The shadow of a solar eclipse over the UK

Posted: March 21, 2015 by tchannon in Analysis, weather

For what it’s worth here is the eclipse stuff I mentioned earlier.


Figure 1, cloud abruptly cleared at midday.

There are 755 days of data from the Chilbolton Observatory, Hampshire, England, one of the worlds primary cloud research sites. Our interest is thermal radiation data. The collected is parsed from web plots, the raw data is available a month later, vast and I generally don’t process it. Either way this is one of the few sources of high resolution data in the world where there is public access.

Software here takes the data and processes to a new all-in-one thumb and a data file, gives a fair idea.

Unfortunately it was a cloudy morning, mostly unbroken low cloud, just about make out the sun. If it had been clear this would have been wonderful given the vast array of instruments.


Figure 2, detail, doesn’t add much meaning.


Figure 3, thumbnail of all the parameters. See previous Talkshop posts for meanings. A sub home page at Chilbolton Observatory, the actual home page is useless, NERC standard.

Met Office data

I also archive the Met Office hourly data a question being does anything show up?

Yes, which surprised me. Quite a few stations show a temperature drop at 10 hrs.

An often featured station is at Katesbridge Northern Ireland


Figure 4, Katesbridge, only a few parameters.

The data covers all the UK. A rough plot of all stations is done daily to a 23 page PDF for the 24 hours of the day.
PDF is here (1.3MB)

Figure 5

Figure 5

Averaging all UK stations produces a figure of an 0.6C temperature drop at 10hrs. This would be much more under clear sky and sharp at a finer time quantisation.

The 1999 total eclipse from memory didn’t notice as chilly, I thought it was a sweltering hot and humid day. A 2011 Talkshop article shows a plot from a web page elsewhere, still there, doesn’t imply a hot day, how odd.

I need to add there are problems with the Met Office data, an open support ticket goes back to early January. Hopefully this is not fatal to an understanding of the plots.

Post by Tim

  1. Your presentation is inverted to the scale. At start of eclipse, Pyranometer reading (solar irradiance) is first to drop like a rock. This “with delay” is followed by the abrupt drop “not increase” of prygomjeter reading, the piss pour estimate of atmospheric radiance. Followed by the decrease in temperature and exitance from the near surface.

  2. Richard111 says:

    Had clear skies here at Milford Haven yesterday, after a bit of early fog. Had a good view and tried to take a picture at 09:27. Didn’t realise how shaky I had become! Lousy pic but you can see the sun was about 80% covered.

  3. ren says:

    Sunspot number: 27 – New regions: 0

  4. ren says:

    Graph shows that the strong solar wind causes a drop in temperature in the stratosphere. above the polar circle.

  5. tom0mason says:

    So a short but deep variation in the solar radiation reaching an area of this planet’s surface does impact the ground level temperature (and apparently the air pressure) in that area.
    Logically then a quite small variation in solar output would cause quite a change in ground level temperature across the planet.
    Not quite the message that the IPCC put out » The TAR states that the changes in solar irradiance are not the major cause of the temperature changes in the second half of the 20th century unless those changes can induce unknown large feedbacks in the climate system. .

  6. Steve C says:

    Conditions in south Notts were perfect – we had a splendid view of the whole two hour show, with just a little atmospheric haze. I just watched it out of my kitchen window, through a home-made filter. Sorry, no photos – not my hobby, and I knew there’d be superb pics online! – but I did monitor the light level (6500 lux at 08:30 dropped to <650 and "weird light" at max, returning to a little under 18000 at 10:30, all figures "eyeball averages" allowing for haze variations) and also have a fine set of maritime weather fax images recorded from Hamburg Meteo on 7.88 MHz. These show the reception over this path becoming "a bit iffy" as the signal began to dip down towards the noise level 40 – 45 minutes before H-Hour, descending into "night-time" noise levels and complete unusability around the eclipse maximum, then returning to normal good daytime reception conditions about 45 minutes later. An interesting illustration of how quickly our ionosphere responds to changing solar conditions in real time – which is what I, as a lifelong radio enthusiast, was most interested in. Hope I shall still be around for the next one … 😀

  7. oldbrew says:

    @ Richard 111 – re shaky pics

    Try a mini-tripod, cheap at Amazon or ‘pound shops’. Makes auto-focus work better too.

  8. tallbloke says:

    And set your ‘self timer’ to two seconds to give the camera/tripod time to stop shaking after you press the shutter button.

    [Tim: Okay I just got it after puzzling over comments. 🙂 –Tim]

  9. Richard111 says:

    Gosh! Thanks. 🙂

  10. Tenuc says:

    Hi Richard. Looking forward to watching your funeral… 🙂

  11. tchannon says:

    Steve C, Now that is very interesting, I have a thing about the ionosphere as a vast untapped resource of atmospheric data. What you are revealing is an eclipse cutting a swath through morning propagation conditions. I am surprised although in hindsight, why not?

    This might yield up to a web trawl for writeups on previous eclipses.

  12. oldbrew says:

    Then there’s the Allais effect, sometimes claimed not to exist.

    ‘An attempt is made to explain the Allais effect, the anomalous behavior of pendulums during a total solar eclipse, by Poisson diffraction of gravitational waves into the lunar shadow, with the waves emitted through large mass motion in the center of the sun by a thermonuclear fusion reaction driven agnetohydrodynamic dynamo.’ [eh?]

    Conclusion: ‘…with the moon acting as a kind of a lens, increasing the intensity of the gravitational waves coming from the sun in the center of the lunar shadow by Poisson diffraction, it becomes plausible that the wave intensity might become large enough to influence the motion of a high Q-value pendulum’

    And the Wang eclipse…

    Click to access abstracts_6839.pdf

  13. tchannon says:

    “Will Janoschka says:

    Your presentation is inverted to the scale. At start of eclipse, Pyranometer reading (solar irradiance) is first to drop like a rock. This “with delay” is followed by the abrupt drop “not increase” of prygomjeter reading, the piss pour estimate of atmospheric radiance. Followed by the decrease in temperature and exitance from the near surface.”

    Not a good idea hitting on me on this one Will.

    The context is air temperature. The thermometer shows a delay.

    Yes the instruments are piss poor for high resolution work. In this case the logger output will be available next month (professional but not good enough), nevertheless sampling is about 2.4s and excess quantisation is used, offsets are not zeroed.

    These Kipp & Zonen instruments are perfectly good though for rough work and mostly they are all we have so don’t knock it. As a bonus if we were at sea I could see the instruments from where I am sitting, ie. I have a good idea of conditions there. The sun came out suddenly here too.

    I included the pyranometer net (computed here, an approximation) mostly to show the cloud cover change at midday, best cloud detector there is. (, ignore the sales bull)

    Next month I might decode some of the netcdf, see if anything useful appears. I think there is a pyrheliometer there too, whatever, the CNR4 will be nearer to what you want, still way short of anything able to show agw stuff. Such things have existed, far to complex and expensive when no-users have the faintest interest.

  14. Kelvin Vaughan says:

    Were any ground temperature measurements taken?

  15. tchannon says:

    So far as I know there is no soil temperature reading at Chilbolton but there is a ground pointing CNR4.

    This was installed (I assume) for bridging before the older instruments were removed for factory service. An article mentioning this

    The data is copious, 2MB per day compressed. There are thermocouple readings and instrument body temperatures. From that I should be able to compute a detailed _relative_ picture.

    I mention relative because gaining absolute values is very tough.

  16. tchannon says:

    Will, pondering why you mention opposite maybe this is a classic where strictly I am wrong, splitting display context, pyrgeometer ought to be plotted negative (or incoming negative). On my part this is perhaps lazy but plots can get awkward if multiple axis are needed.

  17. tchannon says:

    I’ve just decoded another netcdf, Chilbolton have a CHP1 pyrheliometer but I don’t know which tracker. (solar tracking mount) File header contains whatever the installer typed in, quite a lot of info.
    Photo of one mounted on a tracker, it’s the small tube at the bottom of the photo just above the text SOLYS 2

    One of these

    All being well that was looking directly at the sun.

  18. ren says:

    Distribution of ozone in the stratosphere is that shared the polar vortex over the North Pole. Circulation in the stratosphere over North America is completely reversed. This means that the influx Arctic air to southeast US and to Europe.

  19. Richard111 says:

    Hi Tenuc, it was a good funeral. The 21 gun salute with the replica cannon was a clever touch. 🙂

  20. Steve C says:

    @tchannon – There is certainly some information in ionospheric studies, although obviously upper-atmosphere ionisation is relevant more to space weather than to climate directly. The last big “burp” from the sun, a few weeks ago, knocked out my weather fax reception for the whole day, which was simultaneously interesting and annoying! If you’d be interested to see the eclipse faxes, email me and I’ll pass ’em on, with a bit of info on interpretation.

    A more usual way of monitoring the ionisation is done at far lower frequencies (20 – 30 kHz) than around 8 MHz: a search on SID monitoring (Sudden Ionospheric Disturbance) will find plenty of interesting stuff about it. Basically, you just set up a simple receiver tuned to the station you’ve chosen, feed the signal-strength information – usually just the rectified signal! – to a datalogger and look for ripples in the curve. Stanford has some useful information on setting up a station here, on their Space Weather Monitors page:

  21. tchannon says: March 22, 2015 at 1:26 am

    “Will, pondering why you mention opposite maybe this is a classic where strictly I am wrong, splitting display context, pyrgeometer ought to be plotted negative (or incoming negative). On my part this is perhaps lazy but plots can get awkward if multiple axis are needed.”

    Tim, My bad! I did not notice the wee dip at 9:30 that was your partial in clouds. Your expanded Chilbolton does indeed show the atmospheric wide angle radiance lagging the dip in insolation!
    Since the air temperature never rose ’twas hard to dip. 🙂

  22. tchannon says:

    Will, I wish there was better instrumentation. Anything like that is kept strictly away from us lot or is selectively used.

    When the full data is available the actual thermocouple streams will be in there, without the post sensor meddling: we get the origin of that too.

    With luck there will be 7000 data points as floats across say 2 hours of the event.

    Over 3 instruments
    2 channels of short wave all sky incoming
    2 channels of all sky long wave outgoing
    2 channels of long wave from instrument
    1 channel of long wave ground to instrument
    1 channel of short wave ground reflected to instrument
    1 channel wideband sun tracked narrow angle
    4 channels of instrument body temperature

    Maybe various other instrument data about the atmosphere, water vapour, etc.

    As I recall the instrument datasheets give a response time of 5 seconds but without saying exactly what that means. Sample rate is faster.

    Out of that lot something can be made. I’m not expecting anything dramatic, confirming understanding will do.