Europe – Atlantic and Coastal

Atlantic / Coastal Europe – The Warmer Parts

Some with both Atlantic and Mediterranean coasts. Some perhaps suited to the Central Europe group, but small enough and near enough to water to be strongly water moderated; but not so cold as the North Atlantic group. That transition from cold North or volatile East to warmer and more moderated places.

Madeira Islands (Portugal) – 654

Fascinating. Here we get an added thermometer just in time to make a nice “little dipper” in the baseline period. Then it leaves right at the end, with a rise as the result. A 1980 “splice” jumps us up a ways, and then a 1990 Pivot to a great heating run. Guess a group of islands off shore, being responsible for a 1200 km or so radius of surrounding surface deserves the full treatment. About 1.5 C to 2 C “warmer” than before. Wonder if there are any sea surface records or a history from the Azores or Canary Islands that might be compared… Ah, yes, in Africa, The Canary Islands. Need to check back here after I’ve got Africa done.

For now, though, we’ll just have to wonder why the middle of the Atlantic has a hockey rink while Gibraltar does not…

Madeira Islands Monthly Anomalies and Running Total by Segments

Madeira Islands Monthly Anomalies and Running Total by Segments

Portugal – 636

Portugal is running along dead flat at about -1 C right up until the 1961 start of the UEA / CRU baseline, then it gets more thermometers (about 20 of them) and a cold plunge. (Strange how in the USA that happens in 1951 while in Europe it happens in 1961… purely a statistical quirk, I’m sure /sarcoff> ) In about 1981 we yank them out again (and maybe yank out the non-warming or colder ones?) and get a nice “bump” up. Then the 1990 to date period gets a lot of persistent pot stirring of thermometer change. The Pivot looks to flatten the trend some (probably as we’ve now got a nice cold baseline built in… don’t want to be outrageously out of touch with reality…)

I love the way the “low going spikes of hair” just plunge in the baseline period, and get completely clipped off just after it. 2 C of “clip” in one thermometer swap. That 1990 change is interesting too. A big spike up of “hair”, then a downward “neutralize” spike, followed by a tuning period where the “up hair” ends up just about 1/3 C more than the down hair and with all risk of random volatility squashed out of it. Nothing more than +2 C up or -1.5 C down. Wonderfully played. Wouldn’t want those -3 C spikes of the late 1990’s messing up such a beauty of a “Little Dipper”…

Portugal Monthly Anomalies and Running Total by Segment

Portugal Monthly Anomalies and Running Total by Segment

Spain – 643

Ah, Spain fascinates me. Long slow drop into cooling. Even the 1980’s thermometer cuts can’t fix it. Barely a pause. Flattened, but not rising. Somebody better do something! So we have a big “double bullseye” in the 1991 to 1995 range with a spectacular jump up accompanied by a LOT of thermometer change. We take ’em out, we put others in, we take some more out, stick some others in. “Where did I leave those damn jet airport records?” you can almost hear someone swearing. And in the end, we are ‘back to zero” were we were in 1857 and 1901. But “not to worry” as the baseline is nice and cool in comparison. I think this one may qualify for “The Big Dipper”. It certainly took a lot of work…

Spain Monthly Anomalies and Running Total by Segments

Spain Monthly Anomalies and Running Total by Segments

Gibraltar – 653

Who cares about Gibraltar? I’m sure someone thought there was no need to tune it up, as it will simply be homogenized in to everything else. So we have an interesting “rolling” aspect (ocean oscillations?) and in general we start and end at about zero. No global warming in Gibraltar, though there is the cloying hint of a general drift lower with a halfhearted neutralizing splice at 1990. There is also a ‘bullseye’ splice at about 1952, though with little impact.

Gibralter Monthly Anomalies and Running Total

Gibralter Monthly Anomalies and Running Total

France – 615

France, having a “Hide the Decline” moment. Nice long steady downward trend, but just too big to truncate. So, a little rearranging and “Presto!” no more decline!

France Monthly Anomalies and Running Total by Segment

France Monthly Anomalies and Running Total by Segment

Belgium – 606

Nearly dead flat, then a Hockey Stick blade glued onto the end. Nice, very nice. Who knew Belgium was so big in hockey?

Belgium Monthly Anomalies and Running Total by Segments

Belgium Monthly Anomalies and Running Total by Segments

Netherlands / Holland – 633

So here we are, “cheek by jowl” with Belgium. Not a hint of warming. No hockey stick. A trend line that rises about 3/4 C in 300 years, but with significantly warmer excursions in the past. Zero crossing all the way. I love the Dutch (and not just because some of my ancestors came from that part of the world (via England). But because they can be cussedly persistent about being honest with themselves about the reality that confronts them. If you would keep the ocean away, you can not afford to fudge your numbers…

This graph REALLY needs to be looked at in full size, so click on it to see the “trend” up close and personal.

Netherlands Monthly Anomalies and Running Total

Netherlands Monthly Anomalies and Running Total

Luxembourg – 629

Yeah, not exactly “coastal” but near France and needs to go somewhere, so why not here?

A short record that starts and ends at zero, but with a bit of a dip in the baseline middle as thermometers come and go. Again with the curious match to the 1961-1990 UEA CRU baseline rather than the 1951-1980 GISS baseline we see in the Americas data. Surely just an accident… (Though I do wonder if they haggle over Asia and Africa? A reprise of global colony politics? Must be fun to hear them discussing changes to the baseline…)

Luxembourg Monthly Anomalies and Running Total

Luxembourg Monthly Anomalies and Running Total


Other than continued admiration for the Dutch and wondering why the French have let themselves be “sticked” in the end; my major observation is that if you gave me this basket of trends and graphs and said “Average them and tell me what it means” I would say “Averaging them would mean nothing, they are all different.”

Furthermore, I’d be really curious why thermometer counts jump all over at the entry and exit from the baseline intervals (and with “coincident” changes of anomaly trends…)

And finally, I’d ask “How can Spain, France, even Portugal have long duration cooling trends, that suddenly spike up, if most of the ‘CO2 effect’ was already present in the air long before then?” And why does it happen right when you get a change of “Duplicate Number” attesting to a changed process?

Postscript: It looks like Gibraltar deserves a Carbon-Guilt free visit (drive through France and Spain to get there, they can use the warmth ;-) and spend a bit of time at a nice cafe…

And, given their stalwart stand against tides and storms, both natural, and undoubtedly some political “heat” over their record: everyone planning a visit to Europe simply MUST visit Holland. All time spent there is completely Carbon-Guilt Free. Heck, they even have windmills…

About E.M.Smith

A technical managerial sort interested in things from Stonehenge to computer science. My present "hot buttons' are the mythology of Climate Change and ancient metrology; but things change...
This entry was posted in dT/dt and tagged , . Bookmark the permalink.

5 Responses to Europe – Atlantic and Coastal

  1. tarpon says:

    The warmists added more thermometers so they could pick and keep the ones that were hottest.

    Science for dollars.

    REPLY: [ Not “hottest”… “less cold”… watch this space ;-) I’m working on a France detailed look… -E.M.Smith ]

  2. E.M.Smith says:


    I’m Shocked! Don’t even begin to think it would be DOLLARS!

    This is Europe we’re talking about here. Euros, please! ;-) and maybe a few Pounds Sterling…

    (Though I’m trying desperately to hold onto a hope that it could just be a “Stupidity Splice”… it’s becoming harder to hold that option open from day to day. A “Stupidity Splice” ought to be pretty much “Ham Handed” across all comers. This is looking very much custom tailored to each country… especially things like that “Don’t do the Dutch, they will notice that we don’t have the same numbers” and the amount of pot stirring needed to get Spain to warm up… that is just NOT a stupidity “one size fits all” goof. IMHO, of course…)

  3. Alan Davidson says:

    Very interesting results! I think the link that you mention between the different baseline periods and the trends you see in suddenly added thermometers, is a really significant indication that deliberate manipulation of the temperature trends is being done both in NASA and CRU.

    By the way Canary Isles is near to the African coast but is part of Spain. Several islands Tenerife, Lanzarote and several airports so I’d guess a number of reporting stations and data should be part of Spain’s records on-line somewhere.

  4. Bob Highland says:

    Well, E.M., as far as I can see you’ve absolutely nailed it. Your country-by-country, month-by-month analysis of the (presumably, comparatively) raw data is an inspired choice. It seems such an obvious thing to do; but then, all really good ideas seem obvious once someone, in this case your good self, has actually taken the time to do the hard yakka. One might have thought that someone with the time and resources provided by a handsome research grant would have had a crack at this kind of analysis. Maybe someone even has, and then quickly destroyed all trace of the evidence, sensing an imminent end to a rent-seeking career in ramshackle climate “science”.

    I particularly like your approach to averaged cumulative anomalies which so neatly disposes of the missing-data issue. I’m always leery of fabricating and infilling missing data, especially when the methods used are so arbitrary. Just using the real data that actually exists, assuming there’s enough of it, seems to me a far purer statistical approach. Indeed, your method would appear to be unexceptionable, which probably explains why, curiously, nobody has turned up yet to take exception to it.

    I also love the fact that you’ve used THEIR data as is, which, even if it has been messed with in questionable ways to fudge the raw readings, still supports your conclusions. (Is that scurrying I hear the sound of them rushing to retrospectively adjust the data once more?)

    I’m looking forward immensely to your final summary of this exercise, when you get the time to do it, which will no doubt elegantly postulate and conclusively prove that the modest apparent warming of the last 30 years is no more than an artefact of land thermometer changes and counts. This surely warrants elevation into a paper of some kind so that it can be subjected to review. (But not necessarily journal “peer review”, that vile, corrupted process so beloved by those who have the power to manipulate it.)

    Oh, and by the way, if any of the usual suspects does try to criticise your methods citing obscure statistical techniques, I suggest you enlist the assistance of a poster who uses the moniker ‘VS’. This is a person of no mean ability in the field, who has stirred up a hornets nest in this forum:

    His (brilliantly supported) contention is that the nature of the temperature and CO2 concentration data series are such that correlating them to show a connection by the usual statistical methods is invalid, even when it contains the hockey-stick data. It would be great to have him go to work on your truer representation of the real data.

    More power to your elbow, Michael!

  5. E.M.Smith says:


    The Canary Islands has it’s own “Country Code” of 159 so it’s easy to break them out. Often a part of a country that’s in a different geography gets it’s own “Country Code”. The most extreme of these is Russia, that gets a European code for West of the Urals and an Asian code for East. I’ve already run the report, just not done the graph… Pivots about 1992 and rises about 1.4 C pronto. More station swapping…

    Yeah, it took me a while to catch on that the thermometer counts always ramped up and back down on the baseline periods and that provided a nice way to fabricate a “little dipper” and follow it with a hockey stick if needed. While I hate to say it, I’m good enough, now, at this identification of anomaly trends stuff that I could create any pattern you want in the “product” by selection bias in the stations. Just run my “hair graph” for a set of stations and pick the ones you want in / out in each part you splice together… During a “warm phase” put in high volatility stations in warm places. During a “cold phase” put in low volatility stations in cold places. You catch the “rises” more than the falls… Want a cold bias? Just reverse that. Put IN the high volatility cold places when a cyclical process (like PDO) is going cold, then swap to the warm low volatility places when it swaps to warming. You capture a lot of cold excursions, but miss the warm ones.

    So, like it or not, I could custom craft a replacement for GHCN to show a far different result (and only using their methods of “filling in” missing data, selection bias, and splicing stations).


    Take a look at the “France – Hide The Decline!” posting for a preview of “things to come” ;-)

    And thank you for the kind words. I’d tried to make dT/dt dirt simple and bulletproof. Considered several options, but just didn’t like the “gap problem” being fixed with filling in (and it’s seem quite reasonable to me to just wait for the next valid monthly value. It ought to “self heal” from any broken values rather than propagate them.) Frankly, the best thing about it is that it IS so simple. There’s not a whole lot to complain about or attack.

    Step1: Make month to month anomalies for each thermometer. On dropouts, wait for a valid value, then make anomaly. Can’t see much of a way to say that June in Paris is somehow not valid to compare with, oh, June in Paris…

    You could argue that a missing 4 years ought not have all the change lumped into year 5 (so 0.1 C / year would end up 0.4C dumped into year 5) but then you are just averaging that month in with 11 others and potentially more thermometers so it ought not have much impact; and what impact there is would be a momentary jump in year 5 as ‘catch up’. Which is exactly what’s in the data… you don’t REALLY know if it was 0.1 C / year or 0.4 C in year 5.

    Step2: Average them by year for a given place. (That is, add up Jan-Dec and divide by 12. If each month rose by 1 C, then you would get 12 / 12 or… 1 C. Kind of hard to argue with that…)

    Step3: Make a running total of net change over time.

    If you are gaining 0.1 C in each year, the running total over 10 years would be 1 C and you would have risen by 1 C. Hmm. The Dirt Simple thing again…

    And that’s it.

    Yeah, UHI is “still in there”. Better left in and visible than “sometimes / often broken” like GIStemp.

    Yeah, there is no homogenizing into ‘dead zones’. So Bolivia “ends” in 1990 when their data ends. Again, seems like stating that’s all you really know…

    Yeah, it does not make a regular geometric grid of boxes. But you can easily just think of a country as just a large grid / box. Heck, a lot of them are smaller than the GIStemp Grids… In GIStemp there are cases of a single thermometer filling in 1200 km in a couple of different directions. I’d prefer to say that “Diego Garcia” is doing FOO than to say 1200 km radius around Diego Garcia is maybe doing BAR, or not.

    So at the end of it all, me and my elbow are rather happy with the results. It seems to “listen politely to the data” fairly reliably. Sometimes you find warming, sometimes cooling, and sometimes odd pot stirrings of the thermometers…

    At some future time, after getting the forensics part out of the way, I may come back to it and try to make a version that “hides the splicing and pot stirring” better, just to see how well that works. Or maybe not ;-)

Comments are closed.