Posts made by dzuremar

@Gkikas-LGPZ @idefix37 Thanks for the report and info. Cca 3 weeks ago I added new Italian radar source (open data finally !). However due to a bug in my code, the Italian radar did not recognize its no-data values. As a result, Italian rectangle overwrote France values with zeroes. Should be fixed now.

@Stenar Hello, to cover Trinidad and Tobago radar, we are using data from Rainviewer (https://www.rainviewer.com/). The data have been missing there for some time, it happens from time to time. Consider approaching them directly.

It seems that was a problem with data synchronization being delayed. It works normal now, but I am not 100% sure if the problem is fixed completely. If you stumble upon this issue again, please include the exact time (and your timezone) of the occurence and also your location, so that we can track the problem better. Thanks for your patience ! :-) This definitely should not happen, USA radar data are usually delayed by 5 minutes tops when there is active precipitation around the site.

@Tomber42 It's still updated in 5 minute intervals ;-) nothing changed there. However, it takes around 10 minutes for the data to be displayed in the client (e.g. data that correspond to Germany precipitation at 14:00 are usually displayed around 14:10). I'll ask around whether there is a possibility to speed up this delay. The 20 minutes delay you experienced earlier was some kind of server hiccup. Usually it should be around 10 mins.

@Tomber42 Germany is free open data. Here is the list of countries we process ourselves: Taiwan, Czechia, Germany, Estonia, Spain, Finland, France, Italy, Netherlands, Norway, Poland, Romania, Sweden, Slovakia, United Kingdom (covers Ireland too), Canada, USA.
The rest is via Rainviewer.

@Tomber42 @idefix37 yes, that time 24 min means everywhere. Could have been some temporary internal server issues on our side. Impossible to be caused by rainviewer, plus we still get German data directly from DWD, not via Rainviewer ;-)

@ecavaleiro Thanks for feedback ! Yeah, scale is right, but I missed a small problem that caused the range mask not to update so the image was still clamped to the wrong radius. Should be fixed now ! About the update time - I can't do anything with the delay unfortunately :/ it depends on the data source (both https://www.ipma.pt/en/otempo/obs.radar/ and rainviewer.com)

@Tomber42 Of course ! I have exchanged a couple of emails with Oleksii already. He did a lot of hard work to cover such a huge area and is always very helpful. Plus his API is easy to use. Hopefully next client version will display credit for Rainviewer.com (like Blitzortung.org for example).
We still prefer our own handling of radar sources, especially raw data, but it's a huge amount of work. So Rainviewer at this moment covers the additional countries that were added in the past 2 weeks. Our original cover that has been here since last summer hasn't changed (USA + most of Europe).

@idefix37 Thanks for the report, difficulties with watermark removal. The borders are still visible in the postprocessed radar image. I have disabled this source for now, it should be fine since surrounding countries covers Belgium just fine.

@Gkikas-LGPZ @csabatatar Yes this happens often, it's a configuration error. The routine that draws those light gray circles does not know anything about the underlying radar data (over which it draws the circles). Sometime radar sites have incorrect range in the configuration (e.g. off by several kilometers), or the data can change. Plus, very probably I will be reworking the central/south american radar so probably this will be fixed to a degree :-), but still you can find these things often.

@csabatatar Oh boy, thanks for mentioning me in the comment. Now that I can see precipitation in the image, I see that the problem is not different range. The map projection is wrong. Note that the precipitation "clouds" are the same, but the scale is wrong (the island is way bigger on the Windy image). I already contacted Rainviewer.com , its creator Oleksii usually responds and fixes these minor problems quickly. It will be fixed soon. Thanks for the post !

@idefix37 Hello ! We use AROME with resolution 0.01° (1.3km). Both of the AROME resolutions have 00, 06, 12 and 18 ref time runs. There are also 03, 09, 15 and 21 ref time runs which we don't process. Advantage of AROME 0.025° (2.5km) is that it includes pressure-level layers. AROME 0.01° provides only surface-level layers, only wind and relative humidity are additionally in 20m, 50m and 100m altitude. However, AROME 0.025° also covers much smaller area - a square that covers whole France and a bit of surrounding area.

@beardedcook Hello. No, the light gray circles show our current coverage and range, nothing beyond that is available unless we find other radar source that supports larger radius.

@campas12 What is your location ? It is possible we do not have coverage there, or the responsible radar was temporarily out of order, or of course - it could be a bug.

@ubugnu Exactly, this single radar is part of our Spanish composite and has been out of order for the past few days: http://www.aemet.es/es/eltiempo/observacion/radar?w=1&p=am it is most likely temporary and hopefully will be online soon, this happens from time to time in Spanish radar source.

@idefix37 Yes, they are included in the UK radar we have from metoffice, but the images are in 15 minute intervals, and are usually delayed by another 15 minutes :/ not the best data source. This is also the reason why UK radar in our animation is usually displayed as too much behind (with the "crosshatch" pattern) in the last ~10 minute interval of the animation.

@tomber42 Actually, super interesting and very important ! We have to update our code accordingly to keep providing ICON after 01.10.2018. Thanks for the notification !

Pravdepodobne to moze byt velkostou obrazku, pokial ma viac ako 1MB tak neprejde.

@gkikas-lgpz Hello ! Yes exactly :) here I talked about it to a more detail here: https://community.windy.com/topic/5458/thunderstorms-forecast-suggestion/5 .
However, ECMWF now also supports similar forecast for lightnings density (https://www.ecmwf.int/en/about/media-centre/news/2018/ifs-upgrade-improves-extended-range-weather-forecasts) which we plan to implement as well. So soon, there will be more options !