Now and future weather tools

The term “heavy weather” makes perfect sense to any sailor who’s ever been smacked around by tons of agitated ocean. Such a sailor is usually willing to devote time and money to better understanding the dynamics of marine weather and collecting data and predictions during a voyage. That’s good, because weather is a very complex subject, and – despite the fabulous evolution of weather resources ashore – getting the information we want at sea is still not easy.

As soon as the telegraph was developed in the mid-19th century, observers started sending measurements to central offices where meteorologists could create synoptic (large area) weather charts, monitor systematic trends and produce regional forecasts. Today, human observers are joined by satellites, balloons, buoys and other automated stations sending in massive quantities of data collected by some of the world’s most powerful computers. Last year, National Weather Service director, John W. Kelly Jr., flipped the switch on a new 786-processor IBM computer that can make 2.5 trillion calculations per second (TeraFLOPS). “This new supercomputer puts us closer to reaching our goal of becoming America’s no surprise weather service,” he said.

Truly, there are fewer and fewer weather surprises – at least ashore or close to it. We can tune in to our local TV weather and watch a 3-D animation of tomorrow’s forecast, the result of those TeraFLOPS organizing and modeling the data with a little help from some experienced people. We can use the Web to get less snazzy but quite localized predictions or even go directly to nearly real time sources – looking at precipitation echoes on a 100-mile Doppler radar screen, a table of wind and swell conditions radioed in from an offshore buoy or an image from a satellite 500 miles overhead. Surely we can avoid heavy weather when we stay close to the coast, but everything changes when we head offshore. It’s almost a cruel irony, but the farther offshore we go, the more exposed we are to serious weather – the leaner the available resources become and the harder they are to receive and assimilate.Voice, text and fax

To be sure, dedicated organizations like NOAA’s Marine Prediction Center (MPC) in Camp Springs, Md., provide voice and text weather via SSB, NAVTEX, satellite and Internet – but they tend to be a heck of a lot drier and more granular than what’s available ashore. For instance, if you were bound south to Bermuda on July 4, this is the most relevant part of your high seas text/voice synopsis: WARM FRONT EXTENDS FROM 41 N 92 W TO 37 N 55 W. AREAS OF DENSE FOG WITH VISIBILITY BELOW 1 NM EXTENDS WITHIN 300 NM N OF FRONT. There are also the unofficial but very useful weather nets like Herb Hilgenberg’s South Bound II and David Jones’ CaribWX on which SSB/Ham-equipped voyagers share weather data with a central prognosticator. However they get the information, most navigators will draw it out on an ocean outline chart, since offshore forecasts without physical landmarks are hard for most people to grasp.

Sailors want weather graphics, particularly to understand the trends that can affect a little ship possibly dodging systems capable of significantly greater speeds. Enter the basic black and white weather fax chart, provided by agencies like MPC worldwide and probably still the most popular weather tool offshore. Traditionally printed out by dedicated SSB receivers (always slowly and sometimes with noise-induced anomalies), faxes are now available several other ways. Companies like OCENS in Seattle and Xaxero (pronounced za-za-ro) in Auckland, New Zealand, market software/hardware systems that demodulate the SSB signals and download them to a PC and capture the faxes as clean raster images. Some satellite services broadcast faxes or offer them online to download. Globalstar, for instance, recently added 9,600 bps data transmission to their sat phones, and Weather Predicting Simplified author Michael Carr reports that he was able to download Web-based faxes with a Globalstar phone throughout an Atlantic crossing. He added that he would never mess with SSB faxes again, if he can help it.

Faxes are a mature tool, with twice daily (usually) individual sheets with analysis, 24-, 48- and 96-hour forecasts, sea state forecasts and more. Particularly interesting are the 500-millibar (mb) charts that Steve and Linda Dashew call “the most valuable forecasting tool at our disposal” in their excellent Mariner’s Weather Handbook. These charts define the pressure map and wind currents about halfway up in our atmosphere, a view that is quite different than, but closely related to, the activities below. The 500-mb charts provide hints to the potential paths surface systems will take and also suggest how weather systems might become amplified. Making sense of them, however, requires study. You need to compare different charts, each in a time series, to really get the picture. Used well, faxes are probably the best tool available, but they’re certainly difficult and bland compared to shoreside technologies. There should be a better way, and it looks likes it’s coming in the form of vector weather data – also known as gridded binary (GRIB) files.GRIB files

The acronym GRIB seems to have dual meanings. First of all it’s a very specific World Meteorological Organization computer file type for passing around all sorts of geopositioned weather data or model predictions, usually in 1, 2.5 or 5 lat/long degree data points. The term is also used to describe various proprietary file types to pack lots of weather information – much of it true GRIB data – into an efficient file that can be transmitted easily and then turned into imagery by dedicated software. In many ways, GRIB is to weather fax imagery as vector charts are to raster charts. The first companies to exploit it are the developers of PC charting programs who are already adept at mixing and displaying vector data like routes and charts. Current versions of MaxSea, RayTech and Nobeltec’s VNS can all display GRIB data, and, though the technology is very young, its potential is terrific. Data is displayed in layers over a zoomable map, crisp and in color. You can choose display types and mix layers as you wish. For instance, if traditional barbed wind roses don’t read easily for you, you can change them to scaled arrows. You can compare surface pressure to wind to sea state to whatever data with simple clicks, making their relationship comprehensible. A click on the chart will also give you point forecasts for that spot, and, because weather has a time dimension, it can be animated across the screen, as seen in both the RayTech and MaxSea versions.

Currently, the various software companies offer their users free downloads of basic short-term data or subscriptions for more extensive data by email. Even the full data sets don’t have some of the more arcane – but useful – fax information like 500-mb conditions, and even the compact format makes for 20- to 50-kb email attachments which can be slow and costly via present offshore methods. The main gripe with GRIB, is that most of it is what’s called model output.

Despite the introduction of TeraFLOP machines, weather modeling is still an emerging technology, drawing raised eyebrows from many a professional forecaster. Lee Chesneau, an MPC meteorologist, says that model data is great in the sense that any related information is useful, but he cautions that it needs evaluation. Working shifts at the MPC office, pros like Lee look at output of several models along with current conditions and even recent history of the various models’ accuracy. They synthesize everything, sometimes even tweaking the model parameters. Then they cut and paste the best material from various model layers to their own forecast layer, draw in front and system analyses/forecasts and output the layer for eventual distribution as faxes or as the basis for narrative text.

Most official meteorological offices could be outputting pure vector data weather charts, but they aren’t. In the last few years the MPC has added color to some of their Internet-distributed charts (though still slight compared to the U.S. Navy’s snappy FNMOC products), but it may be some time before they can add vector output to their burden, especially in this era of shrinking federal budgets. The bottom line: human evaluated vector weather data is not free. That’s why Nobeltec is the only vector data provider, which it gets from a private company called Applied Weather Technology, and it’s fairly pricey at $10 for a single five-day forecast or $500 for a year. The displays, however, with their sharply drawn fronts and system motion predictions, are handsome and useful.

Another important feature of vector weather data is its intelligence, that is to say it can be used in equations, a fact both RayTech and MaxSea take advantage of in their optional routing modules. Give the program a destination and a polar diagram of your vessel’s abilities on various points of sail at different wind speeds, and it can develop isochron lines representing a set of points your boat can reach in a fixed time. Then it calculates the best set of points in the given “smart” wind conditions, and you have an optimal route to sail.

While none of the GRIB weather displays I saw fully realized the medium’s potential – and all had a bug or two – the good news is that these companies are squashing bugs and adding features continually. Nobeltec president, Michael Neal, says they are listening to their users carefully and expect to add more data types, more frequency (VNS now displays only one data set per day) and possibly even a way to query a live meterologist. Raymarine’s new RayTech product manager Lewis Chemi has a lot of experience with weather data and big ambitions for RayTech.

The future of GRIB is bright indeed. Besides meteorological offices eventually offering evaluated vector data, the models themselves will no doubt get better. Their output is already largely available to folks with the will and skill to process it. To see what a clever – perhaps obsessed – programmer can do with this stuff, you just have to visit, which uses NOAA’s Wavewatch III model to generate very specific wind and wave predictions. Dan Martin, a former futures trading software programmer and windsurfer based on the north shore of Maui, Hawaii, says he’s been on this project since 1994 “with intense focus” and claims to have written the first Java program that reads GRIB data directly and dynamically. He also says that years of comparing model output to actual buoy conditions have convinced him of this model’s accuracy. His code will reference instantly a 30-mb GRIB file downloaded to his server twice daily and put up prediction graphs for any spot of your choosing or one of his preselected virtual buoys. He is planning to introduce a subscription service that will deliver similar results in text form as email or even to a WAP-enabled cell phone – for a projected 10 cents per message. No doubt we’ll see all sorts of interesting GRIB-style products in years to come.Direct from the source

A totally different strategy for bringing weather imagery to an offshore PC is to grab the raw stuff directly from the devices that produce it. One of those sources is the constellation of satellites that continually photograph the earth and radio the imagery down to the central meteorological offices and anyone else with the means to receive them. OCENS and Xaxero both market the compact helical antennas, receivers and complex software needed to track the satellites; obtain the desired visible and infrared images; and apply the coast outlines, color palettes and vessel routing information all useful to interpreting them. It’s like having a crow’s nest 500 miles up. You might catch a hurricane forming up at the same time the meteorologists ashore do. Colorizing the infrared temperature images can reveal thunderstorms embedded in cold fronts or eddies in the Gulf Stream. Satellite direct data can be stunning and it’s free. However, the systems are fairly expensive, and the user must master both the software and the image interpretation. OCENS even offers two-day seminars for using their products.

Fast evolving technology suggests that several other similar schemes are in our future. For instance, Furuno’s new NavNet vessel networking system talks in Internet style TCP/IP protocol. It doesn’t mean much now, but Furuno is thinking ahead. When broadband wireless reaches offshore vessels, NavNet could facilitate the feed of, say, a 150-mile range Doppler radar weather image directly to a boat’s monitor in real time. Such radars already blanket the U.S. mainland and are starting to measure wind speeds as well as precipitation density. It seems safe to assume that their range and abilities will grow; not to mention the satellite-based radar systems in development that can measure wave heights, wind speeds and cloud shapes. Another data source from which sailors might appreciate a direct feed are the hundreds of coastal and offshore weather bouys. Since they already send their readings ashore by radio wave, perhaps one day we’ll be getting their output directly to our vessel.

Finally, there are improvements afoot in a boat’s ability to monitor weather conditions directly, though first we should mention traditional tools like thermometer, barometer, anometer and eyeballs. The Dashews have spent a lot of time in remote oceans aboard their series of Deerfoot designs, and they note in their weather book that perhaps the primary reason to learn marine meteorology – and to carefully monitor changing conditions – is so that you can recognize quickly when an offshore forecast is wrong and be ready to guesstimate what is really going on. Agencies like the NWS may have a goal of no surprises, but the truth is that there are vast stretches of ocean with fairly light data collection and fairly general forecasting.

The Dashews are quite keen on using radar for early warning of squalls. In fact, they are claiming that as the technology’s most important function. It seems likely that ship board radars will continue to improve, and perhaps acquire some of the capabilities of the land and satellite units discussed above. Other “over the horizon” sensors are coming. The Boltek company makes a product called Stormtracker which can detect and map lightning strikes 200 miles away. It consists of a tiny direction-finding antenna, a PC board or PCMIA card and software that interprets the data. At present, their market is concerned with enterprises like chemical and fireworks factories and the legions of dedicated crazies known as storm chasers (as seen in the movie Twister), but – at about $500 for a complete mobile system – some of us dedicated storm flee-ers might want to add this to our onboard weather kits.

Presume for a moment that numerous voyagers, commercial and pleasure, expand their ability to collect weather data and put it in computer-friendly, GRIB-style formats, and fast, worldwide wireless does eventually become easy and inexpensive. Won’t we all share our data with professionals ashore and each other afloat, much as commercial ships report into NOAA or yachts compare notes on SSB weather nets? Perhaps one day we’ll get very user friendly, long-term and accurate forecasts of whatever section of ocean we are sailing across. The human element

In the meantime, the situation with offshore weather requires the attention of a knowledgeable human on board or in direct touch. Plainly spoken, understanding and making best use of what’s available out there is not trivial. If TeraFLOP computers and trained professionals screw up predictions sometimes, how’s a sailor trying to manage a boat at sea going to do? Certainly, it makes sense to start studying the weather well in advance of a given voyage. You might use the splendiferous resources of the Web to develop a few sources that really work for you and follow them carefully for a week or so before a trip, perhaps printing out the latest charts and forecasts before you depart. Hopefully you’ll start your trip with a thorough idea of what’s supposed to happen and then modify it with whatever forecasts you can receive offshore as well as your own observations. You may also want to read the books mentioned in this article or take one of the weather seminars offered by several organizations like our own Ocean Navigator School of Seamanship (ONSOS).

You might also consider a different route, so to speak – hiring a personal forecaster, or router. This is a professional – usually a trained meteorologist – who spends his days hooked into the weather data monster and can advise you with very specific and efficient email or marine telephone messages. Ken McKinley of Locus Weather – who also teaches for ONSOS – says he usually begins a routing assignment with an interview. He’ll find out what a boat is capable of and what the crew is willing to take, in effect how many tons of heavy weather water are tolerable. A boat on delivery, for instance, may be more willing to trade comfort for speed.

McKinley gets the data he likes to use on the Web, primarily satellite imagery and fax style charts, and he uses it to prepare preliminary and departure forecasts, suggest routes and consult during the passage. This personal service, available from a growing number of ocean routers, is understandably not cheap. It may also run against the grain of some yachtsmen who are out there in part to be self-sufficient, but it must be comforting to have a wired-in expert like McKinley keeping a weather eye out for you.

If all I’ve written here seems a somewhat paranoid exhortation to spend a lot of time and money before ever heading out to sea, I’d like to end with two notes. The first is an acknowledgement that I’ve suffered the abuse of 30-foot seas in a small vessel with almost no weather equipment or knowledge, and – like many before me – I survived. The second is a remembrance of beam reaching over the huge but smooth swells that concluded that whole gale – a fantastic sensation I’ll never forget.

Contributing editor Ben Ellison is a freelance writer and Ocean Navigator seminar instructor.

By Ocean Navigator