How to make optimal use of weather data, while keeping a close eye on the opposition, by offshore racing navigator Will Oxley.
IN this age of GPS, sophisticated navigation software and the internet, the role of the racing navigator has shifted dramatically towards fine-scale weather forecasting and boat positioning.
Consider the following hypothetical scenario for a race from Sydney to Southport. It's 1730hrs. We have had a great afternoon of racing in the last of a south to south-easterly push and the breeze is now easing and backing into the east as the new high pressure system ridges across the NSW coast.
Looking at the boats around us we can see we are doing well, but now the fun begins. Should we be inshore or offshore tonight? What about the current? What is the opposition doing?
The weather models have done a good job of predicting the general trend, but as usual things have happened quicker than predicted. The CSIRO SST chart downloaded before the start, suggests that current is negligible until Seal Rocks, then flowing south at 1-3 knots through to north of Byron, with the maximum flow off Smoky Cape. The imagery shows fingers of cooler water right in on the beaches, suggesting some relief from current on the bricks.
I have been looking at the NSW coastal observation stations for the last few weeks and last time we had this weather pattern, the data recorded showed a switch from a light onshore wind to a land breeze between 2300-0100hrs, so I am expecting a similar pattern tonight.
I log on and check the coastal stations and write in the ship's log the last three hours of observations. I check my timeline. The 00z US GFS GRIB file was available at 1500 and the high resolution meso scale WAM charts should be available now.
The next surface chart will not be available till 1845 and the evening Quikscat pass will not be available until 1900. I'll have to log on later to grab these.
For now I open UGRIB and download the latest GRIB, then download the meso scale WAM images.
I examine the GRIB file in UGRIB, stepping the images back and forth, to see what is predicted overnight for the NSW coast. I then open the GRIB file in Expedition and run a route using our polar. Given our proximity to the coast the routing solution is “pretty average” (ie, wrong!). Nonetheless I note down where we are expected to be at several key times during the night.
I have made up my own current GRIB that morning using the latest CSIRO image and I incorporate this file into the routing. I observe that the routing solution has us hugging the coast right up to Byron Bay.
I look at the predicted wind speeds and directions overnight: the speeds look pretty optimistic but at least the model seems to have The charts also show that well offshore the easterly wind holds in until about 18Z (0500am). “In or out” is not going to be a clear cut decision.
I go over all the data again, making some notes in my “wet notes” book. It is now 1805hrs. I wait five minutes (making a full log entry while I am waiting) then log back on and record the latest coastal observations.
Interestingly, the wind speed on the key coastal stations has dropped 5kts in the last hour while we still have a healthy 14kts eight miles offshore. I compare this new information with our game plan from before the start. The evening Quikscat pass at 1900hrs is going to be very helpful in resolving our strategy. Let's hope we are not in a blank spot on this pass! I make more notes, then go on deck to brief the watch leaders on how I see the night unfolding.
Working in isolation
In the scenario above, I have tried to give you a small insight into just what your navigator might be doing down below while the rest of the crew is busy on deck!
One drawback of being a racing navigator is that you seldom get the chance to race with another navigator. As a trimmer or a helm there is plenty of opportunity to watch and learn from others, but for a navigator you are it.
In order to progress my craft I read extensively and talk with others about how they deal with particular navigation/meteorological challenges. There are plenty of books around describing the traditional art of navigation and weather forecasting, but almost nothing about making the most of the opportunities that arise as a result of increased access to the internet, before and during a race, and so that is the focus of this article.
Where to start?
The first thing is to ensure that you have a reliable onboard internet link. For races along the Australian east coast this is best achieved by a Telstra Next-G connection. Teething problems in the change from CDMA to Next-G have been well publicised, but you should now be able to get a good connection up to 40nms offshore with a good setup.
I have had the most success with a professionally installed Ericcson W-25 fixed wireless terminal and a decent marine antenna.
Further offshore you will need another solution to continue to access the internet. This will invariably be much more expensive and slower, so I will restrict this article to how best to make use of a Next-G connection.
As a navigator a key question is “what am I allowed to access during a race?”. Read the Notice of Race and the Sailing Instructions for advice on this.
The relevant ISAF Racing Rule is No.41 Outside Help. The rule says: “A boat shall not receive help from any outside source, except: d) help in the form of information freely available to all boats.”
I have heard many liberal interpretations over the last few years on just what “freely available” means.
Fortunately the term was interpreted in Royal Yachting Association Case number 5, 2005 (See Box 1) and by US Sailing interpretations Question 931
The jury at the 32nd America's Cup in September 2006 also had cause to rule on a question regarding Rule 41 and they referred to the RYA interpretation in their response.
Basically the interpretation is that it's okay to pay for the internet connection but there should be no further cost to obtain the information. US Sailing goes further and says that in addition, “information freely available” means that it is easily obtained by all boats in a race. They define “easily obtained” as meaning that the information is available from public sources that competitors can reasonably be expected to be aware of and can locate with little effort.
This clarifies that “free” websites which magically appear before a race with weather information are NOT allowed.
There is an incredible number of websites providing weather information and it is very important to focus on a limited number that will provide you with the information that you need. Do your web searching for all the sites you will access well before the race.
I break down what I need into a number of categories as follows:
1) Near real time observations and radar 2) Government forecasts
3) Charts (analyses, prognoses) and 4) Weather models
4) GRIB files
5) Fleet information (news, positions)
I then set aside some time for a series of web searches and trawling through known websites to find useful images or data. Next I paste the URL directly into a word processing document, under one of the headings above and write a description beside it. This is a long process but gets much easier as you build up links which can be used in other races.
The benefit of putting the direct link into the document is that you can go straight to the page rather than through the home page. This saves time and it focuses the mind when you are tired and can't remember where to get the data you need.
If possible note down the time that the page updates so that you only revisit when it is updated and so don't waste time and money.
Having found a load of web links, I pare these down to what I consider the essential URLs. I then save the word file as a “single file web page” with the name and year of the race. Once the web page is open I can access the URLs directly. I generally use a right click and open the file in a new web page. I then save the images and use a viewing program like ACDsee to open them.
A big advantage of ACDsee is that I can open up to four images at a time and view them all on the same page. This allows me to examine particular weather features in some detail (see Figure 1).
This takes plenty of time and effort each occasion you do a major download. However, if you have some programming skills it is possible to write simple scripts to automate much of the work and this dramatically speeds up the process.
Create a timeline
I put together a 24-hour timeline for every race I do. This timeline includes radio sked times and update times for GRIBs, weather charts, Quikscat I also maintain a backup timeline for when I am outside of internet range. This lists transmission times for weatherfax images and voice weather forecasts.
These timelines mean that at any point during a race I know where I can get the next useful piece of weather information. It also allows me to manage my time so that I'm not napping just when there is a new GRIB file available.
Access to GRIB files
The one meteorological model that is readily and freely available in Australian waters is the US GFS model. It is available in three-hour time steps forecasting out to 180 hours. The model is available in several resolutions. Use the 0.5 degrees resolution model if possible. The model is run every six hours (00Z, 06Z, 12Z and 18Z) and generally available 5-6 hours after initialisation. This means the 00z run (1000am EST) can be accessed between 1500-1600 EST depending on where you source it.
There are a variety of ways to access GRIB files for navigation/routing programs such as Expedition , Maxsea or Deckman. My favourite is an excellent piece of software called UGRIB. The program and downloads are completely free.
UGRIB facilitates efficient downloading, then viewing of the high resolution (0.5 degree) GFS model. The downloaded file can also be imported directly into any of the navigation programs listed above.
One big advantage of UGRIB is that it allows you to download precipitation data, in addition to the standard 10m winds and surface pressure fields. This really helps in estimating the position of fronts and trough lines.
A backup option for downloading GRIBs is to use the email robot Saildocs. Again, this is completely free and uses the high resolution GFS model.
Complete information on how to use Saildocs can be obtained by sending a blank email to firstname.lastname@example.org.
In recent times both these options have become accessible through Expedition and Deckman.
Tracking the opposition
Many yacht races around the world now have websites where followers can track your every move. Of course this means that if you have access to the internet, then you can also track your opposition from onboard. For the east coast races in Australia, the Cruising Yacht Club of Australia has an excellent system that updates every ten minutes or so. The standard yacht tracker is very graphically intensive but there is generally a text web page of the standings available which is much easier to access on a slow link.
In my opinion this frequent access is a double-edged sword. The advantage is that you can keep a very close eye on your opposition and, using the SOG and COG data, infer the conditions that others are experiencing. On the other hand it is then harder for a yacht to make a breakaway from the pack and this can shut down some tactical options.
The other negative is that the crew ask for this information on a very frequent basis, which means it can be hard for a navigator to get any rest!
When I have time, I type position information for a subset of the fleet into Expedition, but this is a laborious process. In contrast, many of the world races send this information to each yacht on a regular basis and also make the data available as a small downloadable text file on their websites.
With a bit of initial fiddling, this file can be directly imported into Expedition, Maxsea or Deckman. If the CYCA could adopt this practice it would greatly enhance the value of the data to the fleet. It would also allow avid shoreside navigators to make more use of the wealth of information now available.
Using the information
Before any race you should have spent some time watching the lead up weather and the forecasts so that you can I like to have the big In the days leading up to the start I download the GFS GRIB (and any other GRIBs I can get my hands on) and using my routing software I run through the race. I look at the effect of things moving quicker or slower or the winds being stronger or weaker than expected.
This continues to build my Before leaving the dock on race day I solidify my thoughts and present them to the afterguard as a detailed forecast and view of how the race will unfold. This allows us to make any final decisions on sail selection, watch systems, gear on board, etc. If these briefings are interactive I am then forced to justify and refine my thinking, which improves our “race plan”.
I then give a less detailed briefing to the entire crew as we motor out to the start.
During the race
Once the race starts we begin to execute the plan, taking note of how our key opposition are setting up. My job now is to continue to monitor and refine the race plan. Is the wind stronger or weaker than forecasted and is the direction as expected? If not, why not?
I continually monitor coastal observation stations. What is the wind strength and direction, how relevant is that observation station, etc? Past experience, expert advice and reference to the Bureau of Meteorology (BOM) website can provide valuable information on observation stations.
Better still, visit the observation stations if you are ever driving past, and make some notes. (My wife thought I had lost the plot while driving round the North Island of New Zealand after the America's Cup, when we stopped and clambered up hills to view some of the weather stations I had used during my time as weather coordinator at Victory Challenge!)
If there is any rain around, keep an eye on the sky and on the weather radar to make the best use of the outflow from cells and to avoid getting trapped in their lee where possible. If the rain is moving offshore or up the coast, have a look at the weather observations as the cell passes that area. This will often give you a much better clue as to what wind speeds are expected with the approach of the rain.
Every six hours download a new GRIB and predetermined set of data, review the strategy and make sure you are still on track. In the unlikely event that it has all gone pear-shaped, don't be afraid to restart your thinking about how best to get from where you are now to the finish line.
Predicting wind strength
Another key question asked of the navigator is “how much breeze will we get”? This is especially relevant when wind speeds are getting near a crossover on the sail chart.
It has been well drummed into us by the BOM that coastal forecasts talk about averaged 10m winds while gusts can be up to 40 per cent stronger. However, if the Bureau forecast is for 20-25kts and you come on deck and say “max wind 25kts with gusts to 35kts” (+40%) it can make it pretty hard for the afterguard to choose the right sail combination!
I use a number of different sources to come up with the expected mean and max/min wind speeds. As a rule I look at the latest GRIB and meso WAM images and note their winds for the next 6-12hrs. For the GFS 10m GRIB data, I add my own fudge factor to account for mast height, the type of weather expected, and where we are in the world.
One way to deal with this is to use a general formula: TWS(@masthead) = TWS (@10m) x (h/10)a, where h is masthead instrument height in metres and a is a constant which is normally in the range 0.11 to 0.14. This is recommended by Nick White, the brains behind Expedition.
If I have access to information on the wind speeds in the upper level (eg, 925Hpa) I use these data to provide a good clue to maximum speeds that might be expected in the gusts. I also look at the wind speeds recorded by Quikscat to check that they gel with the forecast numbers.
Finally, I do always try to look at the current BOM forecast just to make sure I haven't missed anything. I then come up with my best guess!
The navigator is the glue that holds the watches together on a boat, ensuring consistency and optimum performance. Programs such as Expedition and Deckman provide excellent graphing facilities and these should be monitored closely ð especially when there is a gear change or change of helm. The navigator is in the best position to judge whether a change has been positive or negative and the results should be reported to the deck, using as much diplomacy as possible if the results are not good!
After the finish the navigator should calculate the corrected times to determine how you have finished against the boats ahead (if there are any) and how close your competition has to be to beat you. Good news I hope!
Will Oxley is a racing navigator and marine ecologist who has completed more than 200,000nms including three round-the-world races, ten Sydney Hobart races with one line honours victory on Skandia and major international and local regattas including IRC victory at last year's Hamilton Island Race Week aboard Yendys.
Supplement- Further Cases, January 2005 to June 2007.
Question 3: Is weather information sent to a mobile phone, to a receiver or to a computer by a weather bureau as part of a dedicated subscription service “freely available” for the purposes of Rule 41(d)? Is the cost of that service relevant? Is information available to all on the internet “freely available”, given that a subscription has to be paid to an internet service provider?
Answer 3: Once a subscription has been paid to a generally available and non-specialised communications service, such as an Internet Service Provider, a telephone service (mobile or terrestrial) or a television licence, any information that is then available to the general public, or is available to all competitors in the event, and that can be accessed readily and at no further cost (other than the cost, if applicable, of a standard rate call or connection) is “freely available”. The Notice of Race and Sailing Instructions may change Rule 41 to widen or narrow this.
1. www.marine.csiro.au/remotesensing/oceancurrents/SE/latest.html (accessed 13/02/2008)
2. www.bom.gov.au/products/IDN60700.shtml (13/02/2008)
3. www.bom.gov.au/marine/winds.shtml (13/02/2008)
4. www.bom.gov.au/difacs/IDX0894.gif (13/02/2008)
5. www.GRIB.us (03/02/2008)
6. manati.orbit.nesdis.noaa.gov/quikscat/ (13/02/2008)
7. www.ericssonw25.com (30/01/2008)
8. www.ussailing.org/appeals/supplements/05-08/Question_93.pdf (03/02/2008)
9. jury.americascup.com/multimedia/docs/2006/09/decision_[05.09.06].pdf (03/02/2008)
10. www.acdsee.com (03/02/2008)
11. www.iexpedition.org (13/02/2008)
12. www.maxsea.com/ (13/02/2008)
13. www.bandg.com/Products/Deckman-Tactical-Software/Deckman-v8/ (13/02/2008)
14. www.GRIB.us (03/02/2008)
16. Rolex Sydney Hobart Race web page example rolexsydneyhobart.com/standings_lite.asp
18. Formula from Expedition help file. TWS(@masthead) = TWS (@10m) x (h/10)a, where h is masthead instrument height in metres and a is a constant which is normally in the range 0.11 to 0.14. See www.iexpedition.org (13/02/2008)
19. manati.orbit.nesdis.noaa.gov/quikscat/ (13/02/2008)