Windguru: inconsistencies?

I use Windguru a lot. I have always found their forecasts to be quite accurate to the hour. Obviously there are occasionally variations due to local phenomena but as a guide the wind/rain forecasts are good. I only look elsewhere for the visibility - Windguru doesn't show fog (but there are probably others better than me who could work it out themselves!!!)
 
When people say it's not accurate - how do you know? How do you log it?

I'm involved in an outdoor event in Azerbaijan and have been keeping a log (screenshots of forecast & actual into evernote) since January , windguru forecasts have been accurate within a few M/S when compared with historical data from the local airport. Within a day or 2 anyway, longer term the winds will lessen or increase and the time scale change but general trends are pretty good.
 
Back to #13. This output is direct from a computer model. The models carry humidity as one of the variables. I guess that there is an algorithm that relates humidity to cloud amount at various levels. You can have varying amounts of cloud at different levels in reality as well as in the models. I would not take the numbers too seriously - if at all.

Re accuracy, see #12. The only change is that the U.S. GFS now runs on a 0.125 degree grid and provides output -GRIBs - on a 0.25 degree grid. For all or most practical purposes, it will not be bettered - on average that is. Claims of accurate prediction of small detail, say <50 km are strictly for the birds. See http://www.metoffice.gov.uk/learning/science/hours-ahead/mesoscale for example, although some of the numbers eg the 40 km global grid length are out of date.

I am not aware of any organisation/firm etc, other than national weather services which can analyse fine detail let alone predict it. Most start with the GFS 0.25 degree data and add no other information. Apart from now casting national weather services will not predict small detail such as thunderstorms in other than a general or probabilistic manner.
 
irrespective of how its produced , whatever models run etc etc, all I care is proof of the pudding , even if the forecasts were just guesses !!!,

In my experience ,in my sailing area , west of scotland , XC and wind guru have proved to be the most accurate and reliable (which for WOS has been suprisingly high) , this proof through use more than satisfies my needs , im too stoopid to make up my own forecasts nor have time nor inclination .

Quiet scary at time how accurate they can be , can almost set your watch by them !!!

All we need is a weather controller to dial in some decent breeze , getting a bit cheesed off with current crop of winter gales
 
markhomer said:
In my experience ,in my sailing area , west of scotland , XC and wind guru have proved to be the most accurate and reliable (which for WOS has been suprisingly high) , this proof through use more than satisfies my needs ...
Click to expand...

My experience too. Whatever model they use to interpolate between source data points works very well for me.
 
I have said on several occasions that which GRIB or GRIB derived product you use depends on ease of access to you and ease of use – again to you. The basic data are the same. As ever with meteorology there are caveats. In this case, there can be short-lived differences due to precise update times.

The U.S. GFS, the basis for virtually all these services is remarkably good. We can see quite short weather windows several days, 5 or 6 ahead. Almost uncanny at times


We prefer to use zyGrib not because it is a better forecast but for several practical reasons.

· I can choose my own area of interest – currently Scillies across all of Biscay.

· I can select up to 10 days ahead, with intervals from 3 hours and grid spacing down to 0.25 degree. NOTE these are not interpolated but model grid point values. The GFS calculates on a 0.125 degree grid and provides data on a 0.25 degree grid.

· I can choose which elements I want to download including sea state and, even, high level winds.

· I can view on a grid or select any point and get a time sequence.

· I can download directly even with low bandwidth that will not support a browser – recently my situation in France.

· With no WiFi, I can use email to get the same data and view it in zyGrib

· The data are saved automatically. When SWMBO asks “What did you say for Camaret to Audierne tomorrow?” I can always tell or show her.

Sites such as XCWeather, WindGuru, Passageweather etc are equally as good but with much less flexibility. A plethora of tablet apps do much the same as zyGrib but, again. With a little less flexibility.


If limited area models are available I am happy to look at them – if they are free. I would not pay for any. I have a listing of many, certainly not all, GRIB or GRIB derived services at http://weather.mailasail.com/Franks-Weather/Grib-And-Objective-Forecasts-Reviewed.
 
Last edited:
Sorry to bump this to the fore, again...but in recent days, I've noticed that the Windguru site I've been looking at all summer, is largely annotated in a language other than English...

...but I'm very sure it wasn't, until this week. Any idea why it seems to have changed? Here it is: http://www.windguru.cz/se/index.php?sc=28765
 
dancrane said:
Sorry to bump this to the fore, again...but in recent days, I've noticed that the Windguru site I've been looking at all summer, is largely annotated in a language other than English...

...but I'm very sure it wasn't, until this week. Any idea why it seems to have changed? Here it is: http://www.windguru.cz/se/index.php?sc=28765
Click to expand...

If I try to look at that it changes "se" to "int". Are you seeing it in Swedish, by any chance?
 
I can easily believe it...but why, I wonder? I did try changing that address back to what it ought to be, but my laptop automatically displays the furrin' version. Most odd.
 
dancrane said:
I can easily believe it...but why, I wonder? I did try changing that address back to what it ought to be, but my laptop automatically displays the furrin' version. Most odd.
Click to expand...

Are you on a train on the East Coast Main Line? Their internet access goes through Ericsson in Sweden, so clever websites think you are Swedish. What version of the BBC news website do you get by default - UK or world?
 
I'll give it a try tonight if I can remember, thanks JD. Out & about with only the phone at the moment...and naturally, the phone doesn't make the same mistake.
 
dancrane said:
I'll give it a try tonight if I can remember, thanks JD. Out & about with only the phone at the moment...and naturally, the phone doesn't make the same mistake.
Click to expand...

If that doesn't bring enlightenment, have a look at the cookies on your browser. I've just checked, and www.windguru.cz has put a cookie called "langc" with value "en" on mine, which may be significant ...
 
Wind guru is primarily a windsurfing weather site and tends to over estimate the wind...in my forecast areas of interest anyway.

The programmes those weather sites run ( XC weather , Big Salty etc) seem to have variable accuracy depending on you ' local' spot. Topography probably has a lot to do with it and the alogarithms maybe aren't tweaked enough to cope? Or something fancy and computer geeky like that.

For my areas of interest XC seems more accurate, Wind Guru very optimistic, Met Office generally rubbish.
 
Regarding WindGuru's web address and cookies, I've just stopped all cookies from attaching themselves to my laptop...and deleted all previous attachments...

...I daresay I'll regret that within the hour, but meanwhile I am getting WindGuru in English again. :encouragement:

Although, curiously the address at the top of the page is still... http://www.windguru.cz/int/index.php?sc=28765 ...which I thought was the part giving the trouble.
 
You must log in or register to reply here.

Share this page

Top