openWRF free high resolution GRIBs now cover the entire Med

D

davidgal

New Member
Joined
4 Nov 2011
Messages
15
Visit site
The openWRF model runs now covers the entire Mediterranean in 12km and 4km resolution. The model output is freely available and focused on the sailing community's needs in the following forms:

- Downloadable Grib Files - "WRF Gribs" (Site also provides Skiron model Gribs)

- "Wind & Waves" interactive application that can choose any area and then produce meteograms and meteocharts for points of interest

- Regular weather charts for all areas covered - "Charts".

- Lightweight Grib files (wind and significant wave only) that can be obtained via saildocs email system.

Link to openSkiron.org site

Enjoy!

wrf-index-map2.jpg
 
davidgal said:
The openWRF model runs now covers the entire Mediterranean in 12km and 4km resolution. The model output is freely available and focused on the sailing community's needs in the following forms:

- Downloadable Grib Files - "WRF Gribs" (Site also provides Skiron model Gribs)

- "Wind & Waves" interactive application that can choose any area and then produce meteograms and meteocharts for points of interest

- Regular weather charts for all areas covered - "Charts".

- Lightweight Grib files (wind and significant wave only) that can be obtained via saildocs email system.

Link to openSkiron.org site

Enjoy!

View attachment 58976
Click to expand...

Well thanks. However I bet it will be just as rubbish as all the others. Try VHF ch 66. Proper.
 
I found the site without a problem

It does not explain whether the model used is any different from PWC and PWG as used by PredictWind for example
Is the forecast a more detailed resolution that other sites?

TS
 
tudorsailor said:
I found the site without a problem

It does not explain whether the model used is any different from PWC and PWG as used by PredictWind for example
Is the forecast a more detailed resolution that other sites?

TS
Click to expand...
The model used for the openWRF gribs is the NOAA WRF model version 3.7 with the latest ARW core. Basically it is the same non-hydro-static mesoscale model known as NAM (North American Mesoscale Model) that provides all detailed regional forecast products for the US.

openWRF is run several times a day on relatively small domains that cover the entire Med. Initialisation data is from the NOAA GFS at 27km resolution and the topography, terrain coverage etc are from the latest Modis data sets. Altogether it is a totally modern non-hydro-static model providing larger area coverage at 12km resolution for 5 days and smaller area coverage at 4km resolution for 48hrs

Predict wind does'nt exactly provide insight to their models but at least the 8km model and the 1km models are also most likely non-hydro-static models that are good for simulating local weather phenomena. I doubt if they have much coverage at 1km resolution as computing power would be the limiting factor.

The openSkiron Gribs also provided on the site are at produced by the University of Athens with their SKIRON model which is an older type of model based on the ETA model (which used to drive NAM in the past).

Are the openWRF gribs worth anything (which seems to be the concern in this thread)? - If you are in the Med just try them out. In a couple of days you will know for sure. I have a lot of very positive feedback from users and I myself spend most of the year sailing in the Med and I will use any grib files that I can find.

Up to two years ago the only free and open Gribs one could find for the Med were GFS gribs. GFS is a fine hydro-static global model that is designed to provide synoptic scale weather forecasts and not regional forecasts reflecting local phenomena. The European ECMWF model is also global and hydro-static and may be better and slightly higher resolution than GFS but has anyone ever seen a free grib file from them?

Coastal sailing in the Med is very much affected by local weather conditions that are not at all modelled well by GFS. I personally felt this discrepancy and this is what drove me to establish a site that can provide Grib files from regional non-hydrostatic models on a free and open basis for the sailing community. Sailors helping one another and no more. I turned to all the national weather services and other producers of local weather and asked them to share their grib files openly and freely. Only the University of Athens agreed to share the SKIRON output which I post process and publish it in usable form for leisure sailors.

Last year I began running productions of WRF gribs on 3 clusters of used servers (72 CPU's in total). I found these gribs to be very accurate and started to publish them on the site as well.

The bottom line is what Gribs from regional models that reflect local weather, are freely and openly available for the Med today? Not too many:
- Meteo Consult have 11km gribs with wind only for 48 hours
- Meteo France Arpege - gribs at 11km over all the Med
- Meteo France Arome - gribs at about 3km covering France only
- openSkiron gribs - 11 km over all the Med including WAM wave data
- openWRF (at the openSkiron site) - 12 and 4 km covering all the Med with WAM wave data too

Besides the above it is only GFS which is first tier forecasting and lacks the more detailed second tier which is the forte of the non-hydro-static regional models.

The proof is in the pudding! Try the openWRF gribs. If they are not good - please let me know.

David
 
Last edited:
OpenWRF is the only forecast that got it right today for the Kolpos Lakonikos. We usually check Windy's ECMWF, but that hasn't even got remotely the right idea today. It's a tricky area here, the weather systems from both sides of the Peleponnese seem to merge here and constantly fight for attention, pushing the border east or west and you can have completely different weather on either side of the Gulf depending on which one is winning at any given time.

For same-day forecasts we normally use Meteo Consult's spots (and create custom ones for where they have none), which do however only predict accurately until midnight, but usually very precise. Then they drop off into fantasy land until midnight rolls around and you get an updated forecast - seems they use a different (more detailed?) model for the same day forecast.

We've tried OpenSkiron last year in the Balearics a couple times, but it wasn't as good as Meteo Consult there, and I can't remember whether we've used the OpenWRF or the OpenSkiron model. But today after groping for any kind of idea about the weather, I've remembered this one and gave it another try and voila, it matched what was actually happening (which is 25 knots gusting to 40 from N-NE). Will check it regularly from now on! :encouragement:
 
You must log in or register to reply here.

Share this page

Top