@KrisLoves-Cats We had to postpone the launch a bit. Right now, the plan is to launch it on Jul 21. Here's a sneak peek. We will also launch the forecast predictability (% in the header).

@KrisLoves-Cats We had to postpone the launch a bit. Right now, the plan is to launch it on Jul 21. Here's a sneak peek. We will also launch the forecast predictability (% in the header).

Hi @whaletales
could you try again now? This seems like a rare problem with our payments provider. If it still occurs, we'd like to know.
I've issued a refund for your payment, since it was only 3 days since the transaction. Sorry, it didn't work out as intended.
Can I ask what your expectations were and how we did not meet them? We value the feedback, thanks!
It seems we are back on track! Thanks for reporting, guys.
Thanks, everybody, for reporting. Special thanks to Travis for providing the details on this! We are on the case.
Hi @gaosi231
Windy API is WYSIWIG. There are no customizations on our side possible. Check the meteoblue API documentation on this. It might be more helpful for your case (hope so).
https://docs.meteoblue.com/en/weather-apis/maps-api/radar-and-satellite-tile-api
Hi @yongsua
It's not possible. Although it would be interesting to see your comparison between 00z/12z ref times and other ref times of the given day, I'd doubt it could be more accurate every time. Of course, there can be scenarios and locations where it happens more frequently, but in the big numbers, this is very improbable.
Out of curiosity, which weather model and location are you referring to?
That's the last time the model data were complete and ok.
UPDATE: The Met Office has changed the way data is uploaded. That's the cause for not getting the data right, and probably the reason for incomplete and late data as well.
We are working on the fix on our side to get at least the late data.
@mac_harris
Met Office sends the model data late again.
You can check the status here as well:
https://datahub.metoffice.gov.uk/support/model-run-status/atmospheric#mo-uk
Re-marking as unsolved again.
Thanks for reporting, we'll take a look at it.
@santeramo