluable as prior to this, primarily only openCPN was the go-to open-source option that would not be able to handle motorboat and CO2-based measures of optimization. The primary contribution is its attention to detail and reproducibility for science computation that is wholly lacking on the open-source playing field.
Specific Comments
There are still a handful of revisions I believe the paper needs to undergo to be finalized for publication.
Throughout the manuscript, there are numerous undefined acronyms in this section that either need to be noted as a footnote, or explained to the reader. E.g., GFS, OSCAR, AVALON, GUTTA, openCPN, especially in Sec 1.1.1.
The scientific notation is very hard to follow. It is very hard to distinguish a vector quantity from a scalar. Can you use typographic convention to aid the reader? Hard-to-follow notational conventions induces increased effort in assessing the research contributions in Sec. 2.1 because of this mismatch and lack of clarity. In addition, the authors flip between radians and degrees. For consistency and legibility, they should remain the same throughout unless where deemed necessary for more intuitive understanding for the reader.
Sec 5.1: Should more than just surface current be used for the ferry? It seems the draft is in excess of 4m, so potentially 0, 2, and 4m relative z-levels could be employed for even further increased fidelity in the optimization at the expense of computational complexity.
There is a large and important question when assessing graph edges throughout the paper and that is what coordinate system/projection/transformation is assumed. This is a very important piece of information missing from a geodesy and nautical navigation standpoint.
The conclusion section is too long and should be a synopsis of the contribution and highlights of the results that a reader should and must take away from reading the publication. No new results or new discussion should be present in the conclusion.
Technical Corrections
Ln 6: A least-CO2 algorithm in the presence of
Ln 12: Two-digit percentage? Two-digit quantity? Suggest clarification on this improvement as its unclear on the units / tangibility of statements. Two-digit pounds of CO2 emission for example is not as impressive as say two-digit percentage of overall CO2 expenditure.
Ln 14: 3% shorter as measured by time, or distance? Based on the authors’ prior words “path elongation”, it is confusing to the reader to tout a 3% shorter result.
Ln 17-18: If you are using winds, then meteorology should be included in the list of knowledge bases pulled from
Ln 37: CE-Ship model is an undefined concept or acronym, it also doesn’t seem to be used elsewhere so no need to use the acronym unless it is most commonly known by that name
Ln 44: Need a reference for this statement. The reviewer agrees the estimates often are in fact in the 2-5% range but these sources are not mentioned here. Suggest including the reference that assesses the fuel savings (on average) to be <10%. Some open literature is easily searchable /citable for 2-5% estimates.
Ln 105: risk attitude seems an unusual term, the more common scientific term in the literature on human cognition in the context of decision support systems refers to it as risk propensity
Ln 141-141: Suggest renaming STW and SOG to be velocity through water and velocity over ground, as it is contradictory to state you are taking the vector sum of speed with something else (in this case ocean current). In a similar vein, the authors state the forward speed F is a vector. Speed is only the magnitude, hence it’s recommended such quantities take on the definition /name of velocity, rather than a speed – forward velocity F in this example.
Ln 183, shouldn’t this be modulo 2*pi radians or 360 degrees?
Ln 260 Collinear in what transformation space/projection? Lines of constant bearing (rhumb line) or great circle lines?
Ln 282-285, this provided approach works for Cartesian measurements and coordinate systems, but the proposed research application is that of nautical navigation. How do the authors attend to this? At a minimum, a projection is required somewhere.
Ln 312, shouldn’t a ceiling function be used in the interest of safety of navigation? Drivers and sailors with differing risk propensities may have different agreement with recommendations if they are pessimistic vs optimistic edge weight estimation.
Ln 315 do the authors mean “the same outcome” ? Weather is highly nonlinear though so what analyses has been done to understand the tradeoffs for these two interpolation schemes in a dynamic nonconvex environment?
Ln 326 The sentence ordering makes it seem that VISIR-1 is the improvement of Dijkstra for dynamic edge weights when I believe the authors intend to credit Orda & Rom 1990.
Ln 341,347 FIFO-hypothesis is the correct English spelling.
Ln 416 “straight” by what measurement? Constant bearing/dead reckoning, or shortest path on a sphere?
Ln 665-666: From layman’s understanding, your findings confirm those of prior work in bibliographic citation [Sidoti et al., 2023] in importance considering both current and leeway for sailboat routing optimization. Can you be more specific regarding what “this” refers to when the authors state “This is, …, the first of its kind assessment”?

Citation: https://doi.org/10.5194/egusphere-2023-2060-RC2
- AC2: 'Reply on RC2', Gianandrea Mannarini, 19 Feb 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2060/egusphere-2023-2060-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2060-AC2

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2023-2060', Anonymous Referee #1, 30 Nov 2023
I recommend that this manuscript is accepted for publication with minor revisions.
The submitted manuscript describes new updates to the VISIR-2 weather routing software. The model has been developed from Matlab into Python and is available as an open-source model with no license. A brief literature review is provided, alongside an in-depth methodology, providing a detailed explanation of each component of the model. The paper finishes by demonstrating VISIR-2’s ability to minimise CO2 emissions in two case studies – one for a motor ship and one for a sailing ship.
The paper excels in its scientific reproducibility and I would like the congratulate the authors on the well-organised manuscript with impressive levels of additional detail provided, including the user manual. I believe that the modular implementation, released under a freely available GNU General Public License, has strong uses for the scientific community and the wider community. The paper excels in its presentation quality, in particular the presentation of the very clear figures. VISIR-2 benefits greatly from its validation and provides very useful and detailed insight into the model’s computation time.
That said, I believe the paper could improve in the following aspects.
I believe the manuscript would benefit from a clearer description of the inherent novelty of the model within the method section. The novel aspects of the model are not highlighted sufficiently in the method, and I finished reading the section wondering which parts of VISIR-2 form the novel features. I would also like to see this same novelty more clearly described in the introduction – here, the authors describe additional features well, but the section fails to comment on the novelty of these features in the context of the body of literature in this field.
The paper heavily focuses its novelty on the fact that their model has been developed as open source, modular and free to use (which VISIR-2 largely does with exceptional care and quality). While this is indeed useful, I believe the manuscript could benefit from a discussion on the novel quantitative implications of their CO2-saving results. (In fact, the authors also touch upon this in the first paragraph of their abstract: “…its quantitative impact has been explored only to a limited extent…”.) I don’t currently fully understand what novel academic question they try to answer with their analysis, or whether it is just used to showcase VISIR-2 (which it does a great job of). I believe that these quantitative novelties are present in the paper, but more work is needed to outline them in the results/discussion section. While outside the area of model development, the results of this paper are great work and I believe it would be of use to advance the field more generally. I would ideally like to see an explanation of how their quantitative results/discussion contribute to new science. This could also be brought out briefly in the abstract.
On a similar note, the results/discussion does a great job of highlighting the potential of weather routing as a CO2 reduction measure. However, there is almost no discussion on how their results compare to other studies in the literature. Where do their CO2 savings fit in the literature? Do they agree/disagree? What relative contribution is the paper making to this body of literature? I would like to see more discussion on this. That said, the level of detail in the results/discussion is brilliant and very commendable, a great job on that.
Some of my above suggestions have been completed to some extent in the conclusion - but should be strengthened in the other sections. No new information should be presented in the conclusion.
Alongside this, I have the following minor comments:
I believe a figure at the start of the method section that provides an outline of all steps of the model would help readers to understand the model structure more generally.

h-hat is mentioned at the end of page 5 but I cannot see this variable in Figure 1.

Removing collinear edges is a great idea, one which I will test myself!

I found Section 2.3.2 Space interpolation confusing. I would consider rewording. A better description would be useful. Same with Figure 5. Please explain the meaning of Head and Tail and give a more clear description of the two interpolation methods. Indeed, please state the difference and why the two are necessary.

The least squares fit for the blue lines in Figure 9 doesn’t seem to work. I’m not sure if an alternative is possible, but if my interpretation is correct, the solid blue line (total computation time) should not fall below the dashed blue line (Dijkstra component of computation time). In fact, this is the same for the red and green lines also.

Section 5.1 Environmental fields – should the URL next to “a lower resolution (0.4, URL)” go as a footnote?

Figure 10 b – super clear and very interesting. An engaging plot.
Citation: https://doi.org/10.5194/egusphere-2023-2060-RC1
- AC1: 'Reply on RC1', Gianandrea Mannarini, 19 Feb 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2060/egusphere-2023-2060-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2060-AC1
RC2:
'Comment on egusphere-2023-2060', Anonymous Referee #2, 09 Jan 2024

General Comments
Overall the manuscript is of high quality and provides a very thorough analysis on a proposed navigational approach for both motorboats and environmentally-driven/impacted surface vessels. Currents and leeways are compared for two separate boat models and meticulously analyzed seasonally over a scoped domain, offering high quality conclusions and results discussion, while also offering rich model bases for the open literature. Another open-source weather routing software is invaluable as prior to this, primarily only openCPN was the go-to open-source option that would not be able to handle motorboat and CO2-based measures of optimization. The primary contribution is its attention to detail and reproducibility for science computation that is wholly lacking on the open-source playing field.
Specific Comments
There are still a handful of revisions I believe the paper needs to undergo to be finalized for publication.
Throughout the manuscript, there are numerous undefined acronyms in this section that either need to be noted as a footnote, or explained to the reader. E.g., GFS, OSCAR, AVALON, GUTTA, openCPN, especially in Sec 1.1.1.
The scientific notation is very hard to follow. It is very hard to distinguish a vector quantity from a scalar. Can you use typographic convention to aid the reader? Hard-to-follow notational conventions induces increased effort in assessing the research contributions in Sec. 2.1 because of this mismatch and lack of clarity. In addition, the authors flip between radians and degrees. For consistency and legibility, they should remain the same throughout unless where deemed necessary for more intuitive understanding for the reader.
Sec 5.1: Should more than just surface current be used for the ferry? It seems the draft is in excess of 4m, so potentially 0, 2, and 4m relative z-levels could be employed for even further increased fidelity in the optimization at the expense of computational complexity.
There is a large and important question when assessing graph edges throughout the paper and that is what coordinate system/projection/transformation is assumed. This is a very important piece of information missing from a geodesy and nautical navigation standpoint.
The conclusion section is too long and should be a synopsis of the contribution and highlights of the results that a reader should and must take away from reading the publication. No new results or new discussion should be present in the conclusion.
Technical Corrections
Ln 6: A least-CO2 algorithm in the presence of
Ln 12: Two-digit percentage? Two-digit quantity? Suggest clarification on this improvement as its unclear on the units / tangibility of statements. Two-digit pounds of CO2 emission for example is not as impressive as say two-digit percentage of overall CO2 expenditure.
Ln 14: 3% shorter as measured by time, or distance? Based on the authors’ prior words “path elongation”, it is confusing to the reader to tout a 3% shorter result.
Ln 17-18: If you are using winds, then meteorology should be included in the list of knowledge bases pulled from
Ln 37: CE-Ship model is an undefined concept or acronym, it also doesn’t seem to be used elsewhere so no need to use the acronym unless it is most commonly known by that name
Ln 44: Need a reference for this statement. The reviewer agrees the estimates often are in fact in the 2-5% range but these sources are not mentioned here. Suggest including the reference that assesses the fuel savings (on average) to be <10%. Some open literature is easily searchable /citable for 2-5% estimates.
Ln 105: risk attitude seems an unusual term, the more common scientific term in the literature on human cognition in the context of decision support systems refers to it as risk propensity
Ln 141-141: Suggest renaming STW and SOG to be velocity through water and velocity over ground, as it is contradictory to state you are taking the vector sum of speed with something else (in this case ocean current). In a similar vein, the authors state the forward speed F is a vector. Speed is only the magnitude, hence it’s recommended such quantities take on the definition /name of velocity, rather than a speed – forward velocity F in this example.
Ln 183, shouldn’t this be modulo 2*pi radians or 360 degrees?
Ln 260 Collinear in what transformation space/projection? Lines of constant bearing (rhumb line) or great circle lines?
Ln 282-285, this provided approach works for Cartesian measurements and coordinate systems, but the proposed research application is that of nautical navigation. How do the authors attend to this? At a minimum, a projection is required somewhere.
Ln 312, shouldn’t a ceiling function be used in the interest of safety of navigation? Drivers and sailors with differing risk propensities may have different agreement with recommendations if they are pessimistic vs optimistic edge weight estimation.
Ln 315 do the authors mean “the same outcome” ? Weather is highly nonlinear though so what analyses has been done to understand the tradeoffs for these two interpolation schemes in a dynamic nonconvex environment?
Ln 326 The sentence ordering makes it seem that VISIR-1 is the improvement of Dijkstra for dynamic edge weights when I believe the authors intend to credit Orda & Rom 1990.
Ln 341,347 FIFO-hypothesis is the correct English spelling.
Ln 416 “straight” by what measurement? Constant bearing/dead reckoning, or shortest path on a sphere?
Ln 665-666: From layman’s understanding, your findings confirm those of prior work in bibliographic citation [Sidoti et al., 2023] in importance considering both current and leeway for sailboat routing optimization. Can you be more specific regarding what “this” refers to when the authors state “This is, …, the first of its kind assessment”?

Citation: https://doi.org/10.5194/egusphere-2023-2060-RC2
- AC2: 'Reply on RC2', Gianandrea Mannarini, 19 Feb 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-2060/egusphere-2023-2060-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2023-2060-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Gianandrea Mannarini on behalf of the Authors (27 Feb 2024) Author's response Author's tracked changes Manuscript

ED: Publish subject to technical corrections (04 Apr 2024) by David Ham

AR by Gianandrea Mannarini on behalf of the Authors (09 Apr 2024) Author's response Manuscript

Journal article(s) based on this preprint

24 May 2024

VISIR-2: ship weather routing in Python

Gianandrea Mannarini, Mario Leonardo Salinas, Lorenzo Carelli, Nicola Petacco, and Josip Orović

Geosci. Model Dev., 17, 4355–4382, https://doi.org/10.5194/gmd-17-4355-2024,https://doi.org/10.5194/gmd-17-4355-2024, 2024

Short summary

Gianandrea Mannarini, Mario Leonardo Salinas, Lorenzo Carelli, Nicola Petacco, and Josip Orović

Supplement

https://doi.org/10.5194/egusphere-2023-2060-supplement

Data sets

Raw data for replicating VISIR-2 runs of this manuscript Mario Salinas https://zenodo.org/record/8321216

Intermediate products (routes) to expedite reproduction of both figures and tables in Sect.5 of this manuscript Mario Salinas https://zenodo.org/record/8233874

Model code and software

Source code of VISIR-2 Mario Salinas, Lorenzo Carelli, Gianandrea Mannarini https://zenodo.org/record/8305527

Video supplement

Ferry optimal routes for whole 2022 Mario Salinas https://av.tib.eu/media/62912

Sailboat optimal routes for whole 2022 Mario Salinas https://av.tib.eu/media/62913

Gianandrea Mannarini, Mario Leonardo Salinas, Lorenzo Carelli, Nicola Petacco, and Josip Orović

Viewed

Total article views: 610 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
421	166	23	610	30	20	14

HTML: 421
PDF: 166
XML: 23
Total: 610
Supplement: 30
BibTeX: 20
EndNote: 14

Views and downloads (calculated since 16 Nov 2023)

Month	HTML	PDF	XML	Total
Nov 2023	24	8	2	34
Dec 2023	77	21	5	103
Jan 2024	105	33	3	141
Feb 2024	84	26	7	117
Mar 2024	49	32	2	83
Apr 2024	51	35	4	90
May 2024	31	11	0	42
Jun 2024	0
Jul 2024	0
Aug 2024	0
Sep 2024	0
Oct 2024	0
Nov 2024	0

Cumulative views and downloads (calculated since 16 Nov 2023)

Month	HTML	PDF	XML	Total
Nov 2023	24	8	2	34
Dec 2023	77	21	5	103
Jan 2024	105	33	3	141
Feb 2024	84	26	7	117
Mar 2024	49	32	2	83
Apr 2024	51	35	4	90
May 2024	31	11	0	42
Jun 2024	0
Jul 2024	0
Aug 2024	0
Sep 2024	0
Oct 2024	0
Nov 2024	0

Viewed (geographical distribution)

Total article views: 609 (including HTML, PDF, and XML) Thereof 609 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 05 Nov 2024

Download

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Preprint (32299 KB)
Metadata XML

Short summary

Ship weather routing has the potential to reduce CO₂ emissions, but it currently lacks open and verifiable research. The Python-refactored VISIR-2 model considers currents, waves, and wind to optimise routes. The model was validated and its computational performance is now quasi-linear. VISIR-2 yields, for more than ten days in a year, two-digit savings for a ferry sailing in the Mediterranean Sea. Sailboat routes with wind and currents can be optimised as well.


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