Routing: how to optimise your route based on weather predictions

One recurring question among the novice offshore sailor is how to optimise your route when sailing. The practice is known among sailors with the French name of Routage or the English name of Weather routing. We do not want to write an article that is too technical here because it partly depends on the navigation software you use. However, we will try to understand the fundamentals so that you can find your route with any system.

Route Optimization - Navigation Software
Route Optimization – Navigation Software


The software to optimise the route

To optimize the route you will inevitably need dedicated software. This is why this is usually a topic for long distance racers. Mobile apps have also appeared recently so you can try them out to see if you are comfortable with them. The best known navigation software among professional sailors are the following:

Optimising the route: understanding the problem to be solved

Let’s try to get to understand how “Weather Routing” or “Routage” is done step by step. First of all, let’s ask ourselves what is the problem we need to solve to optimise the route. We have to go from point A to point B in the shortest time possible. The variables are the intensity and direction of the wind and obviously our boat. Initially we can see it as an instantaneous problem. That is, in relation to the weather conditions we are observing at that precise moment.

Route Optimisation - The Instantaneous problem
Route Optimisation – The Instantaneous problem 

However, unless you have to go to a rounding mark, the problem is not instantaneous. On a long course the boat performance details is the only variable that remains constant (save for damage to sails etc), but the variables wind direction and intensity change over time. For this reason we have to divide our potential route into many steps. Starting from point A we will be able to navigate in a number of directions that fan out in front of us. All the points that I could reach with my boat in a defined period of time is defined an isochrone. For simplicity we say that the first isochrone defines all the points that I can reach after an hour of sailing.

Route Optimisation - The Problem over time
Route Optimisation – The Problem over time

At that point, the software, starting from a series of points on the isochrone, repeats the exercise from each point. From each point a range of new possible routes will open up. The maximum distance boundary given by the sum of the navigation in the first and second hour defines the second isochrone. Proceeding by successive steps, the navigation software will arrive at its destination. Working backwards through the series of segments that defined the fastest route we have the outline of the optimal route. It is an empirical solution to the problem, because trying a mathematical solution would be impossible.

The input factors to optimise the route

In order to make this empirical calculation of the optimal route there some essential input data.

  • Wind direction and strength
  • The speed of the boat relative to the wind

Then, to improve the result, we could add advanced input data.

  • Surface currents
  • State of the sea
  • The choice of sails
  • The crew

Some of these factors are exogenous, such as the surface current adding to or subtracting from boat speed. Others are factors affect boat speed. If the sea is rough we will not be able to maintain the maximum speed that we would have in a flat sea. Similarly, if we have torn a sail that we need we could be penalised. The crew also matters, for example we may have to decrease the expected efficiency if we sail single-handed. We need these additional factors to optimise the route and obtain a more reliable result.

Wind direction and intensity in optimising the route

It is obvious that this data is essential to optimise the route, i.e. to do weather routing or routage. The input files to be provided to our navigation software are called GRIBs (GRIdded Binary). This is a specific format used in meteorology, which defines space-time grids. For each time point, and each point of the grid, the intensity and direction of the wind is indicated. The GRIB file in its visual representation is a wind map that we talked an other article.

Route Optimisation - Input Data - GRIB files
Route Optimisation – Input Data – GRIB files

A GRIB file is specific to an area of the earth’s surface and contains a defined number of snapshots. Many systems that can be used to download the files, such as for example  Saildocs which allows you to define precisely what data you want to download. Saildocs was created for low-bandwidth connections via SSB radio and allows you to downsize the required file by reducing it to a minimum omitting less important information. Other services, on the other hand, download large areas with many snapshots, creating files that are too large to download via satellite. With Saildocs, via email, you send a request specifying the parameters of the request, and you receive the GRIB in response.

You can specify the extremes of the rectangle of the geographical area of interest as well as the precision or grid definition. For example indicating whether the grid should have a point every degree, half degree or quarter degree. The time sequence of the analysis and of the forecasts to be received is indicated below. For example by indicating 0 (the analysis) followed by 3,6,9,12,15,18,21,24. This gives us a photo every 3 hours for 24 hours. But Saildocs is very flexible and we can adapt it to our needs.

Boat speed in route optimisation

The speed of the boat is defined by another input file that describes our Polars. This is nothing more than a table in plain text format with the extension “pol“. The table indicates for each row our performance at increasing angles to the true wind. For each column we have increasing wind intensity. The reference is always true wind, so the software will find a point for our 60 degree wind angle with 10 knots of air.

Route Optimisation - Input Data - GRIB files
Route Optimisation – Input Data – GRIB files

It is important that the polar file represents the actual speeds expected for the winds indicated. The boat designer usually provides calculated theoretical data called VPP. The acronym stands for Velocity Prediction Program and usually overestimates actual average speeds. The boat straight out of the mould is much lighter and the program will be based on its empty weight with basic equipment.

For this you should adapt or even better calculate your own Polars. To calculate them you need to record navigation data, navigation software like Adrena have packages to do it. Polars, if represented visually, are curves, one for each wind intensity with an interpolated line connecting the points in the file

Surface currents in optimising the route

In the Mediterranean you can ignore them but surface currents are very important in many parts of the world. We can divide them into two types, permanent currents and tidal currents. Permanent currents are for example those of the Gulf Stream.

Route optimisation - Input data - Surface currents
Route optimisation – Input data – Surface currents 

Tidal currents are instead dependent on the lunar attraction of the terrestrial water. Their cycle is regular and mathematically predictable with an excellent approximation. For this reason, it is possible to buy files that can be integrated with our navigation software for the areas concerned. It must be said that currents vary slightly with atmospheric pressure but would be a difficult level of detail to integrate.

Route optimisation - Input data - Surface currents
Route optimisation – Input data – Surface currents 

The choice of sails and the state of the sea

Only some very advanced navigation software such as Adrena allow sail choice to be taken into account. In fact, if we recorded our polars by indicating our configuration of the sails and the state of the sea to the navigation software, we could create multiple files. When we go to optimise the route we will be able to indicate to the navigation software which sea conditions we expect. If we have torn a sail, it is possible to indicate it and if our performance suffers take it into account.

The crew in optimising the route

The crew factor is a factor that complements the experience. All navigation softwars allow you to indicate an efficiency factor with which to use the Polars. For example we could indicate that single-handed we sail at 90% of the polars compared to when we are fully crewed.

The sum of all factors

In this example we see how to optimise the route between Bermuda and Newport. The result takes into account wind, boat and Gulf Stream. It can possibly be corrected by choice of sails and crew.

Route Optimisation - The sum of all factors
Route Optimisation – The sum of all factors

Optimising the route: a practical example

The first step in optimising the course is to superimpose polars on course. Given the starting point, calculate the first range of possible routes up to what we have called the first isochrone. From each point of arrival we repeat the exercise arriving to calculate the second isochrone.

Route Optimisation - The polars superimposed on the route
Route Optimisation – The polars superimposed on the route

The optimal route will be none other than the series of segments that reaches the target waypoint first. I am talking about segments because in the calculation we will have to define a time interval of the isochrones. The shorter the time interval the greater the precision in optimising the route. However, the computing power required to obtain a result is also greater. On a short route this doesn’t matter much, but on long distance simulations we have to take it into account. In order to optimise the route over a long distance, the isochrones can be every 6 hours. On a short route every half hour or every hour at most.

Route optimisation - Isochrones and the optimal route

Optimising the route: the main limitations

Optimising the route suffers from a whole series of limitations that the expert will be able to take into account:

  • Inaccuracy of GRIB file predictions
  • Unrealistic polars
  • Space / time interpolation of GRIB files
  • Average conditions versus instantaneous conditions
  • Any local phenomena not modeled
  • Data not available on surface currents

→ The calculated route is only an indication

Examples of input data limitations

The navigation software is based on the input data that we pass to it therefore inevitably the worse the quality the worse the result. In another article we discussed the limitations of wind maps or grib files. In particular, we talked about problems related to light winds, which are often guesstimated by the model. We also talked about space-time interpolation. By this we meant to emphasise that wind fields do not indicate wind shifts when a front passes. Instead, they present us with data in progressive rotation that does not correspond to reality. When we go to optimise a route we have to take this into account.

Passive strategy

Without any information on the weather forecast, travelling from A to B we will have to follow a passive strategy. In the example in the middle of our route the wind rotates from South-East to South-West. This happens suddenly when a warm front passes. If we make no attempt to optimise the route, we will sail on a reach on the first part of our leg. In the second part we would find ourselves with the wind dead from behind having to jibe repeatedly.

Route optimisation - Passive strategy
Route optimisation – Passive strategy

The route calculated by the navigation software

In an example like this, the software will observe a progressive rotation of the wind from the input data and its possible interpolations. This is indicated with the smaller arrows while the bigger ones indicate the real wind which, as mentioned, rotates suddenly when the front passes. With this data the navigation software will calculate an optimal route which will define a curve towards the finish line. If we followed it to exactly we would find ourselves asking why the wind is not progressively rotating as predicted. If we have not studied the synoptic map we will hardly understand that we are due to see the passage of a front.

It is clear that blindly following the route indicated by the navigation software would be a mistake. If we have studied the synoptic chart we would already know what the software is trying to tell us. If we haven’t studied it, we can instantly make a deduction by observing the suggested optimal route.

Route optimisation - Route calculated by the software
Route optimisation – Route calculated by the software

The optimal route

With a wind shift like the one indicated, the optimal course is in fact a two segments broken route. That is, in order not to find ourselves with the wind from dead astern in the second half of the race, it is better to keep more to the left initially. Doing so at the passage of the front we will have an excellent angle to the wind to get to the finish line. The experienced sailor will understand what the software meant with it’s optimised route, but wont follow it literally. By understand the dynamics of the passage of the front the skipper will interpret the result and opt to keep to the left on the first half of the leg.

The overall strategy need to also take into account what others are doing. You should not go off on a flyer but rather stay with the group of boats that have anticipated the front passage and started the leg bearing away slightly off course downwind. A relative strategy, neither passive, nor absolute, taking into account the race and the competitors.

Route Optimisation - The optimal route
Route Optimisation – The optimal route 

How to improve the input data

To get a reliable result we can intervene to try to minimise the calculation error:

  • Updated GRIB files (and high resolution)
  • Realistic polars on data collected during actual navigation
  • More polars for different sea conditions
  • More polars for crew or single handed navigation
  • Software with sails management
  • Data relating to known currents and tidal currents
  • Verification of output data and “back testing

Interpret the results of the Software

Route optimisation - Check the software data
Route optimisation – Check the software data

The result provided by the software must always be interpreted. The difference between a great sailor and a beginner lies in the ability to interpret the data. In navigation I always asked myself “what is it that the software is trying to tell me”? That is, by observing the result of the optimal route you must make the effort to understand why you deviate from the shortest course.

Furthermore, we cannot do the races without taking into account what others are doing. Totally relying on software and taking a different route from everyone else rarely pays off. When it pays it is often the beginner’s luck, because the experienced racer had the same information available. If they have not chosen your route you have to ask yourself why not. Here are some tips to always keep in mind in every race:

  • Check the positioning of the competitors
  • Avoid extreme routes
  • Stay with the group
  • Chose a strategy relative to others
  • Evaluate the risk / benefit of the choices


Route optimisation - Headed for Cape Horn
Route optimisation – Headed for Cape Horn 

It is the skipper who makes the difference

Despite the very advanced software when it comes to optimising the route, it is still the skipper who makes the difference. Here are some summary tips after many real life situations, many mistakes, and over 60,000 miles of racing:

  • Know the weather and the optimal management of the boat in relation to it
  • Know how to manage available resources (boat, energy, sleep)
  • Know how to manage the information available
  • Know how make the most of the tools available to you, such as the route calculation software
  • Know in depth the limitations of input data, models, and tools
  • Know how to identify the route with the best risk / benefit
  • Know how to position yourself with respect to competitors, relative strategy and how to cover others
  • Know how to change strategy while minimising losses if the situation changes
  • Be able to identify the signs of an wrong weather forecast
  • Train and prepare – also in the use of software and polar data preparation
  • Etc.

If you are not sure
→ take the shortest route
→ sail fast

Route optimisation - Headed for Cape Horn
Route optimisation – Headed for Cape Horn