Model grid
-
Different models have different grid size and the forecast is influenced by the grid size. This is relevant if the location is near a large body of water, a change of altitude or even simply near a large built area.
In that respect it would be useful to have a layer that shows the grid of the model in use to help interpret the forecast.
Or maybe the grids are already centered on the location under inquiry but I'd be surprised if it was the case.
Thanks
Nelson
-
@Nelson-Neault
No, weather model grids are not centered on any location.
For the local weather forecast and for the map, Windy interpolates the forecast values at grid points. -
@idefix37
Agreed, as expected.But is there a possibility to show the grid of the model being used so that the user can weigh the amount of faith to put into each model for the particular location under investigation?
-
@Nelson-Neault
Unfortunately, this information is not available in Windy.
We are talking about the surface level grid. But there is another question regarding the orography model of this grid. The grid of points does not exactly match the terrain.
In one place, there is a difference between the actual altitude and the model altitude. You can see these altitudes on the local sounding forecasts.
This is important in the mountains and explains some temperature inaccuracies.
So, for locations between grid points, interpolation will lead to larger deviations.
By the way, you can try METEOBLUE in point forecasts table. This model is a non-gridded model and works well in the mountains. -
So the short answer is that we cannot get a layer that shows the model grid because the information is not available to Windy. Too bad, but can't argue about that.
But the long answer is fascinating. On the graphics that you provided we can see that the model grid can be projected on a theoritical globe at sea level (there is surely a proper name for it). Ground location could also be projected on the same theoretical globe (which is probably what a road map is), so it would be possible to create a layer if the model grid nodes are fixed in space. But I now suspect that the model grid nodes are placed on axes measured in pressure units and or other physical variables. And that would make my request impractical.
-
@Nelson-Neault said in Model grid:
because the information is not available to Windy. Too bad, but can't argue about that.
No, the information is available TO Windy, but it is not shown in Windy as a layer. Some weather apps display the grid of the models they use.
The grid as you say could be displayed asl but with the topography model added. Like this example with GFS
If you know this area in the French Alps, GFS “considers” the Mont-Blanc (4807m) at only 2030m high. This is because of its 22km low resolution.
So your request would not be impractical.
