Understanding Weather Forecasts?

[Kings Gambit]

It's not uncommon to hear from sailors about how NOAA's weather forecasts are often wrong and cannot be trusted. I've heard it more than once just this week, including here on this forum. In contrast, I've found the forecasts to be very reliable and quite useful in planning for weekend or longer excursions and especially for executing them. I think it might come down to the user's understanding of a forecast. First, NOAA forecasts are always provisional pending the next update, the provision being, I suppose, that nothing important has changed since the forecast was issued. Updates are every 6 or so hours. So, if you pick up a forecast near the time of the next update and then don't bother to get that update 30 min later, then it might appear to you that the forecast you got was "wrong" when in fact it had merely expired while you were out. There's also the zonal nature of the forecasts. For example, "the coastal waters from Pt. Mugu to San Mateo Pt., California, including Santa Catalina and Anacapa Islands...." This zone stretches about 95 n.mi along the coast out to about 40 n.mi. If the forecast is 10 to 15 kt, then somewhere in the zone it blows 10 kt whereas elsewhere it might be blowing 15 at the same point in time. But if in your location the wind happens to drop to 5 kt for an hour or two does that make the 6-hr zonal forecast wrong or unreliable? NOAA voice (VHF) and text (web) forecasts are not "real-time" b/c the media do not allow it. Thus, it's more informative to see an animation of the progress of a forecast over the whole zone(s) using, for example, Windyty. Watch this a few times and you'll start to understand how a forecast actually works. However, the animations themselves are also provisional pending new information or updates.

 

[SFS]

Excellent points and advice, KG. When I started flying, I was always amazed at the number of pilots that took a 6-hour forecast as gospel, even though they were looking at it when it was 5.5 hours old.

 

[Kermit]

So basically they keep changing the forecast until zero hour. Then it's 100% accurate.

 

[shemandr]

Yeah, try calling your bookie late in the 4th quarter to update your bet based on changing game conditions.

 

[Kings Gambit]

Yeah, try calling your bookie late in the 4th quarter to update your bet based on changing game conditions.

Well, one can learn how to use 'em (forecasts) or hurl insults at 'em. So, I'm sure we can count on you and Kermit to pack up the car and blow town on the next hurricane forecast, 36 hr out, that puts you in the red warning zone rather than betting on favorable updates?

 

[Kermit]

Well, one can learn how to use 'em (forecasts) or hurl insults at 'em. So, I sure we can count you and Kermit to pack up the car and blow town on the next hurricane forecast, 36 hr out, that puts you in the red warning zone rather than betting on favorable updates?

I honestly believe wind is the most difficult forecast to make. But I do believe what I said is basically true. Not meant to be an insult just an observation. I almost didn't post it because I figured it could ruffle some feathers. (Or reef a sail or two.) For the most part I look at forecasts as their best guess with updates in 15 minutes.

 

[Kings Gambit]

There's always a trade-off between predictions from empirical data that are very specific to the point of being useful in a narrow field only, versus more general where they are useful in a broader field. The more specific, the more parameters that are needed; the less specific (more general), the fewer. Greater accuracy but less generality comes with more parameters (specificity). (Rather like binoculars--the greater the magnification, the narrower the field of view.) I'm sure weather forecasting has had to find its place along that continuum. Otherwise, each person would have to have his/her own personal weather forecaster--which you can get w/$$$--for the narrow field of use by that single individual, on that day, in that hour, etc. So, pony up!!

 

[Benny17441]

Let's talk about needs and use in a sailing context. I can understand the usefulness of a land based forecast updated every 6 hours but a Marine Forecast good for only 6 hours is like an oxymoron. OK, the forecasters are sensitive to be called out to be frequently wrong and have responded by shortening the forecast window and that might be OK for Joe Commuter but to issue a Marine Forecast with a 6 hours expiration time is actually a waste of effort and perhaps could be misleading. If I'm waiting for a 36 hours weather window to make a passage I would rather have their best shot at a 48 hour forecast than their 6 hours job. Like Kermit indicates, "So basically they keep changing the forecast until zero hour. Then it's 100% accurate" We do not need, nor expect perfect accuracy but give us a product that is useful and has been put out to the best of their abilities. We will still be saying they got it wrong but will still be thanking them for their work. It is no secret that Trans -Atlantic and Trans Pacific cruiser may subscribe to private weather forecast services issued and tailored to their specific needs. Updates and consultations are transmitted on a daily basis via SSB radio for very specific areas. NOAA provides the raw data to these private services but only issues forecasts for broad areas of general interest. Can't complaint for a free service. It is their Hurricane Tracking group that really makes NOAA a star as they promptly communicate forecasts and data that save lives and property.

 

[Kings Gambit]

Nearly every marine forecast is a 5 to 6 day forecast at the time it is issued, but then updated every 6 or so hours. So, when you get the forecast--it is the "best shot" at that time. The shorter the forecast's future, the better it is likely to be--as I'm sure we all know.

 

[Rick486]

It's not uncommon to hear from sailors about how NOAA's weather forecasts are often wrong and cannot be trusted. I've heard it more than once just this week, including here on this forum. In contrast, I've found the forecasts to be very reliable and quite useful in planning for weekend or longer excursions and especially for executing them. I think it might come down to the user's understanding of a forecast. First, NOAA forecasts are always provisional pending the next update, the provision being, I suppose, that nothing important has changed since the forecast was issued. Updates are every 6 or so hours. So, if you pick up a forecast near the time of next update and then don't bother to get that update 30 min later, then it might appear to you that the forecast you got was "wrong" when in fact it had merely expired while you were out. There's also the zonal nature of the forecasts. For example, "the coastal waters from Pt. Mugu to San Mateo Pt., California, including Santa Catalina and Anacapa Islands...." This zone stretches about 95 n.mi along the coast out to about 40 n.mi. If the forecast is 10 to 15 kt, then somewhere in the zone it blows 10 kt whereas elsewhere it might be blowing 20 at the same point in time. But if in your location the wind happens to drop to 5 kt for an hour or two does that make the 6-hr zonal forecast wrong or unreliable? NOAA voice (VHF) and text (web) forecasts are not "real-time" b/c the media do not allow it. Thus, it's more informative to see an animation of the progress of a forecast over the whole zone(s) using, for example, Windyty. Watch this a few times and you'll start to understand how a forecast actually works. However, the animations themselves are also provisional pending new information or updates.

Over 35 years of sailing I would venture to say that they usually have the big picture correct, i.e. front coming through....and they usually have the wind direction pretty well, but the wind speed is rarely accurate due to local effects, and just plain mistakes. They tend to be a bit useless around big storms because they load up their forecast with "public advice", like "...watch your children...".

 

[capta]

The reality is that some areas, like the Pac coast are fairly easy to forecast and others like the tropics are not. The weather forecasters are also bound by rules that add a huge margin for error, thanks to some who have sued them in the past.
With so much information available on the web these days and a little self education, one is foolish to rely on any 'forecast'. Do your own forecasting for your area and needs and leave the NWS to do the big picture for those too lazy to educate themselves.

 

[Kings Gambit]

The reality is that some areas, like the Pac coast are fairly easy to forecast and others like the tropics are not. The weather forecasters are also bound by rules that add a huge margin for error, thanks to some who have sued them in the past.
With so much information available on the web these days and a little self education, one is foolish to rely on any 'forecast'. Do your own forecasting for your area and needs and leave the NWS to do the big picture for those too lazy to educate themselves.

Yeah-- the best personal forecasting tool one has IMHO is an observation of the weather "the day before." That is, if you want to know what the weather will be like tomorrow, then look hard at what it is like today.

 

[David in Sandusky]

Our weather forecasting system depends on:

The NOAA forecasts (or Canadian equivalent when we are there)
NOAA weather radar app
SailFlow app for local details, and understanding of the wind field forecast
Weather Bug app especially hourly for precipitation
Local knowledge for sea breezes and shaping/stability of wind by nearby land.
Mark 10 eyeballs.

We've learned that when a low or high is passing nearby, wind direction can be random. Low wind on hot summer days is also unpredictable, but we will probably be motoring in any case. A lot of "accuracy" issues with all forecasts (everyone is using the NOAA data for their forecasts!) come from a weather system moving faster or slower than forecast - radar sometimes allows us to tune timing in the short term. Another source of inaccuracy is wind changes under clouds, particularly cumulonimbus.

The scale of the NOAA forecasts has been mentioned by others. Our NOAA forecast is usually the west half of Lake Erie, so system timing can vary +-2 to 6 hours, depending on system speed.

 

[Ken Cross]

We actually live in an area that is nearly impossible to forecast in terms of wind direction. Our predominant winds are from the West but, the Olympic mountains disrupt the flow and it can switch from a north to a south wind in a heartbeat. This is classic Puget Sound and I've come to expect forecasts to be wrong. I mainly watch for big blows which I do pay attention to in the forecast.

We all know that little boys who tell lies turn into weathermen.

Ken

 

[Newport Bob]

Interesting- I dated a weather forecaster many years ago. Told her " you are in the only profession where you can be wrong 50% of the time and not lose your job"
She transferred to the NE BECAUSE "before I could post my forecast and check/look to changing every few days, but here in the NE I need to relook every few hours"

 

[Stu Jackson]

With so much information available on the web these days and a little self education, one is foolish to rely on any 'forecast'. Do your own forecasting for your area and needs and leave the NWS to do the big picture for those too lazy to educate themselves.

So true. In addition to the ideas and sites mentioned, don't forget Passage Weather, with animations two days ahead. We found it very helpful for our trip north last month & 1/2.

Depending on where you are, the fishermen are always very helpful, too.

Like navigation, anyone who depends only ONE source of info is asking for trouble and didn't read the fine print.

 

[dziedzicmj]

With so much information available on the web these days and a little self education, one is foolish to rely on any 'forecast'. Do your own forecasting for your area and needs and leave the NWS to do the big picture for those too lazy to educate themselves.

If you are serious about having a good local forecast, you better learn how to do it yourself. As Capta said, there so many resources available (NOAA and Environment Canada (in NA), radar maps, 500 mb maps, PWSs, SailFlow, WU, etc.) that you can be pretty accurate. There are some very good books to start with.

If you are not serious a local radar map and SailFlow will do. But if you are not that serious you should not complain that someone else did not do your job that well.

Marek

 

[capta]

Yeah-- the best personal forecasting tool one has IMHO is an observation of the weather "the day before." That is, if you want to know what the weather will be like tomorrow, then look hard at what it is like today.

I don't know where you got that idea or perhaps you are being facetious, but yesterday it was sunny, very hot and almost flat calm. Today it's been overcast, cool, raining at several inches an hour and blowing above 40 knots. Oops?
Just as a side note, Chris Parker's forecasts on Mathew have been about 24 hours behind what I've found on Passage WX, http://globalsailingweather.com and several other models I've run when the internet allowed, all free.

 

[Kings Gambit]

Statistically speaking, of course.

 

[JamesG161]

First of all I am happy that Capta is responding after he weathered Tropical Storm Matthew

The only reliable forecasting model that I have seen is...
What, Where, When model.

What you get is Where you are When you are sailing. (usually about 1-2 hours)

Macro Forecasts ( of 12 hours intervals) do reliably forecast warnings of major events and macro conditions.

I wonder where the phrase of "Keep a weather EYE out" came from. I am amazed that sailor of just 40 years ago survived the weather and the sport continues.

Here is where all of the base weather reporting data comes from...
http://www.nws.noaa.gov/om/marine/point.htm
NOAA is now globally linked, since perhaps the weather is man's common link world wide.
______
BTW we are going sailing in them Mississippi Sound Friday-Sunday with the best macro forecast since July 4th and we are prepared to weather any surprise weather event.

Jim...

PS: Birds ruffle their feathers mainly for weather related reasons.