Accuracy of Clear Outside

[TopperHenly]

Apologies in advance if this has already been asked.

I have used the Clear Outside site and app and I've found it pretty accurate for clear skies on the days when I had went out stargazing in the past.

When I get the monopod I have ordered for use with my binoculars, as I'm now working from home due to the pandemic and don't need to get up as early as I did, I will be able to do out stargazing during workdays even if it means I can only do so in the early hours of the morning. To do this, I intend to use Clear Outside to plan for future dates where I can do this.  I was wondering if for instance if CO showed that there was a clear sky in a week or even of fortnights time, would that forecast be correct on that date?

[Waddensky]

In all weather models, the accuracy of the forecast decreases over time. In my experience, models can predict a clear night a few days ahead, maybe a week depending on the local circumstances, but not much further.

"A seven-day forecast can accurately predict the weather about 80 percent of the time and a five-day forecast can accurately predict the weather approximately 90 percent of the time. However, a 10-day—or longer—forecast is only right about half the time. "

Source

[Pixies]

I find it only reliable for a day or so in advance. You can use the more distant forecasts to get a rough idea of the likelihood of clearer weather, but not much more than that.

Let's face it - how many weather forecasts are accurate more than a few days ahead? Certainly not accurately enough to be able to specify exact cloud cover

[johninderby]

I’ve noticed that the accuracy of CO varies across the country. In some areas it’s pretty accurate but for my area the BBC is far  more accurate than CO. 

Wonder how many others find one or the other better in their area. 🤔

[NGC 1502]

I use clear outside regularly because I’m one of the keyholders for my local club’s dark site. (We’re currently locked down of course).

We always wait until the afternoon of a planned visit to confirm or cancel. Clear outside is probably one of the better forecasts because it tries to predict different cloud layers. However I’d say it’s accurate about 75% of the time, so I use it in conjunction with other forecasts.

Forecasts for the general public often ignore high thin cloud, a hazy sky allows the brightest stars to be visible but that’s not much use for astronomy.

Clear outside is useful, but use with caution.

[Jamgood]

1 hour ago, johninderby said:

I’ve noticed that the accuracy of CO varies across the country. In some areas it’s pretty accurate but for my area the BBC is far  more accurate than CO. 

Wonder how many others find one or the other better in their area. 🤔

Same here. CO isn't the best weather app for Derby. I use online Radar Weather Maps instead. Give me a better idea of what is coming in the days ahead but even that is nowhere near perfect. Sticking my head out and having a good luck around is the tried and tested way though.

[PeterCPC]

I find CO almost useless. Accuracy here is very poor.

[valleyman]

We don't need Weather Forecast here in Wales. It just Rains. 🤣

[Guest]

10 minutes ago, valleyman said:

We don't need Weather Forecast here in Wales. It just Rains. 🤣

Ditto in SW Scotland! 😂

[johninderby]

.

 

Edited November 13, 2020 by johninderby

[Adam1234]

I find CO is more accurate than BBC weather most of the time,  but find neither are very accurate more than a few days in advance. 

Sometimes neither (particularly BBC) are very accurate even on the day. The other day BBC said the sky was clear - I looked outside but the whole sky was covered with cloud. Days in advance I can understand getting slightly wrong as there are many variables in the weather, but the current weather, I cannot understand how they can get that wrong! CO however was giving more realistic data.

[rnobleeddy]

It's best to stop thinking of long term weather as something that can be predicted. It's a chaotic system and the best that forecasters can do is suggest the likelihood of various weather in the future.

So maybe in 80% of their simulations it's clear, and so the app reports that it'll be clear. If it's cloudy, that doesn't mean they're wrong, it's just means that the less likely reality has come true. 

 

Short term forecasts should be a bit easier. I've actually found CO is massively pessimistic, and to be honest, I've assumed it must have a bug or bad data because it differs from other forecasts so dramatically.  The met office forecast seems better, although it'd be nice if they split the cloud in to types.

 

[banjaxed]

I remember years ago an elderly gentleman I knew told me “the only accurate way to know what the weather will do is Look Out The Window.

[Zermelo]

Forecasts have apparently improved a great deal since the introduction of multi-million pound computers, but I still find them frustrating and unreliable, so I tend to set up speculatively (which I'm happy to do as I have simple kit and a garden site).

If you choose to "turn on experimental features" in CO, you get some extra forecast lines that use alternative data sources, and sometimes they disagree with the main forecast. The cloud cover predictions in particular, while obviously critical for us, can be a bit iffy.

In the few months that I've been following it, I have found that it copes with predicting UK weather very much like the others (i.e. not great). You really need to combine a specific forecast with an understanding of the overall prevailing conditions. If it's being made in the context of a very stable system (say, a big high pressure area that's not going anywhere fast) then I would trust the forecasts a few days out (but never more than a week). But often in the UK we have the situation of a succession of systems moving in rapidly from the west or southwest. You might see a forecast of a decent night in three days' time, but in practice it arrives in two, or four, or not at all. Many times I've checked a forecast on the afternoon of a prospective evening session, and the actuality has been wildly different. And of course cloud cover is only part of the story, you may get a clear night as per forecast, but have to contend with poor seeing, transparency, dew, ...

If you want, or need, to be informed by the forecasts then one option is to consult several and rationalize them yourself. Others include Meteoblue , Accuweather, Weather Outlook and Windy.com. 

 

[banjaxed]

Waiting at traffic lights outside a local radio station location it amused me when the weather  forecaster said “dry with sunny intervals” when the rain was bouncing off the pavement. 🙄

[Hallingskies]

For what you pay for it, I think CO is an excellent service.  It’s usually pretty good on the day as far as I am concerned

[old_eyes]

I find CO reasonably accurate on the day, although often out by an hour or so (clear period arrives early or late). The annoying thing is when it holds out the prospect of a long clear period 5-days out, and then quietly finds more and more cloud as we get closer until there is no clear sky except for a 1-hour slot at 3 o'clock in the morning! But that's forecasting for you. Forecasting is difficult - especially of the future!

[Stargazer33]

CO here is next to useless I'm afraid. I usually look on the Met Office site, at the recent observations. Either 'Satellite' or 'IR Satellite' and run the sequence through at x2 speed and see if any clouds or clear skies are heading my way. Not always successful, but it's more accurate than CO. We do have a very strange micro-climate around here though. Nobody really gets it right unless we are in a prolonged spell of weather!

Edited November 13, 2020 by Stargazer33

[Jm1973]

It's hit and miss here in Bristol.

I've only recently started using it, and for a couple of weeks I couldn't find fault with it.  But a then one time I'd waited all week for a clear night, as it had been predicting Green from 6pm to 6am all week. On the day the prediction still stood so I set up about 8pm. By 9pm it was completely cloudy and remained so until 1am when I went to bed, and was still cloudy in the morning.

Just yesterday, a friend asked me how accurate it was. It had been showing red for my area all day and it predicted red for the rest of the evening, yet outside my window there was a bright blue sky and no clouds and it had been like that for hours.

I can understand it being wrong days in advance, but something is very wrong with the prediction model when it is sunny all day and ClearOutside still shows red. Like it can't even tell what's happening an hour in advance. Or even at the present time.

I would definitely take it with a very large pinch of salt.

Edited November 13, 2020 by Jm1973

[rnobleeddy]

1 hour ago, Jm1973 said:

 

I can understand it being wrong days in advance, but something is very wrong with the prediction model when it is sunny all day and ClearOutside still shows red. Like it can't even tell what's happening an hour in advance. Or even at the present time.

 

Yeah, this is the behavior I observe too. Last week CO was red whilst I imaging in clear skies. The met office reported clear skies. 

Anecdotally, CO got a lot worse for me around the time that the Dark Skies app stopped working for me on Android, but I doubt they'd hobble their own API.

[Ouroboros]

I use BBC and CO for looking ahead by hours or a day or two. 
On the night itself (or late afternoon) I use the MET Office rain radar and satellite pictures to see whether it’s going to be clear where I am.  You can track the weather over the last few hours, see which way it’s going and make a reasonable forecast based on that. I find that’s the most reliable information to base a decision on whether it’s worth setting up or not.  
 

PS sticking your head out of the window is not totally reliable. It can be crystal clear, wall to wall skies now. In half an hour it could be pelting it down. 

[Jm1973]

9 minutes ago, Ouroboros said:

PS sticking your head out of the window is not totally reliable. It can be crystal clear, wall to wall skies now. In half an hour it could be pelting it down. 

The night I was on about: when I set up at 8pm it was totally clear, but by the time I'd gone inside and picked a target and slewed to it, the clouds had come out of nowhere.

Must have been 10 minutes tops to go from not a cloud in sight, to not a star in sight.

[johninderby]

Or if you are realy desperate. 😁😁😁

Googling came upwith this.

https://www.wattpad.com/256438176-book-of-spells-sunny-skies-spell

[Jm1973]

22 minutes ago, johninderby said:

Or if you are realy desperate. 😁😁😁

Googling came upwith this.

https://www.wattpad.com/256438176-book-of-spells-sunny-skies-spell

Can't hurt to try.

[Ouroboros]

26 minutes ago, Jm1973 said:

The night I was on about: when I set up at 8pm it was totally clear, but by the time I'd gone inside and picked a target and slewed to it, the clouds had come out of nowhere.

Must have been 10 minutes tops to go from not a cloud in sight, to not a star in sight.

Yes, that is the advantage of satellite pictures. Mind you they’re not completely reliable. At night they’re just infrared which I think work on temperature difference. So it can look clear on the picture but it’s cloudy outside. Also cloud and rain can form out of (seemingly) nowhere after showing no signs on the satellite and radar pics. 

Edited November 13, 2020 by Ouroboros