Wind gusts in ECMWF

Uskar

Hello all,
I use windy for sailing, boating, being outdoors and as a everyday weather tool. I think it is great :) .

However, something that I cannot understand is the values of wind gusts. In particular for the ECMWF model. It is quite often that they are almost about 2x as high as the average wind. So given a forecasted average wind of 10 m/s, wind gusts are forecasted to 20 m/s.

To me this seems as unrealistic high values. Imagine that you are out sailing on the sea with a wind of 10 m/s, then a gust at 20 m/s or 18 m/s would be extremely strong...!

Maybe I do not understand the concept of wind gust as represented in ECMWF, but from my understanding and experience you would typically experience wind gusts at the open sea as an increase of wind strength in about 20-30%, and in some rare cases up to 40-50%. But usually around 20-30%.

Maybe you guys could help me out and bring som understanding to the question.

Thank you.

idefix37

@Uskar
At sea wind gusts can usually reach 40% of the average wind. But in some convective conditions they can be much stronger.

ECMWF provides this parameter for wind gusts:
https://codes.ecmwf.int/grib/param-db/?id=228029

ECMWF now includes special gusty strengths due to deep convective conditions:
https://www.ecmwf.int/sites/default/files/elibrary/2009/17487-parametrization-convective-gusts.pdf

Could you give examples with screenshot of these high gusts predictions (2x average wind).

Uskar

Thanks for the info. Here is a screenshot.

Not 2x average but ~80% increase at 15.00. Still quite high?
But maybe this is due to some sort of "terrain effect" in the model?

Here is a screenshot from a measurement site very close to the location above:

windy 2.png

idefix37

@Uskar

I have checked wind and gusts speed in knots. In % it is about the same (fortunately :)).

w/customized color scale

As you said the fact that there are many islands in this area may explain the higher speed of gusts. For wind speed the model takes into account the roughness of the terrain.
https://codes.ecmwf.int/grib/param-db/?id=244
So the wind is probably a bit reduced while the wind gusts that models deduct from wind at 950 hPa are not disturbed in the same way by the terrain roughness.
In addition near the coast with a very indented sea shore, some points of the grid are over the land and some over the sea. So the interpolated point is averaged between them.

I have plotted virtual grid points at 9 km distance to give an idea of the terrain effect:

If you choose a point about 5NM to the West, you see that the discrepancy between wind and wind gusts is reduced.

Uskar

@idefix37

Thanks a lot for the input and clarifications 🙂. This is really interesting. I have no formal training in meteorology, but work in somewhat related fields (engineering, measurements and modelling/simulations). I guess that challenges with grids/meshes and how to resolve models might have some similarities. Anyhow, very interesting, and thanks again 🙂👍🙏.

Gkikas LGPZ

@Uskar
Additionally, please read https://community.windy.com/topic/6526/how-to-forecast-gust-wind/7