Menu Sign In Contact FAQ
Welcome to our forums

Weather Station - cloud base

In the past year our aircraft has changed its base.

The previous airfield was only 3km from my home so the weather was always going to be the same as at home. I could even pick up the ATIS from home if needed.

The new airfield is just over 40km from home. It is unmanned and has nobody to call and ask the weather, never mind an ATIS or METAR!

I’ve thought about the possibility of setting up a small weather station (if the owner is agreeable of course!) and feed it into a raspberry pi with a 4G dongle to make it available over the internet. It might save some wasted 80km round trips!

But most ‘affordable’ weather stations don’t have a sensor for cloud base. For VFR flight that’s probably the most important piece of info. (Visibility isn’t often a big issue here, and wind is much more estimable over just a 40km difference). But cloud base can be significantly different due to differing local terrain.

Is there any low cost solution to detecting cloud base?

My other thought is to add a webcam. It would be less precise but maybe would give a good enough impression. I’ve little practice of judging cloud base from a webcam! I normally find estimating cloud base to be difficult from the ground, by maybe setting up the webcam with distant known elevation items visible might work well enough.

Has anyone else done this before, and have you got a workable solution for the cloud base?

EIKH Kilrush

I believe the webcam is the best solution. It is not difficult to estimate cloud bases visually in my experience, at least not to the degree of accuracy required to make the go/nogo decision. You only need to know whether the base is: below 1000-1500ft, between 1000-1500 and 2000-2500, or greater than 2000-2500.

The problem with the sensor is that it would presumably only give you an indication of the base directly overhead it? So if you have ONE cloud over it, the indication would be meaningless. You would need the combination of cloud base sensor AND webcam to confirm visually, in any case.

I would have the following approach:
1) Webcam observation
2) Closest METAR observation (in whichever direction, to have an idea of the “current” cloudbase in the area)
3) Closest METAR in the upwind direction, and wind speed at the cloud base altitude (gives you an estimate of when that cloudbase will be overhead your airfield)
4) Windy for cloud base works pretty well I have found.
5) Always bear in mind that METARs report bases AAL (above airfield elevation), so you need to account for the elevation difference between the METAR station and your airfield. Windy reports the cloud bases as heights above the ground, but Windy’s terrain model is simplified, I don’t know how accurately it follows the terrain contour.

You can gain “practice” in estimating cloud bases visually by going on Windy, picking an airfield, check the webcam, and then look at the METAR to confirm the base.

Windy reporting a base of 2400ft in the vicinity of Belfast:

Here is a nearby webcam:

And here is the Belfast METAR:
EGAA 281450Z 22021G33KT 180V250 6000 -RA BKN018CB BKN026 13/11 Q1003

Last Edited by Alpha_Floor at 28 Mar 15:20
United Kingdom

We installed an off the shelf cheap weather station and wind sensor at our airfield, but no-one really uses it.
Cloud base most of us get a rough approximation with temp ° minus dew point° and multiplying by 400’. It’s not bad for accuracy in this area. But then I live about 10km from the airfield so the weather out of the window is pretty similar to that at the airfield except I live in a valley where at certain times of the year we will have fog or mist whereas the airfield is on top of the valley and is often above it.


gallois wrote:

Cloud base most of us get a rough approximation with temp ° minus dew point° and multiplying by 400’

I think this works better in still wind conditions and with larger temperature spreads, but the problem is that it’s not very useful in marginal cases where the cloud base could potentially be an issue. When the temperature spread is a few degrees you could have a base of 500, 1000 or 2000, too much of a margin of error to make a go/nogo decision. Of course when the spread is 10 degrees, you don’t really care if the bases are at 3000, 5000 or 7000…

The aproximation I knew was 2ºC per every 1000ft, so in the example of EGAA above:
EGAA 281450Z 22021G33KT 180V250 6000 -RA BKN018CB BKN026 13/11 Q1003
The spread is 2º, so you would estimate the base at 1000ft (or 800ft using the 400 ft/º approx.). However it’s reported at 1800ft, almost twice that. Presumably the strong gusting wind plays a role here.

Last Edited by Alpha_Floor at 28 Mar 15:31
United Kingdom

I was just giving what we would use as a rule of thumb, but of course there is nothing to beat an actual weather report. Although you don’t say how accurate that Metar was in reality.


Of course! It’s a good rule of thumb to know and use! Just sometimes not reliable enough for dispatch. You could potentially bin lots of flights that would have been fine.

gallois wrote:

Although you don’t say how accurate that Metar was in reality.

Well we can assume that the METAR is a pretty accurate and reliable indication provided by the airport meteorologist. I don’t think it gets any more reliable than that other than getting in an airplane and checking in the air in real time.

United Kingdom

If you install a webcam, save some screenshots before and after each flight. On departure and arrival, note what the cloudbase is (based on the aircraft’s altimeter). Then make a “collection” of matching cloudbase altitudes and webcam snapshots. These will be helpful for reference to better estimate the bases in the future.

It can become a very reliable, accurate, and inexpensive method of judging bases.

United Kingdom

@dublinpilot, you can try googling for diy ceilometer, there plenty of projects on the internet… Some of them appear to be quite mature!
But I agree, that a webcam or two (or three!) might be a solution or at least complimenting the ceilometer.


At my gliding club in Dunstable they had two webcams, one against the hill and one toward the horizon, they pretty tell how much visibility & oktas & cloud-base, nearby it’s Luton METAR CB which needs +300ft due to relative elevation

Windy nowcast with ECMWF is pretty accurate for Cloud-Base & Oktas the afternoons as it gets updated at 12am (and gets better the morning if you pay to get regular updates at 6am or every 6h) but it tends to underestimate visibility a lot vs METAR/TAF, at least in UK SE

Last Edited by Ibra at 28 Mar 19:52
Paris/Essex, United Kingdom

Ceilometers are quite expensive and also require some infrastructure to read them out. Vaisala are the best for this kind of application.

The “normal” celiometers are of a laser type which measure straight up. So one of them will not do that much. It is fine with a layer of cloud but useless for few/sct clouds or for convective stuff which may have huge variations in cloud base. So a single ceilometer at least needs a program which takes the measurements of say the last 10 minutes and calculates a cloudbase from that.

We have 7 of them here at ZRH airport all around and that does allow an observer to get a reasonably good determination of what is around. A couple of years ago I had to do the same in Geneva, which had only 3 of them, and it was much more demanding, even though there are quite a few mountains around to visually determine the cloud base.

A completely different story is the CL61 LIDAR ceilometer, which has much more capabilities than a normal one.

Vaisala have just about everything you need even to produce Auto Metars or to simply set up a high fidelity weather station. But they are not that cheap. You may want to inquire about their tactical met station, which is fully portable and self sustained via solar array or simply talk to them what would be required to get a reasonably good result, without going all the way and produce autometars.

IMHO, you would need Wind speed and direction, a present weather sensor (one should be enough), a pressure and temperature / dewpoint instrument, rain gauge as well as a ceilometer to get a reasonably good automated weather report. It won’t hurt to write to Vaisala’s customer service, they might have a better idea how to go about this.

LSZH, Switzerland
21 Posts
Sign in to add your message

Back to Top