Posts made by kkantoutsis

@Franckielestore

Contact them: support@windy.com

@idefix37

So this division-by-3 issue in 1-h forecast, looks like it's kinda front-end? Will it be solved?

@Franckielestore

Yes it is true, I confirm it, but it only happens in the precipitation of Meteoblue. I think nowhere else.

In any case, windy premium is not a fraud. Also, you don't have to worry, you can take your money back the first 14 days.

@Gkikas-LGPZ

I want in the android app to select a location and then to click compare forecasts. Then I want to see all the models with the maximum days forecast each.

Imagine that 100 mm precipitation is coming after 10-12 days. I want to compare ECMWF with GFS in the 10th day and at the same time to see what ECMWF thinks about the 11th-12th days.

Imagine two football players having scored 10 and 16 goals during a season. If you want to compare them, is it meaningful to ignore the last 6 goals of the second player and to just compare the first 10 of both?

@Suty

Hi. Would it be possible to keep both the ECMWF ones separately? For example what if I want to compare the HRES and AIFS for the tomorrow forecast?

In any case, the 16-day-forecast of AIFS should be in the "compare forecasts" mode, and marketingly speaking in the premium ads, the "10-day forecast outlook" should change to "16-day forecast outlook".

But I stillI don't get it. Is there any benefit, even inside France, to use the 2.5 instead of the 1.3? Why do some prefer the 2.5?

@Markusbouman

I don't understand you and the others. For example, I search for Schiermonnikoog, and I can see the Arome 1.3 km version being active.

What do you mean? Maybe you are talking about its duration, being just 1.5 days?

At the moment, if I choose the ECMWF model alone, its 16-day-forecast is visible, but if I choose the "compare models" mode, both GFS and ECMWF stop at 10-11 days.

Will the 16-day-forecast of ECMWF be visible on the "compare models" mode?

Actually, it is one of the most interesting models. An AI version of ECMWF.

Is there a plan in windy.com to add it?

https://charts.ecmwf.int/products/pangu_medium-t-z?base_time=202308040000&level=1000&projection=opencharts_europe&valid_time=202308040600

@Richard-Burrows

Well, I mean, for any app on Android you can use any version you want: Just you google for the apk of the specific version.

Actually, on my phone more than half of the apps run in the version that I want. Just, when I have free time I "update" some of them for checking the latest versions.

Are there any plans of adding some new models for Europe - Mediterranean (I live in Greece)?

At the moment there are 4:

GFS 22km
ECMWF 9km
METEOBLUE
ICON-EU 7km

@idefix37

Yes, you're right. I had wrongly supposed that the logical operator is "or" but it is "and". Thank you.

I confirm that they don't work properly.

The Windy Radar needs reinstallation from scratch (after the last update I deleted it from my home page and then I installed it again). Also the play button needs to be pressed 2-3 times, 2-3 seconds to start working.

The live camera needs much time to start working every morning after reconnecting to internet.

I don't get any notifications / alerts for my favorites, neither on my phone nor on my email (I have checked both inbox and spam).

Any advice?

What is your opinion to have a new model as the average of all the other models? or with custom weights / coefficients, free to anyone to choose them, for example 0.2 for GFS, 0.3 for ECMWF, 0.4 for METEOBLUE and 0.1 for ICON-EU?

I am sure it has already been asked and the windy team has already analysed it. So, what stage are we at?

This is a feature in the "New" section, in the version 39: "What's new, version 39".

Can someone explain to me what exactly it is, since I thought that it was available in 38 version as well?

@idefix37

So, I mean, can I ask somehow the programmers team to put this feature in the coming updates? how? Are there somewhere suggestions for the next updates?

@idefix37 thank you for the answer.

Two more questions:

1. On the sounding forecast there is the ICON model which does not exist in compare forecasts, while the METEOBLUE exists in compare forecasts but not in sounding. Why does it happen?

2. Wouldn't it be more practical to show the reference altitude (model elevation) in the compare forecasts mode? 5 altitutes: the average geographical altitude of the region and each model's reference elevation of that region.

I suppose you compare the 3h mode with the 1h? If so, it is expected that the 3h is more accurate than the 1h. The smaller is the time period, the less accurate is the forecast and that derives from some cosmological aspects.

Or do you mean something different?

Imagine that in the next 3 days you plan to hike from 500 meters altitude to 3000 meters in a mountain range. If you want to see the forecast to all these locations, it is problematic in the current app (Android).

For example, these locations: 1) Olympos National Park or 2) Zonaria Central Macedonia

The first one says altitude 1834 meters and the second one 2643 meters. You can confirm these altitudes by scrolling on the left, lets say, the Basic ECMWF and finally it says some things like Sunrise, Sunset, Dusk ... and the Elevation...

The point is that the models give fully different forecasts, something that does not happen for small cities with a stable altitude.

For example if you get into Zonaria Central Macedonia, compare forecasts, the GFS says that the maximum temperature in the next 2 days will be 32-32 C, while the METEOBLUE says 18-19 C. This isn't because "they disagree", but because they are referring to different altitudes.

So do you plan to fix somehow this issue? Maybe a linear regression mapping on front-end could be enough?

Finally, if I plan to hike the next days to 3 or 4 thousand meters altitude, what is the optimal way to use the current Android app for forecasting?