The temperature of the Mediterranean sea is quite high this year. In some ares, temperature at the surface is around 28°C. This is of course due to the hot spring and summer.

This situation is not without consequences :

1. a medicane was formed close to the Libyan coasts and hit last Friday Greece creating a lot of damages because of the wind and rain

2. yesterday, in the South of France, because (but not only !) of the intrusion of very warm and humid air from the sea, combined of course with other factors (instability, anomaly in altitude, wind shear), it created a multi cellular thunderstorm which was stationary : each new convective cell (this is why it is a multicellular system) is created in the exact opposite way of the move of the system in altitude because of the wind shear. This creates therefore a stationary thunderstorm which is terrible in terms of rain accumulation because you can get huge quantities of water at the same place during 24h or even more.
   This thunderstorm created floating and a lot of damages in France (more than 600mm of rain in 24h)

Unfortunately, other episodes like the ones mentioned above could happen in the coming weeks,...

@Sybiljamie

If you look your data for Saturday the 13th of August, you see 0.4mm at 5.00AM. This is the total of rain accumulation between 2.00AM and 5.00AM. Then you see 0.9mm at 8.00AM. This is the total of rain between 5.00AM and 8.00AM.

If you go for Windy Premium, by splitting from 3h forecast to 1h forecast, you will see how the rain fall is spread by 1h in place of seeing the accumulation over 3h : for instance, you could have 0.2mm at 4.00AM and 0.2mm at 5.00AM as a potential example

@ginop
First of all, please start to check the posts published on this forum and discussing rainfall forecast.
Example : https://community.windy.com/topic/13586/about-rain-calculate/2?_=1602329137551
There are many others, use the advanced search tool to fine them and to read them. You will find a lot of advices.
Principles are always the same :

1. compare the different models to verify the consistency within the forecast from the different models. It is a kind of statistic approach to give you a level of confidence
2. don’t look only the meteogram which is focus on one location. You should check with the Windy rain map, the rain forecast in an area of let’s say 20km around the location. Often, models foresees no rain at the location but for instance 5mm at only 10km from the location. Thereofore risk is high to get rain at the location.
   Combining 1) and 2) should tell you the probability of getting rain or not

@Joerg78
The model from ECMWF is called IFS. This model has 4 run per day: 0z, 6z, 12z and 18z. So far, Windy gave us access to only 2 run, 0z and 12z (the main run). Now, the Premium users have access to the other 2 run, 6z and 18z

@Hubert406
Il est important de comparer les différents modèles : ICON, IFS (ECMWF), GFS,... C'est à cela que sert notamment l'outil de comparaison disponible avec le météogramme. Vous pouvez comparer les précipitations prévues par chaque modèle et vous forger une opinion sur la quantité de pluie probable.

Une autre chose à faire est de ne pas regarder uniquement le météogramme qui est focalisé juste sur votre localisation. Il faut regarder la carte des précipitations pour voir si autour de votre localisation, par exemple dans un rayon de 15km, il n'y a pas de la pluie prévue par les modèles. Si 5mm de pluie sont prévus à 5 ou 10km et rien chez vous, il faut se poser des questions quant à la probabilité que vous ayez tout de même de la pluie chez vous, le risque est grand car il suffit d’une petite déviation et vous avez les 5mm pluie aussi

@CR
No, in general with Windy, you cannot combine different maps, except pressure, geopotential, temperature and isotherm that can be display on any other maps.

@nwindy
Yes, the way the ensemble model calculates the probability is as you explain : for instance for the rain, each dot of the 21 dots can give a different level of rain (1mm, 4mm, 10mm,...). Based on that set of different levels of rain, the model can calculate several probabilistic values such as the average, variance,... This will allow you to quantify the risk of rain

Now, if the model has to adopt the same approach for many other parameters, such as pressure, temperature,... And for each parameter do the 21 calculations (do not forget that each calculation is the equivalent of a single deterministic calculation), computional limitation will be the issue if the model keeps the same grid. To overcome this problem, grid is bigger, calculation of the 21 dots for the different parameters becomes ok but accuracy will decrease.

Computational calculation is key. In average, weather forecast improves by one day every 10 years thanks to the computer (super calculator) progress

@disaster999
I would like to add 2 comments :

1. as @idefix37 and I already highlighted couple of times, Windy doesn’t make any weather forecast model, Windy doesn’t do any calculations with the models, Windy only publishes the data coming from several models and present those data in a very nice and useful interface. Therefore, Windy has no responsibility regarding the accuracy of the data. Only the institutes making the models have this responsibility. Having said that, it is also the responsibility of the user to study the different models and determine which one is the most appropriated for the location and for what he or she is expectating as accuracy. It is also key not to mix the concept of a FORECAST and CURRENT weather. Do not look at models to check the CURRENT weather but check then weather stations or satellite images.
2. In general, severe and local phenomena are still difficult to be modelized by the current models. It is the case for instance for thunderstorms. The reason behind this issue has a name : convection. Convection unfortunately do break one fundamental assumption used by all models : hydrostatism. At synoptic level, atmosphere is stable and this is what hydrostatism is telling. Convection, locally, creates an air circulation starting from the bottom, from the ground or from the see, to the altitude. Therefore, atmosphere is not stable anymore and we are locally in a non hydrostatism assumption which is not yet well modelized by the models even if they did a lot of improvement, especially Arome. While it is easy to predict that a region will have all ingredients to get potentially a thunderstorm, it is very complicated for all models to be accurate enough to tell you the exact localisation of the thunderstorm, the exact timing and the exact intensity (wind speed, rainfall, hail,...). More severe phenomena, as hurricanes, have the same issue for more or less the same type of reason.

One of the typical feature of a medicane is the hot core which is much warmer than outside. It is very interesting and instructive to follow it at different levels. At the beginning, you could notice the difference at low levels, 850 hPa for instance but day after day, you can see the differences, in the core and outside, also at much higher levels. This is because in the development process of an hurricane, the system releases latent heat, among others because of the condensation of warm and humid air at higher and higher level. Here you can see it clearly on the follwing map at 600hPa : +2°C in the core, -2 or -3°C outside. Huge difference !

Very instructive to follow the development of this medicane

@psweetapple
As a Premium user you can get 1h forecast (you have to activate It in the meteogram) and soon 4 updates a day with ECMWF Model in place of 2 (already available on the desktop)

@vicb
Sorry, I found it, it is still possible to open it via the picker

@vicb
How do you select another location than the favorite, without closing the plugin ?

@Mrfunk
This is difficult to answer because there is no real automatic answer. Indeed, the best approach is to check the cloud cover forecast from GFS, ICON and ECMWF and then to compare with the real observations to do your own estimation of what model is the most suitable one for cloud cover for your specific area.

Having said that, today, IFS from ECMWF is considered as one of the best model in Europe. My ranking for the 3 models would be :

1. IFS (ECMWF)
2. ICON
3. GFS

But as I said above, this ranking is general and not always valid for all parameters and all locations. You should do your own analysis

@gobois64
Malheureusement, ce n’est pas possible d’effacer la liste

@mirawayne
All devices you use

@alexsunny123
Obviously the quality of the data is key but even more important is the methodology and the way you are going to use them. If you plan to do a trip or a journey in a specific location, you need to get the most accurate weather forecast as possible and in that respect, using only for instance a meteogram for that specific location and coming from one model only can be disappointing.

Ideally you should start by understanding what the general situation is, what it is going to influence the wearher at the location (with the pressure map for isntance, wind map, temperature map at 850hPa,...).

In a second step, you should use different maps to see different parameters like rain, cloud,temperature, fog, etc... You should compare the same maps but coming from different models (ECMWF, GFS, ICON, etc...) to give you a certain degree of confidence about what it could happen. The big interest of the map compared to a meteogram is that you can easily see what it could happen not at the location only but also in a close area. if you see with the rain map that a lot of rain is forecasted not for from the location you go and no rainat your location, well you should pay attention because risk is high to get rain at the location !

The reason why many people are disappointed by the weather forecast is because they only use a meteogram coming from one model. Even if the data from this model look accurate, the poor approach and methodology to forecast the weather at a specific location will be much more impacting than the quality of the data.

@gkikas-lgpz
@idefix37
@nandes
This is the origin of the conflict of air massa and is going to generate a lot of snow falls in Spain where the temperatures are quite low. I think that 30-50cm of snow in Madrid on Saturday should be possible !

@Looping007
Is it TCON or ICON ?

@mcevzar
If you follow and study closely and correctly the data from ICON and IFS (ECMWF), you should be able to understand first the situation at synoptic level (IFS and ICON are the best models to understand it) and then in a second step to get good weather forecast for Eastern Europe

@idefix37
Thanks for the tip. Indeed, I was a bit annoyed by the time lapse so I increased the speed to limit it to 1h. I didn’t know I could get rid of it, which is definitely much better !