SUJET :

I would recommend ECMWF data . The only free source of ECMWF that I see is here : www.ecmwf.int/en/forecasts/datasets/wmo-essential , but the only fields of interest to sailors are surface pressure and 500 mb HGT. Predict Wind has ECMWF with surface "wind rain cloud temp cape gust", but I don't know where they get that data.

Attached is a simple bash / curl script to get ECMWF data.

The data offered free to the public by ECMWF includes only a few fields and at 24hr steps which is really not worth the effort. Flipping the synoptic charts on their website gives much more information.
Predict wind purchases ECMWF data and apparently they can afford it.

Thanks for the pointer.
David

Cook écrit: I would recommend ECMWF data.

Hello,
In addition of the comment of David, which emphisize the fact that ECMWF provide a lot of good charts publicly available and very limited number of data in GRIB files ...
Thanks for the script.

Predictwind.com is a commercial company selling products (including their own models, PWG and PWE - www.predictwind.com/why-predictwind-forecast; it has a commercial licence with ECMWF quite expensive (www.ecmwf.int/en/forecasts/accessing-for...s/licences-available).
In fact ECMWF is often use as shortage of the ISP model used by the European Centre for Medium-Range Weather Forecasts (ECMWF, www.ecmwf.int/en/about).
IFS is an hydrostatic model, reasonable for large scale prediction (www.ecmwf.int/en/research/modelling-and-...atmospheric-dynamics) and ICON is a nonhydrostatic model, which take into account the geography, which is quite important for costal navigation (www.dwd.de/EN/research/weatherforecastin...con_description.html).
Looking at the presentation given by M. Buchhold, DWD in September 2017 (www.cosmo-model.org/content/consortium/g..._ICON-EU_WG5_CGM.pdf), it appears that ICON model (and sub-model) has in general a better precision, even for wind (mean speed and gusts).

There are numerous evolutions and changes in NWP models and the way they are used, which means that new data sources will regularly (time scale of couple of years) be provided and old sources removed. Clearly we have to follow the evolution of ECMWF, as it may provide more data freely available.

Hello. Please add model ecmwf and wrf in the app (xy grib) and add snow maps for ARPEGE model.

Hello and welcome,

I would love to add ECMWF gribs to our server. Problem is that access to this model is in the order of tens of thousands of EUR per year and even then the Grib files may not be redistributed. As we are providing a free and open service on a purely voluntary basis we cannot dream of purchasing this data.

WRF is a regional model and usually its runs cover limited areas. There is no general WRF coverage of the world or even large areas. These files are best distribute in pre-cut form such as can be found on openskiron.org . Look for openWRF Gribs.
We are planning to add some large area regional models such as NAM, ARPEGE-EU and ICON_EU to the "on-the-fly" grib server in the future.

Regarding Arpege and snow: The data provided by Meteo France does not include snow fall or snow depth. The only snow related parameter provided is SNOM snowmelt which we chose not to include as it is really of secondary importance.

David

The ECMWF free and open source mentioned in this thread will be added to the next version of XyGrib