Another vote against the bubbles... It's much more difficult to find the correct text label for a layer when it's buried between nonsensical colorful tiles. Not only is it multiple clicks to change layers (where it was previously efficient), it's harder to navigate. MAJOR UX downgrade.

Another vote for keeping the cloud/presssure/precip view of the meteogram in the airgram. If it stays, it's a one stop shop for flying information. Otherwise flipping through meteo/airgram and switching models gets tedious.

@idefix37 Ah, I see. A bit deeper dive on this forum led me to a cool description you gave back in 2023 of the Meteoblue AI model. In case someone else is interested and following along:

@idefix37 said in Meteoblue resolution:

• I'm curious to know what the resolution of Meteoblue is that Windy uses?

The Meteoblue model, called Meteoblue AI by this Swiss company, has no resolution because it is not a gridded model.
In short this model is “averaged” from a dozen of models with different resolutions from 2 to 40 km. Then the result is corrected according to local discrepancies usually observed. So it is only available at point forecast.

• Will Windy be able to visual display Meteoblue forecasts like they do with ECMWF, GFS, etc …?

No, for the same reason said above.
It is the reason why the weather maps in the Meteoblue website are plotted from ICON, NEMS, GFS, NAM, HRRR, IFS 40km, UKMO, MF, GEM, etc…. but NOT from their Meteoblue AI model which is only shown in local meteograms.

Now my understanding is that the predictability percentage is specifically from the meteoblue model, but the value is an aggregate of way more models. Interesting! I look forward to playing with this new information!

It seems like there's pushback every time you guys get rid of the clouds in airgram. With the cloud feature included, it is almost a one stop shop for flying. It has been removed (and reintegrated) 3 times that I can recall. Why is this a constant issue? What are we, as the users, missing from the discussion to keep removing it?

That said, the added elevation wind detail is fantastic and I guess a worthy trade in exchange for the cloud view.

@Ondřej-Šutera Add another user vote for bringing back the clouds on airgram! It's really frustrating to jump between the screens now, especially when forecasting a couple days out because the default after changing views is current time. This means I also have to scroll back to the correct time on the x-axis as well.

Any updates on this?

@idefix37 I didn't realize you could drag the picker around, so thanks for the tip.

I do agree with @Comraddm that a data readout would be more convenient. Maybe in a corner of the screen and out of the way, similar to how some map applications will show lat, long, and elevation of the pointer. I would find it particularly useful for wind and CAPE.

Another vote for this. ECMWF is consistently not well converged for North America.

@Canne It looks like your second test shows the behavior I'm talking about. To me, it doesn't seem like predictable behavior to be using the ECMWF on the map, but then click to get a pointcast and be put back on the GFS when the 'always open as' is toggled off.
I have described how to recreate the issue and I have described what I suspect is happening using high level pseudocode language. Is there something else I can provide?

@Canne @suty Sorry for the late respond guys. In the 'settings for point fcst' menu, I have toggled the 'always open as' option for the model selection off. I'm not sure it is relevent, but the other 'always open' for time is set to 1h and 7 days.

In my profile user settings, I have the 'startup layer and location' set to 'use last *'.

I'm not sure exactly what's happening on the backend, but my take is that the model variable doesn't seem to be properly passed from the map to the point forecast when RE-instancing a point forecast. It may be keeping the last point forecast instance active?

@Suty Try the example I gave. I suspect it's a bug.

Messing with it a little more, it doesn't always choose the ECMWF as the default, it chooses the last model that was used for pulling a point forecast.

An example: I view a point forecast with the GFS, then close the point forecast and change model (in map mode) to view the ECMWF map. If I click a location to pull another point forecast, it goes back to a GFS point forecast instead of the expected ECMWF.

I switch between models quite often as I'm sure many people do. If I click the map for a point forecast when viewing the map using a model that is different from my default (in 'settings for point fcst' menu), it always goes back to the default model. This works as I expect.

I purposefully turned that feature off in the settings hoping that it would keep the map forecast and point forecast model consistent. Unfortunately, the behavior stays the same, except the default seems to be set to ECMWF. Is there a way to get the behavior I desire?

@idefix37 Ah, I see. A bit deeper dive on this forum led me to a cool description you gave back in 2023 of the Meteoblue AI model. In case someone else is interested and following along:

@idefix37 said in Meteoblue resolution:

• I'm curious to know what the resolution of Meteoblue is that Windy uses?

The Meteoblue model, called Meteoblue AI by this Swiss company, has no resolution because it is not a gridded model.
In short this model is “averaged” from a dozen of models with different resolutions from 2 to 40 km. Then the result is corrected according to local discrepancies usually observed. So it is only available at point forecast.

• Will Windy be able to visual display Meteoblue forecasts like they do with ECMWF, GFS, etc …?

No, for the same reason said above.
It is the reason why the weather maps in the Meteoblue website are plotted from ICON, NEMS, GFS, NAM, HRRR, IFS 40km, UKMO, MF, GEM, etc…. but NOT from their Meteoblue AI model which is only shown in local meteograms.

Now my understanding is that the predictability percentage is specifically from the meteoblue model, but the value is an aggregate of way more models. Interesting! I look forward to playing with this new information!

@idefix37 said in What does this percentage sign stand for?:

But in Windy it is based on the predictability of the set of models computed by Meteoblue in their multimodel.

I think I understand. Hopefully it is not the predictability of the Meteoblue multimodel in particular, but uses the same calculation method to aggregate all available models for a location. If so, that's an exciting new feature! (I'm in North America and meteoblue doesn't do a great job near me.)

It seems like there's pushback every time you guys get rid of the clouds in airgram. With the cloud feature included, it is almost a one stop shop for flying. It has been removed (and reintegrated) 3 times that I can recall. Why is this a constant issue? What are we, as the users, missing from the discussion to keep removing it?

That said, the added elevation wind detail is fantastic and I guess a worthy trade in exchange for the cloud view.

@suty I correctly guessed the meaning of predictability, but counterintuitively the value doesn't change when you change model. Is this a bug or is the predictability value an aggregate of all the models?

Another vote for this. ECMWF is consistently not well converged for North America.