Open Mind

Seasonal Cycle in Central England Temperature

September 19, 2009 · 38 Comments

For the moment, just a graph:

Categories: Global Warming

38 responses so far ↓

  • suricat // September 20, 2009 at 12:46 am | Reply

    Tamino: So there isn’t much change in the overall UK energy levels then?

    How can UK residents understand the GW influence that you infer when the influence isn’t observed by the local populace?

    Aye, that’s a tough call!

    Best regards, suricat.

    [Response: You're clueless.

    If you're referring to the fact that the average value (zero point) of the graph doesn't change over time, then you're drawing a false conclusion. This graph shows only the seasonal cycle; the secular change has been removed.

    The secular change which is evident, recent, statistically significant, warming, and bodes ill for the residents of the U.K.]

  • dhogaza // September 20, 2009 at 3:54 am | Reply

    This graph shows only the seasonal cycle; the secular change has been removed.

    While I figured this out, looking at the animated gif. don’t be too hard on poor Suricat.

    He’s only a suricat, after all, and they’re not too smart.

    Don’t expect too much of our denialists, even though they are the ones who are overturning the work of thousands of climate scientists.

  • Kevin McKinney // September 20, 2009 at 10:48 am | Reply

    Interesting to watch this; I find myself wishing for “pause” and “rock” controls for the animation so I could better see when certain things happen.

    The two things I notice are:

    1) Amplitude changes. Naturally, amplitude decreases in the modern warming era. Not as sure about the older amplitude changes. (“Pause/rock” would help with that.)

    2) Wave shape changes. (Shifting between more triangular vs. square sine-like shapes.) Wonder what Fourier analysis shows, as these shapes have differing harmonic structure.

    It’s interesting, too, that this shift is not symmetric; it seems to affect the bottoms of the waves (winter) more than the top, as if the low temp excursions are more constrained somehow. (Ie., wave appears “clipped” a little.)

    You can tell this graph appeals to the musician in me. I’ll await more analysis!

  • Eamon // September 20, 2009 at 12:04 pm | Reply


    I think you’re going to have to add an explanation on the main points to observe, and why they’re important – as I’ve got to say the graph’s a bit obscure to this physicist.

    Looking forward in advance to an excellent piece of scientific prose…

  • Derecho64 // September 20, 2009 at 5:17 pm | Reply

    I think an “anomaly” plot would be more instructive.

  • Deep Climate // September 20, 2009 at 10:35 pm | Reply

    The changing anomaly seasonal cycle is reflected in changes in the shape of the absolute temperature annual cycle (i.e. its higher frequency composition). I would counsel patience …

  • FredT34 // September 20, 2009 at 10:49 pm | Reply

    Hmmm…. my bet is : you’re going to add a second curve – and then a third ? Plus some regressions ?

    Anyway, that’s good suspense. Count me as a visitor next days…

  • George Darroch // September 20, 2009 at 10:57 pm | Reply

    I’m sure this tells us something, I have no idea what it is. I look forward to finding it out.

  • Michael hauber // September 20, 2009 at 11:15 pm | Reply

    I couldn’t see any significant change over time, but I only watched through it twice.

    • Marco // September 21, 2009 at 6:14 am | Reply

      If I’d be evil, I’d point out that “ogling” is not the best way of finding significant changes (as Anthony Watts would have found out if he followed Tamino’s analysis of the arctic temperatures…).

  • TrueSceptic // September 21, 2009 at 11:50 am | Reply

    What’s the betting that we see some Fourier analysis, showing how the components have altered relative to one another in both phase and amplitude?

  • caerbannog // September 21, 2009 at 3:37 pm | Reply

    Sorry for the off-topic post, but I was playing around in scilab, trying the red-noise hockey-stick thing…

    And as it turns out, it was surprisingly easy to generate leading PC’s that looked very hockey-stickish using the much-maligned noncentered PCA method.

    But then I tried something else (dunno how mathematically kosher/meaningful it is) — I generated several hundred leading PC’s from my red noise per above, took their absolute values (about half of my hockey-sticks were “upside-down”) and averaged them all together.

    What I got looked a lot more like a (noisy) step-function than a hockey-stick. I did this several times with different red-noise filter setups and always got pretty much the same thing: the average of all the leading PC’s (many of which individually looked like hockey-sticks) always ended up looking a lot more like a step function than a hockey-stick.

    Any thoughts on this? (My stats background is quite rusty; I have a handle on the basics but not much more than that.)

    [Response: This is way off topic, there's an open thread for that.]

  • William Connolley // September 21, 2009 at 3:38 pm | Reply

  • Deep Climate // September 21, 2009 at 3:59 pm | Reply

    For those who didn’t catch it on the Seasons thread, the rest of this post may be slow in coming for understandable reasons.

    Response: I was looking forward to it [completing the Central England post] too. But I had a nasty fall and I’ve injured one hand, so typing is slow and painful … I’m not sure when I’ll be able to do it.

    Let me join the chorus of readers wishing you a speedy recovery.

  • Hank Roberts // September 21, 2009 at 5:30 pm | Reply

    > nasty fall … typing slow and painful.

    Yeek! If you’d like transcription help, and want to experiment, email an audio file. I can type from that and email it back to you.

  • Hank Roberts // September 21, 2009 at 5:34 pm | Reply

    Good pointer from William; don’t miss the citing papers, and this reply (not paywalled)

  • Scott Mandia // September 21, 2009 at 11:54 pm | Reply

    Well, it looks like you are about to prove what many of us already know: you can whip the deniers with one hand tied behind your back. :)

    Wishing you well.

  • Eli Rabett // September 22, 2009 at 2:12 am | Reply

    Phase shifts!!!

    (Although what would be amusing would be to look first at the difference between the seasonal cycle and a cos.)

    And yeah, where is the Little Ice Age?

  • Eamon // September 22, 2009 at 8:24 am | Reply

    Hope you recover soon Tamino!

  • GFW // September 22, 2009 at 5:02 pm | Reply

    Actually, I was wondering about the secular trend too. Took me a while to realize that “zero” was the average for that year, not some constant baseline average. A note to that effect (or better yet, a number or bar on each frame that shows that year’s zero relative to a baseline) would be good. One could see the height of the bar increase as the shape of the curve shifts.

    Oh, and have a speedy recovery!

  • Jim Arndt // September 23, 2009 at 1:08 am | Reply


    Wishing you well.

    “Eli Rabett // September 22, 2009 at 2:12 am | Reply
    (Although what would be amusing would be to look first at the difference between the seasonal cycle and a cos.)

    And yeah, where is the Little Ice Age?”

    That question is irrelevant since the time series starts with the little ice age.

  • Adam // September 24, 2009 at 8:30 am | Reply

    This link may be of interest (note the site is “broken” due to the use of Excel to create the web pages. I use the “remove this permanently” FFx add-on to clear out MS created junk that covers some links):

    Of note, Philip Eden is a very well respected meteorologist and a Vice President of The RMetS ( His version of the CET probably won’t alter any results significantly, but it might be an interesting comparison.

    (His other site is at

    Hope the hand’s better soon.

  • llewelly // September 24, 2009 at 7:24 pm | Reply

    The year number up at the top of the graph increments 5 years each frame, but the graph section shows less than 5 years.

  • Sekerob // September 24, 2009 at 9:42 pm | Reply

    Thought the first release did not have that vertical lines in. Suggest also a horizontal dotted line at Zero C

    llwelly, 5 Years? That’s I think not the point and certainly would compress the wave too much to make it obvious. 5 Year steps in the animation to again make it better visible.

  • suricat // September 24, 2009 at 10:02 pm | Reply

    Tamino: I’ve just seen your response.

    First of all, I’m sorry to hear that you’ve had a bad fall. I hope you get well soon.

    Secondly, I don’t think people in general particularly remember an “Indian Summer”, or a stretched out winter. They tend to remember extremes with greater clarity.

    Best regards, suricat.

  • Ray Ladbury // September 24, 2009 at 11:23 pm | Reply

    Suricat, You don’t know many farmers, do you?

    • suricat // September 25, 2009 at 11:07 pm | Reply

      Yes Ray, I know a few, but I know many more people that aren’t farmers as well. Many of these, others, can have their livelihoods affected by weather as well, but they tend to remember weather extremes more than extended normal/average weather.

      Perhaps this is a good reason to also look at normal weather. Perhaps there is something that’s been overlooked.

      Best regards, suricat.

  • Hank Roberts // September 26, 2009 at 1:39 am | Reply

    Suricat, you either don’t know what the data behind the chart is, or you’re trying to mislead readers.

    Can you do any better?

  • Ray Ladbury // September 26, 2009 at 2:37 am | Reply

    Suricat, do you really think that scientists aren’t looking at every aspect of the climate? Let me put it this way. Let’s say that you were competing with, oh, say about 2000 of your colleagues for any sort of recognition from making discoveries about Earth’s climate. Don’t you think you”d look at every aspect for anything others were missing that might give you an edge. Now, imagine that you are 10 times smarter than you really are. That’s kind of the way climate scientists experience the world.

    • Bob // September 30, 2009 at 10:37 pm | Reply

      I must say that I have never experienced climate scientists as being of above average intelligence (speaking as a physicist). I also observe, after doing science for many years, that the number of people who have original ideas is tiny. Most just follow the current fashion. Sorry about that.

  • dhogaza // September 26, 2009 at 4:02 am | Reply

    Perhaps this is a good reason to also look at normal weather. Perhaps there is something that’s been overlooked.

    Yes, scientists are know for not looking at normal stuff, like normal weather.

    What scientist would do that, when trying to understand abnormal weather?

    None, of course.

    This is why you’ll eventually be nominated for the Blog Scientist Nobel Award in Climate Science.

    Because you’re the first Blog Scientist to look at normal weather.

  • Hank Roberts // September 27, 2009 at 3:59 am | Reply

    Hm, I was assuming you were talking about the topic, but if you’re not talking about the Central England Temperature data, then it doesn’t matter. Are you just rambling about what people notice and recollect generally? or are you talking about the CET data set?

    • suricat // September 29, 2009 at 12:56 am | Reply

      In my first post in this thread (if that’s what you allude to) I applauded Tamino’s courage to investigate a, seemingly, mundane observation. Inter-annual seasonal variation doesn’t seem to attract much attention and there are many climate atractors that can cause this.

      The CET data is in itself an anomaly. Just like the anomaly for data collection between global CO2 levels and global temperatures. Much more than the temperature record is needed to disclose the atractors that can give rise to a temporary climate modification for any regional locality

      How could I expand on this?

      Best regards, suricat.

  • TrueSceptic // September 29, 2009 at 3:41 pm | Reply

    I find CET figures interesting as this is where I have lived all my life.

    One thing that I’ve noticed is how erratic some months are (British weather, eh?). In particular, Feb can be very cold one year and yet Spring-like the next. If anyone’s interested, download the mean data and compare Jan and Feb since, say, 1951. I remember one Feb being bitterly cold for the whole month and sure enough, there it is in 1986, when it was colder than any other month except Jan 63.

    It looks like there’s nothing to choose between Jan and Feb as coldest month over this period, and you’d expect the linear trends to be similar. When I did this, the 2 trend lines are so close they are hard to see as separate lines. There is a clear upward trend, of course.

    Anyway, just my thoughts…;)

  • Hank Roberts // September 29, 2009 at 7:05 pm | Reply

    > CET data is in itself an anomaly….
    > How could I expand on this?

    Try reading
    instead of expanding on your idea. Having a basis in statistics might help.

  • Ian // September 30, 2009 at 12:38 am | Reply

    Tamino I hope you are recovering from your fall and will soon be back in full swing. It is unfortunate you are incapacitated at this time as your thoughts on the hockey-stick controversy raging all over the blogosphere would be most interesting. Your take on the statistical significance of the, apparent, selective inclusion and omission of data would be invaluable

  • FredT34 // December 15, 2009 at 8:41 pm | Reply

    Please, Tamino, don’t let us finish 2009 in ignorance about Central England !!

Leave a Comment