Windguru: how good?

[TiggerToo]

Is it?

 

[GHA]

Is it?

Can be. But compared to what? The free ones are all looking at the same data anyway so it's just a matter of display.

In a moment there;ll be lots of replies about xXYYZZ is the best or Met office is rubbish all based on one or two particular afternoons from someones memory.

AN option would be to watch a buoy every day and see how it tallies up with the previous forecats.
http://www.ndbc.noaa.gov/station_page.php?station=62304

I recently watched windguru on and off for 6 months on an outdoor event compared to an airport weather station and it was very accurate. Evernote works great for something like that, just cut and paste a screenshot snip from both windguru and a weatherstation every day and see how they tally up.
Please let us know how it goes

 

[dimdav]

I've been using windguru for 5 years now and find it very accurate for the clyde. I compare it to met office, xcweather and windguru is usually right for the local conditions. might not be so accurate for all locations but works for me

 

[MoodySabre]

Works OK for me. As it is for coastal winds it seems better than Met Office 'land' forecasts.

 

[prv]

Can be. But compared to what? The free ones are all looking at the same data anyway so it's just a matter of display.

If you give WindGuru a few Czech blatts per year, you get access to some other models. Certainly not unique to them, but I've found the WRF 9km noticeably better than the basic GFS used by most of the free sites, including WindGuru in free mode.

Pete

 

[maby]

It is pretty accurate compared with most forecast resources. The other one that I use quite often is uGrib. It seems to use pretty much the same model as WindGuru, but you can select an area and download the predictions for the next few days in batch mode - can be useful to have loaded on the laptop if you are going to have unreliable data communications while at sea.

 

[Dipper]

I've always found it very good.

 

[GHA]

If you give WindGuru a few Czech blatts per year, you get access to some other models. Certainly not unique to them, but I've found the WRF 9km noticeably better than the basic GFS used by most of the free sites, including WindGuru in free mode.

Pete

Not having a dig, or rather having a dig at our (mine anyway..) ability to see what we want to see , I find it impossible to make any kind of judgement without some hefty research with numbers which didn't come from our subjective looking up
SO lets have a look at sandetti light for the next little while.....

http://www.ndbc.noaa.gov/station_page.php?station=62304

 

[alahol2]

I think probably the same data and model but I have come to prefer the Windfinder layout. Also the Windfinder 'super forecast' seems to add an extra layer of accuracy.

 

[Koeketiene]

Not perfect but better than most.

 

[prv]

Not having a dig, or rather having a dig at our (mine anyway..) ability to see what we want to see , I find it impossible to make any kind of judgement without some hefty research with numbers which didn't come from our subjective looking up
SO lets have a look at sandetti light for the next little while.....

http://www.ndbc.noaa.gov/station_page.php?station=62304

I don't understand what you're trying to say, but for comparison here's the WRF 9km forecast for the same spot:

Note that this is over a shorter period (they don't run the more accurate forecasts so far in advance) so you need to look at the days and times to compare like with like.

Pete

 

[TQA]

Windguru does it for me. Better accuracy on wind speed/direction than rain though.

If I want to get an overview of conditions over a period of time then Passagemaker is good

 

[Foxy]

not sure who posted it on here and I believe it's just the GFS model; however, www.windyty.com has a nice graphical representation which helps you get a good general view . Lots of features if you search through them.

Passage of lows are always tricky to model though to be fair, so it's always a bit of a lottery when they are concerned.

 

[oldgit]

During a recent prosecution by the MCA , an expert witness stated that a Windguru forecast would not be accepted as a legitimate forecast to base a commercial journey on.Only a Met Office forecast would be approved .
A skipper relied on a Windguru forecast to allow him to undertake a passage and did not bother with a MET inshore forecasts.
The voyage ended badly and the skipper is now doing a prison sentence.
The verdict of the weather expert,well known in RTW racing and major offshore yachting events,was that the raw data supplied by such sites is far to general and simply cannot be fine tuned to take account of local variations.
So if something goes wrong.best not mention your only source of weather was Windguru.

 

[lpdsn]

During a recent prosecution by the MCA , an expert witness stated that a Windguru forecast would not be accepted as a legitimate forecast to base a commercial journey on.Only a Met Office forecast would be approved .
A skipper relied on a Windguru forecast to allow him to undertake a passage and did not bother with a MET inshore forecasts.
The voyage ended badly and the skipper is now doing a prison sentence.
The verdict of the weather expert,well known in RTW racing and major offshore yachting events,was that the raw data supplied by such sites is far to general and simply cannot be fine tuned to take account of local variations.
So if something goes wrong.best not mention your only source of weather was Windguru.

Problem with that is that the Met Office forecast is not particularly reliable (mainly because of in-built vagueness and strong suspicions of a fiddle factor to scare people off the water).

I prefer to use the GFS and ECMWF models and add my own judgement to it. The Irish HIRLAM model is very useful, and I've certainly noticed that it seems to include differences from the main models that tie in with my local experience in certain areas. http://www.met.ie/forecasts/short-range.asp

I presume the UK has a similar small scale hour by hour model, which makes you wonder whether keeping it secret then giving a forecast using terms like 'soon' or 'later' is really that ethical.

 

[GHA]

I don't understand what you're trying to say,

What I mean is that it's unlikely anyone here has much of a solid idea just how accurate any of the forecasts are, it's a lot of work to log the forecasts a few times a day then find a weather station with historical data and a hopefully accurate anemometer at 10m to compare the numbers with.

http://aprs.fi/ has a good selection of ham WX stations with lots of data, but most seem to be in built up areas which might lower the readings a bit. And of course Frank, as ever. has a good selecton. http://weather.mailasail.com/Franks-Weather/Nw-Europe-Official-Weather-Actuals

And I just subscribed to windguru for a year - Ta

 

[Vara]

During a recent prosecution by the MCA , an expert witness stated that a Windguru forecast would not be accepted as a legitimate forecast to base a commercial journey on.Only a Met Office forecast would be approved .
A skipper relied on a Windguru forecast to allow him to undertake a passage and did not bother with a MET inshore forecasts.
The voyage ended badly and the skipper is now doing a prison sentence.
The verdict of the weather expert,well known in RTW racing and major offshore yachting events,was that the raw data supplied by such sites is far to general and simply cannot be fine tuned to take account of local variations.
So if something goes wrong.best not mention your only source of weather was Windguru.

Can you post dates, place and time of the court proceedings as I can't find any trace of it and would be interested to read the transcript of the trial.

 

[prv]

What I mean is that it's unlikely anyone here has much of a solid idea just how accurate any of the forecasts are, it's a lot of work to log the forecasts a few times a day then find a weather station with historical data and a hopefully accurate anemometer at 10m to compare the numbers with.

Ah, I see. Yes, such a comparison would be useful. The only sensible thing is to automate it, not manually collate the data, but I can't be arsed to write the necessary code

The Met Office do do something along these lines, in a limited way. I found it on their site the other week.

Pete

 

[KAL]

Not used it, but I do use a combination of: Met Office, Meteo France, Passageweather, uGrib and good old Sailingweather (Simon Keeling's site - which I find to be excellent).

Although there are differences between these, they aren't usually that wide and the spread of opinion helps me to make up my own mind.

 

[gjgm]

Quite interesting link here as to who is using which Grib models.. at least it might save duplicating.#
http://weather.mailasail.com/Franks-Weather/Grib-And-Objective-Forecasts-Reviewed

Basic lesson I learned from Frank's pages is that in stable weather most forecasts are good and in unstable weather one might be proving better (at that time) than another, so do some monitoring and accept the weather is (currently) unpredicatable. The more, different modeled,sites that have the same forecast, the better of course.
For sea state, I like passageweather.