For as long as I can remember, I have noticed that the color of the wind gust forecast changes to match the color of the steady wind forecast when switching from basic view to meteogram view as seen in the attached screenshots below. It also has the same issue in wind view. I just kept forgetting to post about it until now.
Sorry for replying to such an old article, but are these reference times? I ask this because my limited understanding is that, for example, the ECMWF model is updated using these reference times, but the actual time when the model update is publicly available for viewing within the Windy app is much later (by like 8 to 9 hours after the reference time according to a pinned FAQ article I read on here, but even that article was confusing since it states this and then give the example of 00:00 Z being the reference time for ECMWF and then gives 07:15 Z as the update time, which is less than the 8-9 hours stated in the article's previous paragraph).
To confuse matters further, the "update interval" listed on the app for the ECMWF model states "6-7 hours," which less than the update time given in the example given in the FAQ article. Also, the reference time currently listed at 2:18 PM EDT for the ECMWF model says 06:00:00Z (which is 2:00 AM EDT), but it says that it was last updated at around 10:44 AM EDT (14:44 Z) with the next update time at 3:44 PM EDT (19:44 Z). Why would it be showing a forecast with a 06:00:00Z reference time when it was last updated at 14:44 Z? I suppose that falls within the 8-9 hour update time range listed in the FAQ article though. If that's the case, then why is the next update time listed as being 7 hours and 44 minutes after the 12:00 Z reference time (which is when the next model update runs for ECMWF)? That falls neither within the 6-7 hour "update interval" listed on the app or within the 8-9 hour range listed on the FAQ article.
Additionally, the HRRR model is said to update hourly, but the updates seem to only occur every two hours at about about 33 minutes past the hour. The reference time currently listed for the HRRR model at 2:18 PM EDT is 16:00:00Z, which is 2:00 PM EDT. However, the app states that the last update occurred at 1:33 PM EDT with the next update occurring at 3:33 PM EDT. How can the reference time be later than the update time? I'm super confused by all of this.
I understand many different factors affect the processing and download times of each update, but is there an approximate average of how long it takes between the reference time and when the Windy app shows the latest update for each model? Can anyone help shed some light on this for me? Thanks!
@Korina, thanks, but that is the article I was referring to in my comment.
I have noticed two errors with the NAM model. The first is super simple to fix and that's changing the label from "NAM 3km" to "NAM 5km". The second error is that the line marking where the NAM forecast ends and where the GFS forecast begins is not correct. I consistently appears between 6 AM and 9 AM when it should really be appearing between 12 PM and 3 PM since it is a 48 hour forecast model, not a 42 hour model. In fact, the data shown for these hours is actually from the NAM forecast up until 3 PM, then you can see the GFS model data is shown from 3 PM and beyond. The error is simply where the line is being drawn to mark where one forecast ends and the other begins. 
I have noticed that the NAM weather forecast model seems to accurately display on the map overlays while showing incorrect information on the forecast section. As you can see in the attached screenshots, the speed and direction of the wind and gusts, the cloud cover, and the temperature shown on the forecast do not match up with what is shown on the corresponding map overlays. The overlays seem to be quite accurate, but the displayed forecast data at the bottom of the screen is wildly incorrect.
Thank you for fixing the alignment issue. It appears that the icon forecast model is now the only model that does not align the times properly. Additionally, the bug where the selected time jumps ahead to the adjacent time block to the right when switching from any forecast model to any other forecast model still exists. I will post that bug in a separate thread though.
I just noticed that the forecast timeline now shows "0AM" instead of 12AM like it used to say. This is persistent in all models and views. I don't think a screenshot is really necessary, but I can post one if needed.
@Korina, thanks, but that is the article I was referring to in my comment.
@idefix37 okay, so why does the Windy app show the 12:00Z run time for the ECMWF model as being updated at 19:15Z today? I'm not complaining or claiming there is anything wrong. I'm just trying to better understand the data I'm looking at. Thanks for your help!
@pavelneuman Sorry for replying to such an old article, but I'm quite confused. My limited understanding is that each model begins its update process at the listed reference times, but the time when the model update is actually publicly available for viewing within the Windy app is much later (by 8-9 hours after the reference according to this article). But this is confusing since it states this and then gives lists 07:15 Z as the update time for the ECMWF model's 00:00 Z reference time, which is less than the 8-9 hours stated in this very same article). To confuse matters further, the "update interval" listed on the app for the ECMWF model states "6-7 hours," which is less than the "update time" stated in this article.
Also, the reference time currently listed at 2:18 PM EDT for the ECMWF model says 06:00:00Z (which is 2:00 AM EDT), but it says that it was last updated at around 10:44 AM EDT (14:44 Z) with the next update time at 3:44 PM EDT (19:44 Z). Why would it be showing a forecast with a 06:00:00Z reference time when it was last updated at 14:44 Z? I suppose that falls within the 8-9 hour update time range listed in the FAQ article though. If that's the case, then why is the next update time listed as being 7 hours and 44 minutes after the 12:00 Z reference time (which is when the next model update runs for ECMWF)? That falls neither within the 6-7 hour "update interval" listed on the app or within the 8-9 hour range listed in this article.
Additionally, the HRRR model is said to update hourly, but the updates seem to only occur every two hours at about about 33 minutes past the hour. The reference time currently listed for the HRRR model at 2:18 PM EDT is 16:00:00Z, which is 2:00 PM EDT. However, the app states that the last update occurred at 1:33 PM EDT with the next update occurring at 3:33 PM EDT. How can the reference time be later than the update time? I'm super confused by all of this.
I understand many different factors affect the processing and download times of each update, but is there an approximate average of how long it takes between the reference time and when the Windy app shows the latest update for each model? Can anyone help shed some light on this for me? Thanks!
Sorry for replying to such an old article, but are these reference times? I ask this because my limited understanding is that, for example, the ECMWF model is updated using these reference times, but the actual time when the model update is publicly available for viewing within the Windy app is much later (by like 8 to 9 hours after the reference time according to a pinned FAQ article I read on here, but even that article was confusing since it states this and then give the example of 00:00 Z being the reference time for ECMWF and then gives 07:15 Z as the update time, which is less than the 8-9 hours stated in the article's previous paragraph).
To confuse matters further, the "update interval" listed on the app for the ECMWF model states "6-7 hours," which less than the update time given in the example given in the FAQ article. Also, the reference time currently listed at 2:18 PM EDT for the ECMWF model says 06:00:00Z (which is 2:00 AM EDT), but it says that it was last updated at around 10:44 AM EDT (14:44 Z) with the next update time at 3:44 PM EDT (19:44 Z). Why would it be showing a forecast with a 06:00:00Z reference time when it was last updated at 14:44 Z? I suppose that falls within the 8-9 hour update time range listed in the FAQ article though. If that's the case, then why is the next update time listed as being 7 hours and 44 minutes after the 12:00 Z reference time (which is when the next model update runs for ECMWF)? That falls neither within the 6-7 hour "update interval" listed on the app or within the 8-9 hour range listed on the FAQ article.
Additionally, the HRRR model is said to update hourly, but the updates seem to only occur every two hours at about about 33 minutes past the hour. The reference time currently listed for the HRRR model at 2:18 PM EDT is 16:00:00Z, which is 2:00 PM EDT. However, the app states that the last update occurred at 1:33 PM EDT with the next update occurring at 3:33 PM EDT. How can the reference time be later than the update time? I'm super confused by all of this.
I understand many different factors affect the processing and download times of each update, but is there an approximate average of how long it takes between the reference time and when the Windy app shows the latest update for each model? Can anyone help shed some light on this for me? Thanks!
Thank you for fixing the alignment issue. It appears that the icon forecast model is now the only model that does not align the times properly. Additionally, the bug where the selected time jumps ahead to the adjacent time block to the right when switching from any forecast model to any other forecast model still exists. I will post that bug in a separate thread though.
For a long while now, I have noticed that the various forecast models do not stay aligned with one another when switching between model views as seen below. It seems that the ECMWF and MBLUE models align and the GFS, ICON, and NAM models align with one another.
Something worth noting (but may be it's own separate bug) is that 9 AM was originally selected manually, but then it changed to 10 AM being selected for all of the models when I switch models. This change does not happen if 9 AM was automatically selected as the default time when I first open the forecast.
!
For as long as I can remember, I have noticed that the color of the wind gust forecast changes to match the color of the steady wind forecast when switching from basic view to meteogram view as seen in the attached screenshots below. It also has the same issue in wind view. I just kept forgetting to post about it until now.
I just noticed that the forecast timeline now shows "0AM" instead of 12AM like it used to say. This is persistent in all models and views. I don't think a screenshot is really necessary, but I can post one if needed.
I just noticed another issue related to the automatic timeline scrolling feature. It seems that the ECMWF forecast stops automatically selecting the time as I scroll once I reach 48 hours. As seen in the following screenshot.
@ivo awesome! Thank you for your great work!