How to speed up weather routing

Dr Henry Chen, a retired naval architect, looks at where weather routing has gone wrong and how to improve it.

Recent heavy weather related accidents involving the loss of thousands of containers overboard and ships sinking have created far reaching environmental damage.

It is easy to blame global warming as the cause for increased storm activities and thus let the insurance companies handle sharing millions of dollars of losses hoping such events will not happen again. Unfortunately, the past two decades of recurring accidents seem to indicate that while weather forecasting, computing and communications technologies have improved significantly, the effectiveness of weather routing still lags behind.

Starting from meteorologists practicing the simple art of storm avoidance using 500 millibar upper air and surface pressure charts to infer the direction of storm movement in the 1950s, weather routing technology has evolved into a full-fledged voyage optimisation suite of tools, utilising long range environmental forecasts to save fuel for just in time arrival at the destination. Ships at sea can receive the latest information along with route advisories via satellite broadband communication. While the software can generate beautiful weather graphics of winds, waves and currents, they often cannot relay the accuracy of the forecast and its impact on the voyage plan, surprised by drastic changes in weather from the earlier forecast after the voyage has started.

Major centers around the world produce forecasts beyond 15 days, but acceptable accuracy for wind and wave conditions around major storms remains around three to five days, and even shorter for tropical storms for ship routing purposes. Although each service provider touts their superiority of technology and service, it is difficult to consistently prove that one approach is better than the other when forecast uncertainties can turn a perfectly optimised voyage into a disaster. Furthermore, the widely accessible forecast data on the internet lowers the entrance barrier for the business for startups and disrupters, resulting in price wars and a fragmented market share. Instead of developing better ship models and optimisation algorithms, traditional service providers spend money on marketing and sales efforts, hoping to compete. The recent acquisitions of weather routing companies by classification societies, engine manufacturers and multi -industry weather companies only puts pressure on profitability of the newly acquired business unit. They have to compete for funding in a large organisation whose goal is to sell its main products and services. On the other hand, startups are finding it increasingly difficult to find venture capital funding as the market is not readily scalable and shipping companies are very conservative in adopting new technology without consistently showing high returns on investment as promised.

There are plenty of new enhancements for weather routing as it evolves into a tool for holistic voyage optimisation and fleet management to improve safety and efficiency. Here are some examples:

  1. Full ship motion predictions instead of limiting sea states as constraints in the routing algorithm to avoid parametric/synchronous roll resonance for container or tank sloshing for LNG ships.
  2. Low cost, automated high frequency shipboard operations data acquisition system using IoT devices.
  3. Real-time monitoring/warning APPs for safe navigation.
  4. Meaningful fleet KPIs to help find causes of performance degradation.
  5. Quantify the uncertainties of weather prediction using ensemble forecasts from the European Centre and US National Oceanic and Atmospheric Administration.
  6. Improved algorithms to find the optimum route and speed profile for a range of ETAs to avoid fuel waste by integrating into loading/discharging scheduling.
  7. A risk-based route simulation/selection tool utilising ensemble forecasts to support shore-side and shipboard decisions.
  8. A virtual ship operations center to host the data and display KPI on performance efficiency, regulatory compliance and safe navigation decision supports.

Except the first two items, the rest requires the development of operational digital twins in order to achieve full benefits. It would require a dedicated team of meteorologists, naval architects, engineers, software developers and data scientists with feedback and guidance from ship operators. Although some of the large ship operators and managers can afford and have already started this R&D effort, the speed and cost-effectiveness can be greatly enhanced if the industry as a whole works and benefits together.

We are living in a 4-D world: Decarbonisation, decoupling and disruption of the supply chain have already drastically changed the traditional shipping business environment. The industry is coming together for the digitisation effort as a cost reduction measure. We also need to improve safety aspects as the increased storm activities and a shortage of experienced seagoing crew as well shore side management talent have resulted in many recent incidents. Unless steps are taken to mitigate the risks by improving the technology of voyage optimisation to assist ship operators, managers and owner/charterers, accidents will happen again in these treacherous water


  1. Meteorologists, naval architects, engineers, software developers and data scientists are not sailors. The more ‘experts’ they are in their field, the further they are from any nautical skill.

    1. I am no expert, here is my humble opinion on what happened: Today’s mega container ship designs feature a wide beam and large bow and stern flares in order to carry more containers above the load waterline, while still minimizing resistance with a streamlined underwater hull. In moderately high head or stern seas, the stability varies due to the changing water plane area as the position of wave crests travel along the hull. When the bow is down due to moderate pitching coupled with a slight roll, the large flare is fully immersed in the wave crest. The restoring buoyancy force plus the wave excitation force “pushes” the ship to the other side when the vessel is very “tender” due to changing stability at this instant. A similar action will happen on the other side as the bow pitches down in the next cycle. This repeated pumping action, which can lead to increasingly large roll angles within a few cycles even in moderately high head or stern seas, is called “parametric roll”.
      These unexpected events are contradictory to normal seamanship practice. While captains are trained to head the vessel into the sea to reduce synchronous rolling when close to the natural period of roll, doing this only exacerbates the situation and frequently leads to container damages or lost overboard due to excessive accelerations and lashing failures. Furthermore, design criteria of lashing system on these ultra large container ships are often less stringent than those for smaller container ships because of their huge size. However, once a vessel is in roll resonance, the wave energy will keep pumping into the system, resulting in excessive accelerations as great as the smaller size container ships. Unless mitigating strategies are implemented in the routing algorithm to avoid such conditions and provide real-time warning and seakeeping guidance to officers on watch, such accidents will continue to occur. So working together with all concerned parties hopefully will help the industry.

      1. You have provided Sir an excellent comment very accurately describing both phenomenons : parametric rolling and rezonanse rolling. But You may have a perfect routing system and perfect everything else and ships will loose boxes.

        The reason is very simple . The vessel must be loaded as per instruction contained in : a) stability information and b) container securing manual. .
        In case of b) being not achievable due to impossibility of performance caused by many factors , master /choff may use ( and is using) class approved loadicator equipment and software to calculate conditions not complying with above mentioned manuals.

        The validity and accuracy of the results obtained by loadicator depend on the accuracy of declared by stevedores stow plan and on the accuracy of shippers declared weights. Unfortunately these values are often wrong due to a) mistakes/errors, b) intentional cheating .

        Hence one may conclude that the overall result = GARBAGE IN = GARBAGE OUT. Nobody cured the disease by attacking symptoms but only by eliminating the causes . In this case one of the root causes.

        The voyage of the container starts with stuffing depot in some remote location . One may imagine a voyage plan of a box with multiple way points .

        The way points from stuffing depot to ships rail need to be fixed/repaired first before Your excellent weather systems can be useful and reliable. As i understand, You are interested in fixing -not in mitigating . One can not fix the problem in the middle of the problem . The fixing must start at the source/begining. cheers.

        1. AND I WOULD LIKE TO THANK THE EDITORS TO ALLOW ME TO CHIP IN WITH MY GIBBERISH, DRIVEL AND NONSENSE . Stay assured though that if I am not welcome or too controversial or irritating, just simply send me a message to my email : we will not tolerate your crap anymore , so please leave.
          Stay assured You will see me no more. Rgds.

  2. Yet, this is the time of “information management” – us generalists by experience need to comprehend, balance and digest multi aspect-facts provided by ‘the experts’, before blending our decision with gut feeling and experience

  3. Weather routing has never done me any good. It places me in the difficult decision of following my own judgement and experience, combined with a weather forecast, as against weather routing from ashore that the chartered has paid for. In situations like that you’re in a “Damned if l do, Damned if l don’t” situation. Weather routing also undermines a Master’s meteorological learning curve, since at a minimum he has to follow the charterer’s instructions.

    1. That is absolutely correct and accurate description of a situation the preset day ship-master is in. Damned on all points of the compass if ordered by management or charterers to strictly follow the weather routing advise. You nailed it Sir not 100% but 1000% +++

  4. We all need holistic voyage optimisation……….
    Don’t we?
    What happened to common sense on the ship – altering course and/or speed according to the weather encountered or anticipated?
    Look no further than the livestock carrier lost off Japan last August – due to a typhoon whose track was accurately forecast a week in advance.

    J. Anderson
    Retired MM

  5. Just to remind that weather management is now governed by the much real “nowcast”(coming from Artificial Intelligence, Machine Learning, Deep Learning, IoT, etc) offered to Ship-master.
    But he has to direct and play the song that the composer imagine, and the music is nonsense when starts with stowing thousands of containers without fixed guides through the whole Pacific ocean (i.e. Aleutian Is.) and taking for sure that manual stowage by shore workers do not failed.
    Composers (read architects, designers, IT developers and engineers) must approach who plays the song for the solution of many problems today, or directly reciclying themselves.

    1. ………………………..offered to Ship-master.
      But he has to direct and play the song that the composer imagine, and the music is nonsense ……….
      What a perfect and artistic description of a problem. Sir, You have my deep respect for such an eloquent explanation . I may be wrong of course but I would substitute the word” imagine” with what the composer is seeing in “virtual reality” , which does not necessarily and accurately describes” reality” . And that is surely a pure nonsense.

Back to top button