Hi there, we have contacted Starlink and according to their employees, they have engaged in talks with Google and are working on resolving this blocking issue. It should be fixed within a few days.

Trust me, we are taking this seriously, I almost fell out of the window trying to set up Starlink in our office for testing, but, unfortunately, we are on a different ground station than you guys, so, to our dismay, Windy worked without any problems on it.

Due to on our partner's discontinuation of data provisioning we are receiving no radar data for Mainland China and some parts of India. We are working on fixing this issue, but we currently have no timeframe for its resolution.

We are very sorry for any inconvenience this is causing you.

Hi, as for ECMWF, the 1800Z forecast was imported pretty much on time. The 0000Z forecast is expected to be available at the standard time of about 8:00 UTC. As for the Australian radar, we will take a look, thanks for bringing it to our attention!

Instead of map layers, some users of our Maps Forecast API might see fallback tiles with "No Data" text on them. We have identified the source of the issue and are working on fixing it.

We are on the path to fetching Chinese radar data from another source. It should be done within at most a few days, hopefully it could be in the next 24 hours.

@S-G-17 Hi, yes, unfortunately, our maps are based on quite old OpenStreetMaps labels. We have an ongoing update project, we hope that within a few months it should be done and the labels should reflect current OpenStreetMap data.

We have switched our forecast serving to the functioning regions.

This should mitigate the issue for the majority of users, however, you may face increased loading times for the forecasts, as they are being served from USA and Asia, increasing the connection latencies.

If you read the discussions here, radar "forecasting" is a very controversial feature. That's due to it not really being a forecast. Instead, it's just an extrapolation, where we take the current images from a radar, calculate the movement vectors of different features and show how the situation could look if the features kept moving in the same direction and at the same speed during the next hour.

However - it will not show new phenomena that could emerge and it assumes that they will not disappear. It's a feature that is really popular in some regions, where it is a good predictor of rain or thunderstorms which are coming in the next hour. However, in other regions, with different geography, it can be less useful. For example, it's good for people who live in a single spot and know how storms tend to behave as they approach from different directions. However, without this local knowledge, you really cannot rely on it even within a 1h timeframe. A 3 hour radar extrapolation would basically be complete fiction.

Due to an outage on our provider's side, the data for Sea/Currents and Sea/Tidal Currents layers is outdated. We are currently serving data from June 13.

We are in contact with the provider to find out the reason why there is no new data available. We currently have no timeframe for resolution.

According to the information from our provider, the earliest time at which the issue could be fixed is 18:00 UTC today.

Due to on outage on our provider's side, satellite data from the GOESEAST and GOESWEST satellites, providing coverage over North and South America. We have no timefame for solution yet.

We are sorry for any inconvenience.

Hello, this was an aftereffect of yesterday's outage, it's fixed now.

Oh yes, if you want rain or thunderstorm forecast, you can just use the rain layer to get it :) It's just that the Radar/Radar+ layers show radar data, where the preditction works very differently, as I wrote above.

Due to an internal error in our data import system, point forecasts with a 1 hour step (available to Windy Premium users after clicking on a location in the map) for the ECMWF model are not available for a subset of users.

The issue mostly affects users accessing Windy from Europe and Africa. It does not affect the 1 hour step on the map or the 3 hour steps in the point forecast.

We have identified the issue and it will be resolved with the next forecast, which we expect to receive within the next 3 hours.

We are sorry for the inconvenience.

@tomacat Hey there, yes, we use Google Cloud and that's quite a problem with China, so we route traffic from China through Linode. I can see in our code that it knows about this and tries to calculate the location from the original IP, so there's probably some bug, I'm gonna check it and will let you know if we manage to fix it :)

BTW, if you don't use Tor, can you access Windy from China without any problems?

@小明-姚 @雄-阿 Hello, we had a bug on our server which we use to serve data to our Chinese users. The issue should be fixed now.

Hi everyone, the issue has been fixed and everything should work as usual for you!

@远逸-张 Hi, if you are able to use Windy with VPN, please do that. Unfortunately, access to our endpoints is problematic from the mainland. We are constantly trying to work around the problem, but given our capacity limitations, it's not always possible. Furthermore, we don't have developers in China, so it is basically impossible for us to detect problems in a timely manner.

Hello there! We're really glad that you like our app. If you go to app settings (click on the menu button on the bottom right side of the app and pick Settings in the top row), you will see an "Animation of wind" option. If you switch it from "Normal" to "High," a set of sliders will appear. You can customize the size and contrast of the particles there.

We have received a current forecast for Currents and Tidal Currents now.