Problems downloading ECMWF forecast

Little Dorrit

Little Dorrit

Well-Known Member
Joined
27 Jan 2010
Messages
1,295
Location
South Coast
Visit site
I am not able to download forecasts on XyGrib from ECMWF, am I the only one experiencing this problem? Perhaps I need to reinstall? All other forecasts appear to be accessible.
 
Very little in ecmwf is free, looks like just pressure xygrib can get ->

QetR9AX.png


Computer models available are –

Global models.
NOAA-GFS. 0.25, 0.5, 1.0°. 00, 06, 12, 18 UTC, 1-10 days.
DWD ICON, 0.25°. 00, 12 UTC , 1-8 days.
ECMWF – Surface pressures only, 0.5°. 1-3 days.
Météo Frsnce Arpege, 0.5°, 00, 12UTC, 1-4 days.

The XyGrib FTP Service for Linux, MacOS and Windows - Franks-Weather - The Weather Window

windy.com has the full data but not as gribs.
 
with ECMWF.., only "essential" fields are available in grib format for free - this does not include 10m winds

it is only MSLP.., 500MB HGT.., 850MB Temp.., and 850MB Winds

WMO Essential
 
GHA said:
Very little in ecmwf is free, looks like just pressure xygrib can get ->

QetR9AX.png


Computer models available are –

Global models.
NOAA-GFS. 0.25, 0.5, 1.0°. 00, 06, 12, 18 UTC, 1-10 days.
DWD ICON, 0.25°. 00, 12 UTC , 1-8 days.
ECMWF – Surface pressures only, 0.5°. 1-3 days.
Météo Frsnce Arpege, 0.5°, 00, 12UTC, 1-4 days.

The XyGrib FTP Service for Linux, MacOS and Windows - Franks-Weather - The Weather Window

windy.com has the full data but not as gribs.
Click to expand...
Thanks, that's strange as until recently I was able to download but perhaps I was mistaken?
 
A brief summary of the history of this...

WMO "non-essential" ECMWF fields are not in the public domain - they can only be redistributed under certain rules. The rules for redistributing them in GRIB format are particularly tight

For a number of years, both Squid and Predict Wind were paying EC for the right to redistribute (sell) the ECMWF in GRIB format.

I had a subscription - with both Squid and PW at different times.., depending on their other offerings. You could downloaord the ECMWF surface winds (and other fields) in GRIB format.., and open them in the GRIB viewer or nav software of your choice.

Then, a few years ago the EC decided that both Squid and PW were violating the terms of the redistribution agreement with the ECMWF GRIBs

The agreement required that anyone reselling the forecast provide a "value added" component.., it was not okay to redistribute the forecast in its native form - people were supposed to buy that direct from EC.., but practically speaking, the minimum cost would be unreasonable for a single-user.

For Squid.., what they were doing was re-scaling the surface wind speeds slightly.

I do not know what PW did.., but i suspect it was something similar

The EC decided this was a trivial change and that it did not meet the value added requirement: they stopped providing the ECMWF to Squid and PW for resale in GRIB format. This was a few years ago.

At PW you can still view the ECMWF - but it is like windy - you can only view it in the PW app.., you can not download it as a GRIB for use in other programs.
 
Laser_310 said:
A brief summary of the history of this...

WMO "non-essential" ECMWF fields are not in the public domain - they can only be redistributed under certain rules. The rules for redistributing them in GRIB format are particularly tight

For a number of years, both Squid and Predict Wind were paying EC for the right to redistribute (sell) the ECMWF in GRIB format.

I had a subscription - with both Squid and PW at different times.., depending on their other offerings. You could downloaord the ECMWF surface winds (and other fields) in GRIB format.., and open them in the GRIB viewer or nav software of your choice.

Then, a few years ago the EC decided that both Squid and PW were violating the terms of the redistribution agreement with the ECMWF GRIBs

The agreement required that anyone reselling the forecast provide a "value added" component.., it was not okay to redistribute the forecast in its native form - people were supposed to buy that direct from EC.., but practically speaking, the minimum cost would be unreasonable for a single-user.

For Squid.., what they were doing was re-scaling the surface wind speeds slightly.

I do not know what PW did.., but i suspect it was something similar

The EC decided this was a trivial change and that it did not meet the value added requirement: they stopped providing the ECMWF to Squid and PW for resale in GRIB format. This was a few years ago.

At PW you can still view the ECMWF - but it is like windy - you can only view it in the PW app.., you can not download it as a GRIB for use in other programs.
Click to expand...
Thanks, that's an excellent summary of things; nice to know I'm not missing out although I'm obviously going bonkers thinking I had previously been able to access them.
 
Nick White, the developer of Expedition nav software, mentioned to me that he was working on a way to obtain the right to distribute ECMWF for use within Expedition - but that was before the whole virus thing...

it's a bit problematic, as it was not.., and presumably will not be, legal to use ECMWF in most races - and racing sailors are the ones most likely to pay for it.

a few races modified RRS 41 in the SI's to permit use of information that can only be obtained for a fee - if everyone could obtain it for a fee.., but most races did not do this

so.. you could download it up to your warning signal, but not after that. Even if you really like the ECMWF.., by the time new runs of other free permitted models are available.., you would rather use a new run than an older run of ECMWF
 
Why do you want ECMWF data? Granted it is the most advanced model, but it is only available twice a day as opposed to 4 daily for GFS and ICON global. When it does run, ECMWF comes about 2 hours later than the others. So, for 4 hours each day, GFS/ICON will be based on data 12 hours more recent than ECMWF. For a further 12 hours per day they will be based on data 6 hours later. For much of the time, this will make little difference but, on some occasions, the later data may be critical.

Take a look at the Windy.com Comparison of GFS, ECMWF, ICON-EU and Metroblue,. Differences between models are usually well within the model and weather uncertainty.

ECMWF probably out-performs the rest when used by RTW navigators. For us ordinary mortals, I suggest that you stick to GFS/ICON. If not too near the western area of coverage, ICON-EU might do a little better for the next 2-3 days. When we are cruising, I stick to the GFS.
 
I picked this up from Attainable Adventure Cruising (www.morganscloud.com), a subscription based service for offshore voyaging. The link was given in a comment about weather reporting, where a user stated that ECMWF is now free. I am new to using all this data, so not sure how relevant the link is but thought it would be of interest based on some of the comments here. What do you think?

Great News: ECMWF moves towards the public with weather model data - now free and accessible to all

“Medium-range, extended-range, and long-range forecast charts of temperature, wind, precipitation, clouds, and ocean waves are just some of the products that are becoming available. The likelihood of extreme conditions, as well as tropical and extratropical cyclone activity, are also included.”

Until now, the full access to various forecast charts was restricted to members of WMO (World Meteorological Organisation), the national meteorological and hydrological services of ECMWF’s Member and Co-operating States* and indeed commercial customers.
Click to expand...



ECMWF ECMWF | Charts
 
Last edited by a moderator:
This is certainly a change of heart by the owners of ECMWF, the European Met services including the U.K. The displays are interesting but I come back to my query about usefulness. Are these more useful than the displays on Windy.com? This lets you look in more detail at the same data. For most of us, forecasts issued more frequently, using later data and earlier relative to the base data time are more useful.
 
franksingleton said:
Why do you want ECMWF data? Granted it is the most advanced model, but it is only available twice a day as opposed to 4 daily for GFS and ICON global. When it does run, ECMWF comes about 2 hours later than the others. So, for 4 hours each day, GFS/ICON will be based on data 12 hours more recent than ECMWF. For a further 12 hours per day they will be based on data 6 hours later. For much of the time, this will make little difference but, on some occasions, the later data may be critical.

Take a look at the Windy.com Comparison of GFS, ECMWF, ICON-EU and Metroblue,. Differences between models are usually well within the model and weather uncertainty.

ECMWF probably out-performs the rest when used by RTW navigators. For us ordinary mortals, I suggest that you stick to GFS/ICON. If not too near the western area of coverage, ICON-EU might do a little better for the next 2-3 days. When we are cruising, I stick to the GFS.
Click to expand...

ECMWF has four model runs per day, but the 06/18 UTC runs are much shorter (only to step 90 for the HRES and Step 144 for the ENS model). The dissemination schedule for the files might help in determining which steps you can make use of from a forecast run: Dissemination schedule

Unfortunately, ECMWF has always had a non-redistribution licence for real-time data and will continue to do so for a while yet. Encryption was proposed as a means of avoiding needing to make random/arbitrary scaling and to remain compliant with the licence. This is the option taken by PredictWind, and a few others (such as Windy who make sure that the data is not readily retrievable from its display). There are also guidelines published on what is / is not allowed in the context of the licence: Guidelines for Interpretation of the Rules and Standard Licence Agreement

However, in positive news, by 2025 the ECMWF strategy aims to have an open and free package of data, with changes coming over the next few years to achieve this. The first step was the open data licence quoted above, but it only applies to the historic data (any data that can't be used to forecast). The ECMWF real-time charts have also recently been opened up under the same type of licence, as reported by Piddy :)
 
In GRIB format, no. The only free/open data from ECMWF is the WMO Essential dataset: WMO Essential
ECMWF has added 06/18 tropical cyclone data recently, as well.

Some websites display the 06/18 runs. But for Windy this is a subscriber feature.
Weather.us is also a good site for ECMWF charts, but again the 06/18 is a subscriber feature.
 
franksingleton said:
Not as far as I know. I have not checked but they may just be for the purposes of the 4 D VAR analysis scheme.
Click to expand...

Not just for the 4-D VAR. This 06/18 UTC data comes from an optional programme which was funded by the majority of the Member States (but not all!) in addition to the 'core' 00/12 UTC runs. The output was intended to be the boundary conditions for Member States to initialise limited area models (LAMs). In 2018, ECMWF agreed that these could be part of the commercial catalogue.

The same model is run at 06/18, with fewer steps, and so all parameters are produced as per the 00/12 UTC.
 
By way of introduction, Piddy is my Dad (Andy) and I work at ECMWF, so if you have questions, I'm happy to answer them (wherever I can!) :)
 
You must log in or register to reply here.

Other threads that may be of interest

mogmog2
Maxwell windlass 10 VW looks interesting (to a windlass noob)
2 3
Replies
59
Views
1K
mogmog2
mogmog2

Share this page

Top