@gkikas-lgpz
That may well be, but as an English speaker, and the language being used is English, the appropriate meaning of “basic” is “forming an essential foundation or starting point.” As Windy has the ECMWF model as it’s default, the impression of the description on the Windy forecasts is that the GFS is somehow inferior, simple, limited. In fact, in several ways, the GFS gives a more useful service to me as a sailor than ECMWF.

Windy describes the GFS as a “Basic model.” This is incorrect. The GFS calculates in a 13 km grid at 127 levels with a ceiling of about 80 km. Calculation of the physics is highly complex. Can I suggest that Windy re-phrases it’s comments.

@gkikas-lgpz
That may well be, but as an English speaker, and the language being used is English, the appropriate meaning of “basic” is “forming an essential foundation or starting point.” As Windy has the ECMWF model as it’s default, the impression of the description on the Windy forecasts is that the GFS is somehow inferior, simple, limited. In fact, in several ways, the GFS gives a more useful service to me as a sailor than ECMWF.

Windy describes the GFS as a “Basic model.” This is incorrect. The GFS calculates in a 13 km grid at 127 levels with a ceiling of about 80 km. Calculation of the physics is highly complex. Can I suggest that Windy re-phrases it’s comments.

@idefix37
Ah! Perhaps it is cabin fever with a “summer” of being in harbour and seeing long spells when we have been unable to move. I now see my Windy problem.

GRIB forecasts generally, PocketGrib, WeatherTrack, Weather4D, iGrib, PassageWeather, XyGrib, to name a few, all show clearly the base time for forecasts. It is always the first time for which output is available. With Windy.com, the first time for which output is shown is always midnight regardless of whether the base time is midnight or midday. Not being a habitual user of Windy, this had never really crossed my mind. Presumably, in the evening (c1900 UTC) upload, the information shown prior 1200 UTC is actual data. To be fair, Ventusky does something similar in that they show actual data for the past 7 days. In their case it is more obvious, to me, at least, and I automatically remember that the GFS updates at around 05, 11, 17, 23 UTC. ECMWF updates around 07 and 19 UTC but, not having the same operational commitments, may not be quite so reliable in times of updates. Windy predictions of update times seem overly cautious. Maybe, you should not be so specific in giving next update times.

Maybe it is more of a caveat emptor, but it would be clearer if Windy only showed data from the T=0 times, ie for the 1200 UTC forecast, only show data after that time.

@anhinga
Apologies. I am unaccustomed to your site.
This evening, 5 August, at about 2000 UTC I got a forecast issued by Windy at around 1930 UTC. The forecast period started at 0000 UTC. Presumably9 it is the forecast issued by ECMWF at 0700 UTC.
I do not know how to paste images into this post to make clear what I am saying.

@anhinga said in ECMWF updates:

@idefix37
Thank you. Today Windy updated at about 0730 UTC. BUT, before that time, I should have been seeing the forecast updated the previous evening, ie starting at 1200 UTC the previous day. In fact, at about 0630 UTC, I was seeing a forecast starting at 0000 UTC. This has been happening for a few days at least. Windy appears to be issuing forecasts starting at 0000 UTC before the ECMWF issue time. Windy is not currently issuing a forecast starting at 1200 UTC.

@idefix37
Thank you. Today Windy updated at about 0730 UTC. BUT, before that time, I should have been seeing the forecast updated the previous evening, ie starting at 1200 UTC the previous day. In fact, at about 0630 UTC, I was seeing a forecast starting at 0000 UTC. This has been happening for a few days at least. Windy appears to be issuing forecasts starting at 0000 UTC before the ECMWF issue time. Windy is not currently issuing a forecast starting at 1299 UTC.

I seem to have a problem with ECMWF updating on the app.

If I look at a forecast in the early morning, say before 0500 UTC, the forecast covers from midnight onwards. But, ECMWF issue time is about 0700 UTC. It cannot be the forecast using midnight data
If I look at a forecast at around 2200 UTC, again, the forecast starts at midnight, not midday. What I see cannot be the forecast using midday data.

Can someone please explain what is happening.