One arrow in the quiver

The term grib refers to gridded binary files, geek speak for a compressed data format favored by meteorologists as a means of digitally transmitting weather data. It has become both the favored buzzword of the weather worried and a topic that experts continue to debate — quite possibly leaving sailors with more questions than answers.

Lee Chesneau an ex-NOAA Ocean Prediction Center meteorologist, now an engaging weather seminar speaker, tells mariners to steer clear of the raw model data that grib files graphically provide and opt for the forecasts, surface analysis and 500-mb information already being broadcast by the U.S. Coast Guard and found free to download on the Web. Other weather-wise mariners find grib files to be a racer and cruiser’s best friend. A cottage industry has grown around repackaging weather data and selling to sailors hungry for a magic touchstone that will tell them what conditions lie ahead. The best way to understand what’s right for you is to take a look at both sides of this argument and get a feel as to where you fit. But before we dive head first into the brewing controversy, a little background information needs to be shared.

If there’s one area of seamanship where knowledge is power, it lies in the realm of weather forecasting and the development of a clear perspective of what tomorrow will bring. The more volatile the climate and higher the latitude or the closer you are to hurricane season, the more advantageous a heads-up weather-driven game plan becomes.

The good news for contemporary sailors is that never before has there been so much valuable data within easy reach of those poking along coastlines or sailing thousands of miles from home port. There’s only one catch: if you’re after more than a look at how the barometer is trending, where the clouds are coming from and which way the wind is blowing, you’re going to have to make some decisions about hardware and software, along with just how much of an investment weather wisdom is really worth.

There are three schools of thought when it comes to weather awareness and how to keep pace with a volatile atmosphere. The first is exemplified by a dwindling group of fatalists and stoics who prefer to reject communication technology and take what comes along. They embrace the idea of being prepared for weather contingencies and simply picking the right season for a passage — prepared to endure what comes along. Communication equipment takes a back seat to storm sails, vessel stability and structural integrity. Their gamble, like those who set off without a diesel auxiliary, is one of higher stakes and greater challenges in seamanship.

The second, and perhaps most sensible approach to weather awareness, is the do-it-yourself process that begins with the development of a weather-101 level of knowledge and an equal familiarity with radio signal propagation and what it takes to receive VHF, cellular, SSB and satellite voice and data broadcasts. This approach assumes that the onboard decision maker can read a surface chart and make sense of the contours that make up a 500-mb chart. The premise is that an informed decision maker is actually more reliable than a radio link with a shoreside weather consultant.

The third option is to relegate some, or all, of the weather-info gathering to an expert not on board. With this abdication of weather awareness, comes complete reliance upon an external expert and a communication link that carries no guarantee. Of course, most weather-savvy passagemakers see the ideal approach as an amalgam of all three philosophies.

Buying what you’ve already paid for
Do-it-yourself oriented weather-aware sailors believe that NOAA has done a pretty fine job of sampling and evaluating the dynamics of the atmosphere and put that information at the fingertips of the boating community. So, they ask themselves, why pay others to use the same model data and charts to prompt your decision making? But this argument only holds water for those who have done their homework and can handle tasks like reading information on weatherfax 24-, 48- and 96-hour charts and understanding why a teardrop-shaped isoheight line on a 500-mb chart is an alarm bell indicative of the formation of a cut-off low. And if you want grib files, satellite radar analysis of the sea surface and its associative relationship to wind speed for pre-passage planning, it’s all on the Web. Ben Franklin’s wisdom about the “weather-wise and the other-wise” seems to sum it all up. If you are informed and have developed an onboard decision-making process, you are better off regardless of whether or not you also plan to sign up for a grib-file service, lash up a Sirius satellite weather radio receiver to your Raymarine, Furuno or other networked digital plotter or intend to pay for a concierge custom-forecast service.

Most for-a-fee weather gurus willingly admit that they base their decision making on the same model information and forecast data that NOAA provides for free in voice and digital format, promulgated via the Internet. Their value-added can be two fold: These services specifically focus on conditions in their client’s vicinity and the experts themselves are often experienced, small-craft-knowledgeable mariners. This localized awareness and their understanding of specific model shortfalls, associated with the unique conditions of a specific water body, can be of considerable value. They rightly refer to such expertise as value-added in developing their custom forecasts. And it’s exactly what’s missing in every grib file.

Grib data is based upon purely computer model-driven calculations using complex algorithms and there’s absolutely no human intervention, not even a look out the window. What’s amazing is how accurate these computer-crunched projections can be. In fact, those voyaging in remote Great Southern Ocean regions have been shocked by how often the grib-file forecast matches local conditions even though weather stations and sampling are scarce. This fact alone is a great testimony to how good models are getting, but despite this glowing up side, there are times when the models are 180 degrees out of sync with each other, one calling for calm seas and sunshine and another placing a rousing gale in your path. The grib charts generated will be equally contradictory and there’s no human element ready to add an expert’s opinion.

Developing grib data
NOAA sifts data from the atmosphere via weather stations, weather balloons, ship reports, satellites and ground radar. Their models GFS, AVN and Wavewatch III give a shorthand digital summary of the information tweaked by weighted mathematical formulas. For example, the GFS model covers land and sea areas and generates wind vectors at a 10-meter height over a grid that is either 0.5 x 0.5, 1 x 1 or 2.5 x 2.5 degrees in specificity. The data includes wind speed, direction, temperature, pressure, humidity and much more.

Wavewatch III is another model that streamlines information and offers only high seas data limited to oceanic wind, wave and current that is available in a much smaller file. It’s updated in three-hour intervals and the model is run four times a day. Its grid size is 1 x 1.25 degrees, meaning that each data point is for a rectangular area of sea surface, roughly 60-miles-by-45-to-75-miles (depending upon the latitude of the cell). The Navy has put models like NOGAPS and COAMPS in the public domain and more and more data is being shared worldwide.

Acquiring grib files
Ashore, the public can use high-speed Internet capability to snappily download grib files, but at sea it’s a Slowsky world out there with baud rates that make a 90s shoreside-dialup connection seem impressive. This means that either you’re going to spend huge amounts on satellite phone, Inmarsat or other provider air time or you will need to purchase a plan with a provider such as WeatherNet or Saildocs that compresses grib files and provides them via SSB or satellite service. Feedback from competitors in a recent California-to-Hawaii race had a few owners spending $2,000 on airtime for weather information downloads!

The next investment is in a grib-file reader, a software program that turns a binary string into a color chart depicting wind velocity and sea state as well as other vital meteorological data. Grib Explorer or another Windows-compatible program can be nested on your PC and linked to your communication gear of choice. Inshore, this might be a dockside landline, cellular provider or Wi-Fi link. At sea, an SSB with a digital Pactor-III node can do the job, as can SailMail, Ocens Mail and other email/SSB links. Inmarsat and Iridium and other satellite systems can also tie the grib-file reader to a service provider. Ham operators use products like Ugrib and their Pactor-equipped transceivers to circumvent the costs of a commercial service, but the process is less user friendly and the products are not quite as easy to handle.

Ideally, grib files should be used as an input factor in forecast development, just as the trend of the barometric readings or a shift in wind direction is factored into a forecast. Basing the whole forecast on a grib file run is too monolithic. For example, models that are spot-on with wind-velocity predictions with one type of weather system often miss the wind-velocity implications associated with another. A case in point is the prefrontal trough squalls that arrive well ahead of a cold front, which often pack more of a wallop than the cold front itself. This phenomenon is most noticeable on the north wall of the Gulf Stream and information garnered from NOAA’s GFS model runs are regularly tweaked by meteorologists developing a forecast. They look at satellite photos of the stream’s constantly changing north wall conditions, and check ship reports as well as note how different models read a similar situation.

Leaving a forecaster entirely out of the loop is a gamble with a serious downside, and it’s why skilled navigators use grib files as another tool in their box — not the only tool available.   

Ralph Naranjo is a freelance writer and photographer living in Annapolis, Md.

By Ocean Navigator