What are the best websites for predicting cloud cover?

[Martin-Devon]

I often use the Met Office website to predict a day or so in advance the cloud cover and possible observing nights I can use. Their website is: http://www.metoffice.gov.uk/

However, many times I’ve found it to be rather inaccurate. Does anybody else use or recommend alternative cloud prediction websites that may be better?

[Black Knight]

You could try www.metcheck.com. It's not perfect, but used in conjunction with other sites may be helpful. also Sat24 (SAT24.com, Satellite Weather Europe. Watch the Clouds and Sun over Europe) can be useful for predicting the direction of cloud cover.

[GazOC]

I use Sat 24 with varying degress of success. I actually got rained off twice one clear night....

[brianb]

Metcheck is usually a lot worse that the Met Office ... but you need to take all weather forecasts with a couple of lorryloads of salt. Can't beat the Mk I eyeball backed up local experience backed up by the satellite IR animations, but do remember that sat IT images don't show fog/stratus cloud at all well.

[sixela]

If you want short term predictions, nothing beats the satellite images (coupled with some weather prediction sites; just pick the one that has the closest match to the actual images at "current" time).

A pretty good model of Central Europe is run by the Swiss and available at Meteoblue; unfortunately, the North of the UK is only run at a more coarse resolution.

It does still beat a lot of the sites that simply use the public GFS data (which uses grid resolutions of 30km to 70km --good luck predicting that you'll have cloud cover but someone 15km from you won't-- and with physics that tend to completely fail to predict mist and very low cloud cover).

The Met Offices have data that's a lot better than that, but they don't have products that spit out (free) meteograms unless you subscribe to paid services.

[Nexus 6]

Yes Martin, the BBC weather forcast is pretty awfull. I have started using both Metreoblue and Sat 24, as have been mentioned, much more reliable!

Alan

[Paulb3098]

This site was recommended in Sky at Night, 7Timer! - numerical weather forecast for anywhere over the world

[sixela]

That's one of the GFS-based sites I was talking about. the model runs it uses are more than a bit coarse, but it does also predict things like seeing and relative humidity.

But for more ghee-wizz stuff and very detailed meteograms, if the model is right about things, nothing beats Meteoblue's MyMap server:

meteoblue myMapServer - interactive meteorological map visualisation

Especially fancy:

Meteograms: include relative humidity graphs --do I need to take my 12V hairdryer--, wind speeds, directions and temperatures over time and over altitude (can show bad seeing conditions indirectly), and cloud cover by altitude (shows you which type of cloud cover is expected) over time.

Astronomy/Seeing: some seeing predictions (not as good as ClearSkyClock ones, and if the models agree about the rest those from the 7Timer site are actually often better, but includes jet stream elevation wind speeds and a couple of different indices), low/med/high cloud cover.

Don't know how Sky at Night failed to pick that one up. Ah, investigative journalism is not what it used to be .

[Breakintheclouds]

I've tried 7Timer and Meteoblue extensively and have found 7Timer to be much more accurate for me. I don't know if it's just my location, but Meteoblue seems only to be right in a 'stopped clock is right twice a day' sort of way - it predicts cloud and bad seeing pretty much all the time and so, statistically, is usually right

[sixela]

Which meteoblue model are you using (i.e. are you in the zone covered by the 3km resolution run?)

There's a large difference in accuracy between the fine-grained "Central Europe" simulations and the coarser-grained "Europe" ones. And the model used has fairly crude physics, so it really requires a fine mesh to work well.

The physics have also been tweaked ad-hoc to predict low cloud and mist rather aggressively, because the Swiss were whinging that the old simulations (just like GFS) tended to predict splendid weather when that splendid weather in reality then reaches only down to 50 metres above the ground shrouded in mist . Swiss valleys have a lot of mist, so...

That change affected the coarse-grained model more (in the fine-grained simulation you will usually see pockets of "blue" which gives you a hint things might not be as dire as they look, in the coarse-grained one you'll just see a uniform mass of low cloud).

[p5272]

Hi, I find that 7timer is the most accurate.

[Manok101]

I use the weather channel but I'm not sure if they do weather for Europe it's usually very good. I live In the southeastern US