Totally Amazing Weather Models

Click to enlarge

Here is something totally and utterly banoodley, gob smackingly, amazing.

The screen shot above, from my favourite GRIB viewer Luckgrib, shows a tropical nasty named Teddy hitting Nova Scotia on Wednesday 23rd September…but that’s not the amazing part, given that’s exactly what happened.

The part that just blows me away is that the GRIB comes from a GFS run on Tuesday 8th September…I’m pausing for a moment to let that sink in…that’s right, two weeks before Teddy hit us!

If that doesn’t blow your socks off, well, you’re just not paying attention.

But wait, it gets even more incredible.

This screen shot is from the same model at hour zero, over two weeks before the opening graphic, and it shows absolutely no sign of Teddy—the two lows to the west of Africa became other storms.

That’s right, somehow the meteorologists, mathematicians and programmers at NOAA have built a system that detected a camel fart in the African desert, or whatever it was that touched off the tropical wave that became Teddy, and accurately predicted that a hurricane would strike Nova Scotia two weeks later.

And not only that, the positions of the surrounding systems—high pressure over the Great Lakes and low over the Labrador Sea—are about right, too.

OK, they were a little off on the time and place Teddy came ashore. What am I saying? Anyone who bitches about that is…be kind, John…an ungrateful idiot.

I admit this example is exceptional, but I’m regularly seeing the GFS predict storms, both tropical and mid-latitude, 10 days out. And it’s rare not to get seven days warning.

But Only If We Use It

I  wrote a bunch three years ago, after getting back from Labrador, about how we pulled off our entire cruise without any weather surprises, using these tools and some others, and we had the same experience circumnavigating Newfoundland two years ago.

The big take away here is that all of us yachties need to be regularly looking two weeks out, particularly at this time of year, and increasing our understanding of how to use these tools strategically to plan our voyages, both ocean and coastal, rather than just thinking about the area immediately around us and only a few days out.

There is now little or no excuse for getting caught out far from good storm shelter or with the boat unprepared, at sea or in harbour.

Also, it might be worth giving the next NOAA meteorologist you see a great big hug…oh wait, there’s that COVID thing, maybe not.

And, finally, a big thanks to you Americans who pay for this through your taxes and then share it with the rest of the world for free.

Bad News and Good News

And here’s another thought. There is no question that climate change is causing more frequent and violent weather, and that this trend will only get worse. But at the same time these advances in weather modelling are giving us two to three times longer warning of impending bad stuff—I used to assume three days, I’m now seeing at least six and often nine—so that will go a long way to keep offshore voyaging viable in future years.

And Bad News

However, none of that means that we should go to sea, or even coastal cruise without preparing for the worst, particularly great anchoring and offshore storm survival systems. These models are truly incredible, but, in Phyllis’s words, it’s still

a forecast, not a prophecy.

Further Reading

Like what you just read? Get lots more:

Meet the Author

John Harries

John was born and brought up in Bermuda and started sailing as a child, racing locally and offshore before turning to cruising. He has sailed over 100,000 miles, most of it on his McCurdy & Rhodes 56, Morgan's Cloud, including eight ocean races to Bermuda, culminating in winning his class twice in the Newport Bermuda Race. He has skippered a series of voyages in the North Atlantic, the majority of which have been to the high latitudes. John has been helping others go voyaging by sharing his experience for twenty years, first in yachting magazines and, for the last 12 years, as co-editor/publisher of AAC.

35 comments… add one
  • Petter Mather Simonsen September 28, 2020, 8:43 am

    That was one raving review John, and it sparked my interest.

    I have checked out Luckgrib and it seems not only to be a grib file viewer, but also includes functionality to to download the actual of grib files.
    For the ones less familiar with this software, would you care to share the process and weather models and parameter settings you used in the above chart?
    Maybe the models also be used to detect when polar bears fart, not only camels, as I also live a a bit north of Africa.

  • Robert McDowell September 28, 2020, 10:10 am

    John, it gets even BETTER! I have been busy delivering boats around the Atlantic coast of Europe and the UK for the last 2 months and the model predictions 4-5 days out are getting the wind shifts and force correct to the hour! Last week my wife and I left Mylor (near Falmouth, UK) for a short hop around the Cardiff. It was a slow trip due to engine failure but since we were delivering a sailboat we went with the “backup”! Other than a 2 hour “retrograde maneuver” (ie sailing backwards on the tide) we did well. And we could keep sailing because I trusted the forecast. And to the hour we were able to tell the owner to meet us at the dock! He was astounded that #1 we would keep going without an engine and #2 we arrived at our designated time! I am loving the newest weather models. Oh, 12 hours after docking the back end of the low pressure came through and hammered the Bristol Channel with 35-40kts!

    • Dick Stevenson September 28, 2020, 10:01 pm

      Hi John,
      I agree with Phyllis: forecast not prophecy: but…
      It occurs to me that accurate longer forecasts could be a bit of a game changer in many ocean passages. With three days of confidence, a skipper was likely to have a period of time where any weather might occur on any ocean passage.
      With the longer and more accurate forecasts being discussed and reported, crossing the Atlantic with pit stops in Bermuda and the Azores (and on to Ireland or Portugal) might have only one leg (Bermuda to the Azores) where the boat was exposed. 40-foot boats like mine are likely to do the other legs within or close to the longer accurate windows and the longer boats like yours should easily be fetching landfall within the window.
      When I returned to North America via the “Viking Route” (Scotland, The Faroes, Iceland, Greenland and fetching up in Newfoundland), one of our reasons was the shorter legs. Our longest passage was 5 days and the departure forecasts we used were largely accurate for the full length of each jump.
      And thank you for acknowledging NOAA. I have recently finished a book where one whole chapter is dedicated to NOAA, their growth into the amazing institution it became and is, the decisions and impressive people that made it all possibly and the decision for it to be open source and available to all for free.
      My best, Dick Stevenson, s/v Alchemy

      • John Harries September 29, 2020, 8:06 am

        Hi Dick,

        Yes, that’s true, but what scares the hell out of me, and the reason that I ended the article as I did, is that those new to offshore voyaging may go to sea on poorly equipped and unseaworthy boats on the assumption that more accurate forecasting will keep them safe. That might work for a while, but sooner of later it will end badly, particularly on the Viking route. We have already seen the behaviour when a bunch of boats in the Salty Dog Rally had trouble in a fairly mild cold front:

        Bottom line, the models are getting better, but the weather is getting more violent, so it seamanlike boat preparation has never been more important.

        • Dick Stevenson September 29, 2020, 9:43 am

          Hi John,
          Yes, agree completely. Just like GPS was a game changer bringing certainty of l/l position and enticing many boats farther afield, sometimes much farther.
          It remains a topic for me (some might say a rant) of urging the sailing community to voluntarily “police” themselves before an outside agency enters and provides mandatory rules and regulation.
          Along those lines, I assume you are aware of the RCC Pilotage Foundation’s “Draft Polar Guide”. ( I believe the guidelines are an effort to give voluntary guidance for recreational boats going into polar waters in order to forestall regulatory agencies stepping in. Commercial vessels are already regulated by SOLAS and, so far, SOLAS has restricted the regs to commercial vessels. (I am speaking above my pay grade, but I think that is the gist.)
          I see this effort as wise on many levels, but it was stimulated, in part, by those vessels going into polar waters and needing assistance and who were subsequently seen to be inadequately prepared in either crew or boat or both. I believe the Canadian CG/SAR has born the brunt of this, but the proliferation of recreational vessels in polar waters, both N and S, is catching the attention of many, especially when things go pear shaped. The Polar Yacht Guide is designed to help skippers prepare their vessel and crew in an effort to have pleasure vessels discipline themselves to voluntarily prepare sufficiently.
          I also get the hell scared out of me by some boats/crew who go to sea unprepared and in un-seaworthy boats. Some skippers are just going to go anyway, but many skippers who mean well have been let down by the recreational boat community who continues to abdicate educating the boating community about best practices in crew prep or boat prep. The skippers are one thing, but I especially feel for the partners/crew of some of these boats who (many times blindly) depend on the skipper knowing what he/she is doing.
          My best, Dick Stevenson, s/v Alchemy
          PS, Not sure what happened with the Salty Dog Rally, but the rallys that promise hand-holding and pre-departure inspections and classes have long been another rant of mine.

    • John Harries September 29, 2020, 8:00 am

      Hi Robert,

      Yes, the resolution and close in accuracy has also got much better in recent years due to increased computer power and taking into account land contours:

      Love the “retrograde maneuver” terminology.

  • Lee Corwin September 29, 2020, 9:51 am

    I’m a long time member of SDR. In Chris Parker’s defense would say many of those involved boats had no experience in blue water sailing and were not adequately prepared in equipment nor skill set. We did the trip that year without mishap. The Caribbean 1500 had one bad year as well even with their more stringent requirements for entry. Phyllis’ statement about prophecy is totally spot on. Would note 500mb, gribs and modeling give you the big picture but sometimes not what’s going on in your immediate vicinity and that’s where you are.

    • John Harries September 29, 2020, 10:30 am

      Hi Lee,

      Your first sentence is exactly my point. My full thinking on this is here:

      • Dick Stevenson September 29, 2020, 11:03 am

        Hi John,
        I just re-read your referred-to article and highly suggest it to any boat doing their first offshore jump and especially to those boats heading S this fall: many good points.
        My best, Dick

        • Marc Dacey September 29, 2020, 9:28 pm

          That was an excellent comments section as well. Very emblematic of why I have kept subscribing.

    • Dick Stevenson September 29, 2020, 10:41 am

      Hi Lee,
      I am sorry to hear that you feel Chris Parker needs any defense and that you bring it up. I actually suspect he likely needs no defense at all and that going to his defense obscures the fact that, as you documented, many of the boats should not have been out there in the first place. I am also clear that, once a defense is proffered, a taint is often left behind and is hard to eradicate.
      I benefited greatly from knowing him and using his forecasts and routing suggestions from 2000-2006 from the US east coast down to Central America before I left his forecast area. I am greatly hoping that he is not taking the heat for skippers being upset when their poor preparation became undeniably apparent when the weather was not what they hoped and expected. This would mirror the abuse that Herb Hilgenberg, Southbound II, took as I remember it.
      Regardless of how one feels about outside forecasters and routers, I believe it to be totally the skipper’s responsibility for his/her vessel and crew: the buck should stop there.
      And I would wonder how much being “part of a rally” seductively seduces skippers (and their unsuspecting partners/crew) into blue water passages when they are un-prepared, boat and crew, to do so. I know some of the commercial rally’s sell themselves (in part) on this idea and I do not know whether the SDR charges money. It seems to me, in casual observation, that if a rally is proposed (commercial or otherwise) it must take on responsibility to ensure, to some degree, that participating vessels and crew meet some level of appropriate preparedness. Your statement that “many of those involved boats had no experience in blue water sailing and were not adequately prepared in equipment nor skill set” sounds like the SDR either dropped the ball (if the SDR had entrance criteria) or abdicated having and enforcing entrance in preparation, if there actually were entrance requirements.
      As said, I am responding in general only to the written posts and have no knowledge of the SDR problems.
      My best, Dick Stevenson, s/v Alchemy

  • Philippe Candelier September 29, 2020, 10:05 am

    Being able to review past GRIB file is a great feature. The application for that must include some kind of file/archive management feature.

    The app I mostly use for GRIB viewing is Weather4D. This app will on request download the latest GRIB from various models. But it has the drawback of erasing any older run it might have previously download.
    I have found this annoying, as there is no way to compare consecutive runs for convergence or divergence.

  • Maxime Ransan September 29, 2020, 7:03 pm

    While the example you mention is indeed impressive and most likely reflective of progress in weather modeling, one could have taken a one off example of a poor 3-day prediction and write quite a different article.
    It would be interesting to look at weather prediction data from a statiscal point of view and in particular apply statistic methodologies to offshore passages; this would most likely validate your final point about readiness and preparation for the worse. For instance even if accuracy of daily prediction is quite high, the probability of being right for a number of consecutive days reduces quite quickly and concequently the probability of a single wrong forecast during that time becomes higher.
    I really enjoyed your past article on rogue waves which had a somewhat similar approach and was very convincing.

  • Lee Corwin September 30, 2020, 8:14 am

    When I started ocean sailing insurance allowed short handed passages and did not vet who you took for crew. Now many require vetting and a dictated minimum number of crew. When I started forecasting was poor so you worked with observation of the sky and your recording barometer. You assumed you would see weather and prepped accordingly. You understood it was your responsibility and assumed it as an adult. You often took a newbie with you so another soul could learn the skills and join the fraternity.
    SDR is a non profit. In my view they properly put the onus on the captain where it should sit. They do what they can to have the tools available to the captain to make good decisions but it’s the captains responsibility to make them.
    ARC is for profit to my understanding. Ultimately they also place the responsibility on the captain but do screen to some extent. Both have had difficulties with weather causing serious trouble for vessels and crews.
    I continue to believe this is not the rallies fault. The responsibility rests solely with the captain. I believe that as fewer and fewer Americans voyage or sail and we are more protected by technology (such as you describe in this article) the prior sense that the captain and sailor needs the incremental experience and the skills it engenders has been lost.
    Blue water sailors are becoming dinosaurs. If you don’t want us to become extinct think it makes good sense to support rallies, take newbies as crew and do what you can to have new peoples join the fraternity.

    • Dick Stevenson September 30, 2020, 9:24 am

      Hi Lee,
      Lots of good points about responsibility and the state of affairs in the learning of off-shore passage making.
      Perhaps the SDR would be better off describing itself as “sailing-in-tandem” or a “recreational sailing group” or the like, rather than a rally. This reflects that the “rally” description, benign in and of itself, has gotten a bad name (in my estimation deservedly so).
      I would also suggest shifting the emphasis on some of your language (although I know what you mean): I do think the way these thoughts are expressed does make a difference.
      You say: “Both [rallies] have had difficulties with weather causing serious trouble for vessels and crews.” This puts the locus of the problem on the weather. The weather can’t cause problems for a boat: the boat’s inability to deal with the weather is the problem. That is where the locus of the problem exists. And that is where the remedy exists.
      It is far too inaccurate, far too easy and far too common to assert blame on the weather when things going pear shaped.
      Your feeling that Chris Parker needed defense in your first posting was in this vein: weather and/or poor prognostication was to blame or could be seen as blame worthy even though later in the post you acknowledge the poor boat prep of the troubled vessels.
      My best, Dick Stevenson, s/v Alchemy

  • Albert Stahl October 1, 2020, 12:27 pm

    Forecast and prophecy are synonymes ?

  • Chuck Batson October 1, 2020, 1:34 pm

    John, I recently asked a well-known offshore weather data provider whether they offered out-of-the-box access to 500 mb charts. The response (which I’ve paraphrased here) surprised me: “Surface forecasts have gotten so good we no longer see a need to look at 500 mb data.” Perhaps my surprise comes from first-level thinking? Maybe second-level thinking would lead me to a similar conclusion. What do you think?

  • Marc Dacey October 1, 2020, 7:23 pm

    Pretty sure that afternoon I spent with Lee Chesneau going over the joys of the 500 mb chart a couple of years ago weren’t a big waste…

  • Alastair Currie October 2, 2020, 3:17 pm

    I am only just starting down this road of Grib files and viewers. An observation, this year, my employer’s drill ship fleet, in the USA GOM, has been hit with a lot of Hurricanes or other high wind systems. That is not really the issue, the issue is the speed that they develop and then arrive at the rig site. There is normally sufficient notice to prepare the well and then disconnect 9000′ of marine riser from the well and let it dangle below the rig. The drill ships then weathervane into the wind and use their dynamic positioning (DP) thrusters to manage the motion. This year, they have been getting notice of the systems as usual but the time to high winds and seas is insufficient to unlatch i.e. from bearable weather to disconnect weather is now coming too fast. The end result is that the well is safe, but we cannot get the equipment into a position that allows the riser to be disconnected. The DP system ends up working flat out, high power thrusts and huge fuel consumption as the rig tries to stay within the watch circle based on riser stresses and depth of water. There have been a few excursions into the yellow zone as big gusts and wave combinations push the vessel faster faster than the desired thruster reaction. However, as the power ramps up and the thrusters orientate, the rig gets pushed back into the green circle. There are incredible back up systems should the rig excursion get to red zone and then exit that, so flow from well is very low risk.
    Your right about technology, prediction times and how our safety as sailors should benefit from that. I do wonder what climate change will mean in real terms, the unknowns, the magnitudes of speed, direction changes, force, pressure, surges, rainfall. It is worrisome for voyaging.

    • John Harries October 3, 2020, 8:21 am

      Hi Alastair,

      That’s really interesting, and more than a little disturbing in that it would seem to indicate that maybe that climate change is happening more quickly than our ability to keep the models updated to compensate, a possibility I had not really thought about.

      • Marc Dacey October 3, 2020, 9:26 am

        I have some Atlantic pilot books and Jimmy Cornell’s ocean route planning guides. Unseasonable weather events are making both less helpful in the short run, but it is hard to quantify by how much because one would need perhaps decades of weird weather to affect the assumptions gathered over 200 years of observations that comprise the pilots. We are left with the rather simple observation that higher heat makes for more energy potentially capable of driving stronger weather events in or out of season. Planetary disruptions aside, it makes adventure cruising…even more adventurous, at least potentially, because the models will have to keep pace with the chaos.

  • Robert McDowell October 8, 2020, 4:02 am
    • John Harries October 8, 2020, 7:29 am

      Hi Robert,

      It is indeed. I wonder if it will include GRIB data? My guess from the wording is that it will, but maybe not immediately.

  • Eric Klem October 20, 2020, 2:04 pm

    Hi John,

    Very interesting, that is quite incredible.  I don’t let myself look more than 7 days out and don’t put too much stock in stuff more than 5 days out.  I guess that I am going to have to start watching a wider window and learn to know what to do with the forecasts that are further out.  Thanks.

    My own observation is that the latest models handle land effects well in some circumstances and poorly in others but at least it seems to be fairly consistent.  For example, where our mooring is will fill in 1-2 hours later than predicted in a southerly sea breeze but also be 5-10 knots stronger than predicted and northwesterlies will be 5-15 knots stronger than predicted near shore and pretty close once you are 10 miles out.  The conditions that it tends to get right are stuff like easterlies.


  • P D Squire November 12, 2020, 7:54 pm

    The one properly troubling aspect of the GGR race is that, as you say; the sea is more rough more often today than it was in 1968. NOAA provides tools to offset this somewhat but they are denied to GGR entrants:
    “They will be navigating with sextant on paper charts, without electronic instruments or autopilots. They will hand–write their logs, cook with kerosene and determine the weather for themselves.”
    Seems to me that this is one area where “retro” could be compromised in this race.
    I suppose it could be argued that the permitted boats could generally be made stronger than Suhaili but I wouldn’t like to be facing these new angrier storms on one. Much rather avoid it.

  • Tim Newson, Practical Sea School November 21, 2020, 9:44 am

    This article made me chuckle when I read it.

    I was crossing Biscay this same week – in a little Sadler 29 foot bilge keeler – on a delivery trip with owner and their friend as crew. Leaving from Chichester on the 19th, we passed Ushant early on the 21st September where I could grab some mobile data for a weather update. Luckgrib showed a front coming in on Thursday afternoon, but focused in North of Biscay and leaving us with a manageable 20-25 knots in the South of Biscay coming from the NW, a nice broad reach. Although I’d figured we would probably be around Finisterre by then.

    Of course we weren’t. We had a problem with dirty fuel so didn’t make good miles at the start of the crossing when we were becalmed. Then on the Wednesday, inexplicably, the owner decided to sail SE while I was off watch for 5 hours. All this meant we ended up 60nm north of Finisterre on Thursday afternoon. It turned out that rather than being focused in the north the strongest winds were in the south of Biscay. That night we experienced 50 + knot winds and 7m seas. Under bare poles we were still overpowered so I tied all our warps together and trailed them in a big loop. This slowed us down to more manageable 4-5 knots. It also meant any tired helming decisions (in the dark with an inexperienced crew) were mitigated. To cut a long story short I decided to run downwind and sailed into Gijon, arriving on Friday afternoon. We avoided any big dramas or knockdowns.

    So yes – these new models can be amazing – they can also still be fallible and must never be relied upon. As you say in the article, you must always prepare for the worst. Thankfully we had.

    I couldn’t agree with you more about the GGR; the current rules are a travesty. Like the 2018 race, the 2022 edition seems headed for success in doing yet further damage to the public image of sailing. It certainly is failing at what it was originally was billed as – the inspiration for the next generation of sailors – it’s a great shame and an opportunity missed.

    • John Harries November 22, 2020, 10:34 am

      Hi Tim,

      Great comment with a lot of good lessons. Good on you for using what was at hand to slow the boat down and make things safe. And a good summary of the GGR and the opportunity missed to advance offshore sailing, rather than give it a bad reputation.

This is a guest