Faith in Weather Forecast?

[Kenomac]

How many of you actually put much faith in weather forecasting? I've read many posts over the past years on this forum stating themes such as "I never go out if the forecast is for winds above 15 knots," "only fools get caught out in bad weather with forecasting as good as today's," etc. etc.

We use three different marine weather apps which generally, are very good. Pocket gribs, Weather4D Pro, and Buoyweather (we subscribe). But as you can see from the pocket grib photo and from the video, you don't always get what you think. Here we are in what was forecast to be calm seas less than 1 meter and winds of 2.2 knots sailing around the southern foot of Italy. What we had for this six hour period was a consistent 26-36 knots of actual wind, 20-30 knots apparent wind, and 2-3.5 meter swells which we were able to sail comfortably using only our staysail to maintain 6.5-8 knots of boat speed.

It was a 50 hour passage with various combinations of sail and wind speed, even including a round up under full sails just south of Messina, when the wind speed increased from 20 knots directly behind us up to 55 knots in just a few moments.

All in all, it was an average passage which didn't really include anything unexpected, since we now allow for weather variations. Mrs. Mac looks a little tense in the video, she really isn't, it's just the way she's sitting. She's actually quite used to sailing in these conditions. I welcome your thoughts on the matter. Do any of you actually count on and put any faith in the forecast? I've found it's always better to be prepared for the worst even when 2.2 knots is forecast. ;-)

Ken

https://m.youtube.com/watch?v=s0HgonCDFxI

[Dockhead]

Ken, in smaller bodies of water with a lot of land influence -- there is a lot of chaos in the system, and a lot of chaotic events which can't be forecast, especially in summer. Land creates convection, especially in summer, creates sea breeze and land breeze, creates summer thunderstorms, and creates katabatic winds, all of which is different, and far less predictable, than gradient wind which is create by larger systems and basic pressure differences. The weather models work amazingly well with gradient wind, and so further from land and in larger bodies of water, the forecasts are much more reliable.


But you know the old sailor's weather adage -- "Long foretold; long last. Short notice; soon past." This is of crucial importance to us -- a sudden blow caused by convection or other land effects cannot create a dangerous sea state offshore (what you might get in harbor entrance, over bars, tidal races, etc., may be a different matter). No amount of wind, blowing for less than a few hours, can be really dangerous for a strong cruising boat like yours. Just heave to or run off and let it pass. To get a really dangerous sea state, a lot of energy has to get transferred into the water, over a large geographic area and over a longer period of time.

So what that means is that we don't actually have to worry all that much about unforecast, land-influnced weather events -- dangerous weather events are those that create dangerous sea states, and you can always see these coming pretty far in advance. These events are created by large weather systems and big pressure differences, and you always see that kind of thing at least a couple of days ahead.


Here is what to be afraid of:




It's hard to photograph sea state, but this is higher than our first spreader (which is 10 meters above the deck), blotting out the sun and blanketing the sails in the trough. This occurred last year in the North Sea, building up slowly after 24 hours of a F9. Even these are not actually dangerous, except that they are already big enough for the tops of them to start to fall off:




So pretty soon you are steering around the breaking ones, praying, and steering cross-ways down the wave faces trying not to speed out of control.

Unfortunately we got hit by one of these, and it was like having a building collapse on you. Knocked us down. An experience I don't want to repeat.


But this was not unforecast. This was a big weather system which was foreseen even a week ahead, and which lasted for a number of days. On the day we sailed, it was supposed to be F8 for three days, which would have been ok as we were sailing downwind, but after 24 hours of it and then building to F9, that was already too much, in the Atlantic Ocean with 3000 miles of fetch.

[TurninTurtle]

Depends on the forecast...and where you are.

They are getting pretty good about a lot of stuff for up to about a week in advance.
Some things they can't predict 12 hr ahead.

Better/more data over land, notably continental US, with all the little weather instrumentation at airport and TV/radio stations etc. means more accurate predictions.
Out over open water they get some data from ships... sometimes. So its not quite as accurate.

They have pretty good satellite pictures and can detect a lot, but actual wind speed and temperature data is important.

[Kenomac]

Dockhead,

Here's a good example showing what you wrote about. Look at the sea state in this video, maybe a 1 meter swell if even. We're sailing under full Yankee and full mainsail as I'm noticing some windsurfers off our Port side. I'm recording the wind gusting from 9 knots up to 38 knots and how the boat is handling the gusts, not a big deal at all. Who would guess that only moments after I pressed the stop button on the recording, that we were hit with a continuous 55 knots and completely rounded up. I called Pam and Charlie up from down below to help out, and it only took maybe two minutes to furl in the main and Yankee. Sailed under bare poles and a little motoring for the next hour, then back to staysail and Yankee for the rest of the day.

Sorry, but the reflection of the American flag is obstructing the wind speed indicator, but if you watch the flag, you'll get the idea.

https://m.youtube.com/watch?v=Id16ggHPZUk

[Dockhead]

Originally Posted by Kenomac

Dockhead,

Here's a good example showing what you wrote about. Look at the sea state in this video, maybe a 1 meter swell if even. We're sailing under full Yankee and full mainsail as I'm noticing some windsurfers off our Port side. I'm recording the wind gusting from 9 knots up to 38 knots and how the boat is handling the gusts, not a big deal at all. Who would guess that only moments after I pressed the stop button on the recording, that we were hit with a continuous 55 knots and completely rounded up. I called Pam and Charlie up from down below to help out, and it only took maybe two minutes to furl in the main and Yankee. Sailed under bare poles and a little motoring for the next hour, then back to staysail and Yankee for the rest of the day.

Sorry, but the reflection of the American flag is obstructing the wind speed indicator, but if you watch the flag, you'll get the idea.

Yep, typical Med weather. No harm done on a really strong boat like yours.

One thing you might try in such a situation is to quickly heave-to -- much faster than furling sails. Then if it's comfortable, just hang out that way until it blows over, or you can furl from that position. All this assumes you have searoom, of course.

The katabatic storms in the Adriatic are incredible -- unforecast and incredibly violent. But they don't last more than a few hours, so you just gut it out.

[Kenomac]

Just to be clear, we don't view the weather shown in my post #1 video as being anything out of the ordinary, we've come to expect it and view it as a good sailing day. A day when we can easily average 7.5-8 knots without stressing about anything. But there are folks on this forum who would be quite frightened by experiencing the same conditions when they only expect a sunny 2.2 knots and flat seas.

I know one time, my daughter and her friend were onboard when we got hit by a thunderstorm off Mallorca, they were very scared, so we ended up detouring back to Mallorca for a few hours, then starting off again.

There seems to be many people who place way too much faith in the weather forecast and go out only on days forecast to be mild. I think it might be better to head out sometimes on days when it's blowing harder to get used to 20-35 knot conditions. Then when you get hit by 50+ knots, it isn't such a BIG DEAL. You'll know what to do.

[oldsalt_1942]

My dad used to say, "If you believe a weatherman you'll believe what a politician tells you, too."

[belizesailor]

Originally Posted by Dockhead

Ken, in smaller bodies of water with a lot of land influence -- there is a lot of chaos in the system, and a lot of chaotic events which can't be forecast, especially in summer. Land creates convection, especially in summer, creates sea breeze and land breeze, creates summer thunderstorms, and creates katabatic winds, all of which is different, and far less predictable, than gradient wind which is create by larger systems and basic pressure differences. The weather models work amazingly well with gradient wind, and so further from land and in larger bodies of water, the forecasts are much more reliable.


....

Modern forecasting is amazing, but you have to understand its limitations to use it properly. Most models, like GFS, are not very good near shore, but quite good in open ocean. "Meso scale" models, like COAMPS, are being developed to improve upon this, but only currently focus on limited areas.

If you sail in a given area long enough and learn the localized effects there then you can apply your own local corrections. For example, near coastal Belize, I know the near coastal winds will be backed, the closer to shore the more they are backed, versus the GFS forecast. This land effect decreases as you move away from the coast, until about 20 miles off the coast...no more land effect. Land effect off Honduras is even more dramatic, significantly changing both speed and direction.

Many take GRIB files to be gospel, but they are just raw uncorrected model output. Also, they dont show frontal boundaries.

Various convection metrics, like CAPE, can be useful for evaluating the likelyhood/potetial strength of localized squalls. Some GRIB viewers will overlay these metrics. Here in the squally W Carib, I use this a lot.


I dont know the Med, but expect that with all the surrounding land mass there is probably a lot of localized land effect.

Ken, how far were you from land during the higher winds?

[zeehag]

as i do my own, i have great failth in me. i donot have any faith at all in the words of others.
been reading weather charts since age 8, 1956. why should i change now?
i use many sources to read what is coming-- and no words unless by storm 2k contributors of merit, which are backed by sat fotos from many different sources.

[YPSILANTI]

Do you know what little boys that tell lies become when grown up?



Yes!! weather forecasters.

[Kenomac]

Originally Posted by belizesailor

Ken, how far were you from land during the higher winds?

When we got hit suddenly by 55 knot winds we were only less than one mile off shore. During the video where it was predicted to be 2.2 knots and we had 20-38 knots we were 40 miles from land in all directions.

[zeehag]

Originally Posted by YPSILANTI

Do you know what little boys that tell lies become when grown up?



Yes!! weather forecasters.

or presidunce

[barnakiel]

Originally Posted by Kenomac

How many of you actually put much faith in weather forecasting?

(...)

I do. Then again, I must, being trained in meteo and working as a weather router.

I am not going to bite the feeding hand. ;-)

More seriously though, 'faith' is probably the worst word to use as it immediately introduces a vague scent of mysticism and palm reading. In fact, it is all about plain probability and proper understanding of stochastic analysis. Science, but not high science.

To put it in everyday language: 100% correct is what you are observing, and the further forward you forecast, the lower the probability of the forecast being spot on, caeteris paribus.

Because there are many variables, you can only discuss them one at a time (two, if you are versed in multivariate analysis), but you can manipulate any number of them at a time. The point is our brains are wired to use 3 variables at a time only (two variables in data plus time shift).

For the above reasons, I generally discourage people from studying the weather forecasting 'a little'. Having 'a little' knowledge may be (surprise !?) in this case worse than having none. Similar to finance, contraception and flying skills.

So, to sum it up. I have 'zero faith' in forecasts but I can tell the probabilities and I know what is meant and implied when forecasters say their things. Do not try this at home though: find a wx webservice that seems to best match your area (yr.no, noaa, metvuw, AUS BOM, etc.) and allow for the unexpected. The further the forecast, the more unexpected is expected.

Cheers,
b.

[Kenomac]

Originally Posted by barnakiel

I do. Then again, I must, being trained in meteo and working as a weather router.

I am not going to bite the feeding hand. ;-)

More seriously though, 'faith' is probably the worst word to use as it immediately introduces a vague scent of mysticism and palm reading. In fact, it is all about plain probability and proper understanding of stochastic analysis. Science, but not high science.

To put it in everyday language: 100% correct is what you are observing, and the further forward you forecast, the lower the probability of the forecast being spot on, caeteris paribus.

Because there are many variables, you can only discuss them one at a time (two, if you are versed in multivariate analysis), but you can manipulate any number of them at a time. The point is our brains are wired to use 3 variables at a time only (two variables in data plus time shift).

For the above reasons, I generally discourage people from studying the weather forecasting 'a little'. Having 'a little' knowledge may be (surprise !?) in this case worse than having none. Similar to finance, contraception and flying skills.

So, to sum it up. I have 'zero faith' in forecasts but I can tell the probabilities and I know what is meant and implied when forecasters say their things. Do not try this at home though: find a wx webservice that seems to best match your area (yr.no, noaa, metvuw, AUS BOM, etc.) and allow for the unexpected. The further the forecast, the more unexpected is expected.

Cheers,
b.

It sounds like we already do what you recommend. The forecasts I used as examples in the initial posts were less than six hours old, I updated in Messina, and anytime we were within a few miles of the coast. I'd also been following a forming Mistral off France for three previous days and had chosen our anchorages and timetable to enable us to follow a relatively calm period between two periods of increased wind. The timing worked out perfectly.

I really just want to stir up a little debate for those people who put an unusually high level of faith in today's weather forecasting science and encourage those same folks to get out and practice reefing and heavy weather tactics for when the forecast isn't so accurate.

Thanks for your contribution.

Ken

[meirriba]

Originally Posted by Kenomac

When we got hit suddenly by 55 knot winds we were only less than one mile off shore. During the video where it was predicted to be 2.2 knots and we had 20-38 knots we were 40 miles from land in all directions.

The gribs around Messina show ~20kn. What area you have referred to with ~2.2kn?

As to reliability you must take everyhing as probabilities. I keep on file a forecast received mid-Atlantic during ARC2003. It warned of a 'major tropical event, gusts over 80kn'.
The actual wind was 15-20kn.