Open Mind

Temperature Record of the Week: Switzerland

April 3rd, 2007 · 27 Comments

This week’s temperature record of the week is from: Switzerland.

Five Swiss stations have reported data to the European Climate Assessment and Dataset Project (ECA). For each station, the data span the time interval from 1901 to late 2006. In order to remove the seasonal cycle, I’ve smoothed the data on a 1-year timescale using a wavelet transform:

swiss.jpg


One of the stations — Saentis — is much colder than the others (because of its high altitude). So for that station only, I’ve plotted the data on a different axis: the right-hand axis is for Saentis temperature, the left-hand axis for the other four stations.

Each station shows numerous ups and downs, but they all show a sustained rise during the modern global warming era (1975 to the present). So, I’ve also computed the rate of temperature change (in deg.C per century) for all five Swiss stations, from 1975 to the present:

  • Basel 5.8
  • Geneva 7.2
  • Lugano 5.6
  • Saentis 4.3
  • Zurich 5.0

It’s quite clear that Switzerland is getting warmer; in fact each Swiss station is warming faster than the world average (1.8 deg.C per century). Even the slowest-warming Swiss station (Saentis) is warming more than twice as fast as the global average, while Geneva is warming fully four times as fast as the global average.

UPDATE:

One reader asked some interesting questions, and another provided a link to further data, so I’ll update this post.


Why the precipitious drop in one station around 1960? Did the location move from the city to the country?

The data from meteoswiss don’t show this precipitous drop; unfortunately their website is only available in German and French, so I don’t know what corrections thay may have applied. My best guess is that yes, there was a station move which caused the discontinuity in the data, and that the meteoswiss data correct for the station move while the ECA data do not. If there was a station move, then certainly a correction is necessary.

geneva.jpg

Incidentally, when computing the global average temperature at NASA GISS, such corrections are made for stations in the U.S., southern Canada, and northern Mexico, but not for the rest of the world. Denialists like to claim that such corrections are the only reason for the trend in global average temperature, but in fact: 1. corrections make the data better not worse; 2. they’re only done for U.S. and near-U.S. stations, so have very little effect on the global average anyway.


Why does one station show cooling since 1985, seemingly counter to the trend?

Actually it doesn’t show cooling. Although the numbers get lower, analysis shows that for neither data set (ECA or meteoswiss) is the trend since 1990 (not 1985) statistically significant. But the question remains, why does Saentis not show the same warming since 1990 as other stations? My guess is that its temperature history is different because of its high altitude.

saentis.jpg


Were there any other changes (third variable) that may affect the temperature readings other than climate?

There are lots of variables that can affect temperature measurements other than actual temperature. Station moves and instrument changes are two of the most important. U.S. and near-U.S. stations are corrected for these factors when computing the global average, but as far as I know other stations are not.

But the “urban heat island” effect has very little impact on global average temperature trends. In fact, the trend for rural stations turns out to be even greater than that for the globe as a whole. There’s more information about this in another post.

UPDATE #2:

A discussion arose on RealClimate about the distribution of temperatures. So I’m using this post as a holding place for an interesting graph.

I took the daily mean temperature for Geneva, Switzerland from 1990 to 2000, removed the seasonal pattern to define temperature anomaly, then computed a histogram of the data. Anomaly so defined has a mean of zero (by definition) and standard deviation 3.0527 deg.C. This graph shows a histogram of daily temperature anomaly, together with a normal distribution (in red) with the same mean and standard deviation (click on the graph for a much clearer view).

geneva2.jpg

UPDATE #3:

An idea was suggested, that the difference between the data for Geneve and Basel was indicative of “urban heating” of Geneve:


Just follow the Basel and Geneva temperatures for a moment. They start out the same, then gradually around 1920 Geneva edges above Basel. That difference increases gradually until there’s arount 1 degree C between them. The suddenly the Geneva station moves to the outskirts of town and ZAP! Basel and Geneva line up again - but not exactly. Geneva is a bit cooler than Basel for 20 years or so, then around 1980 the two are in sync. Then in 1990 Geneva edges above Basel again!

What could be causing Geneva to gradually warm in relation to Basel, more so when measurements were taken at the Geneva observatory, and less so when taken at the outskirts of town? Tamino, what I think you’ve got here is a rather exellent example of an urban warming effect. I’ll just take a guess here, perhaps someone can verify for me: Geneva grew faster than Basel over the last 100 years or so, correct?

However, another reader (Andrew Dodds) investigated the population history, and reports:


The population of Geneve has been constant since 1970; Basel has been dropping. Per capita CO2 emissions, used as a proxy for energy usage, has been roughly constant in that time suggesting that total Swiss UHI would be constant for this period.

I managed to find population data online, confirming Andrew’s statement:

populat.jpg

We can see that Geneva has not grown faster than Basel, in fact Basel grew larger until about 1970, and since that time its population has declined.

We can also compare directly the temperature time series from Basel and Geneve. There are two sourced of data: ECA and meteoswiss. Here are the differences between temperature anomaly from Geneve and that from Basel, using 1-year averages, according to ECA and meteoswiss data:

compare1.jpg

The ECA data show considerable differences. Change point analysis (using SCUSUM) identifies three change points: 1929, 1963, and 1986. Linear regression indicates no statistically significant trends during the time periods demarked by the change point analysis. Therefore the analysis indicates that the changes are not smooth, but step changes.

compare2.jpg

It appears to be incorrect to state that “…gradually around 1920 Geneva edges above Basel. That difference increases gradually…” and “in 1990 Geneva edges above Basel again!”

In fact the numbers indicate that the differences between the Basel and Geneve reports are not smooth but step changes. The first, in 1929, I have no explanation for. The second is due to the station move in 1962. The final one (in 1986) seem to me most likely to represent a genuine temperature change; Geneve warmed more than Basel.

It’s well to remember that these are the differences between Basel and Geneve, and that both stations show strong warming during the modern global warming era. But these data offer no evidence of an “urban heat island” effect — it appears to be minimal, or absent, from these temperature records.

One final comment is in order. The Geneve data show a change which is due to a station move rather than climate. When computing the global average at NASA GISS, such changes are corrected for in U.S. and near-U.S. stations, but to my knowledge, not in other stations. Denialists often point to the corrections as the only reason for a trend in global temperature, but since the U.S. is only 1.5% of the area of the globe, they actually have very little effect on the global average. The Geneve data show us two things: 1. the corrections make the data better not worse; 2. non-climate factors can go either way — in this case, the station move made the history of Geneve (based on ECA data) seem too cold in recent times (compared to the preceding time interval) rather than too hot.

Non-climate factors surely affect more station histories than just Geneve, Switzerland. But those who believe that the only non-climate factor in temperature records is the “urban heat island” effect, are misleading themselves. And those who believe that all non-climate factors in temperature records lead to false warming are statistically naive (at best); the non-climate factors can go both ways, and the likelihood is that the positive and negative ones mostly cancel each other.

Tags: Global Warming · climate change

27 responses so far ↓

  • nanny_govt_sucks // Apr 4th 2007 at 2:25 am

    Where are these stations located - rural or urban settings?

    Why the precipitious drop in one station around 1960? Did the location move from the city to the country?

    Why does one station show cooling since 1985, seemingly counter to the trend?

    Were there any other changes (third variable) that may affect the temperature readings other than climate?

  • Adam // Apr 4th 2007 at 10:18 am

    None of the cities are very big (the whole of Canton Zurich is about 1m people), and Zurich, Geneva & Lugano are on the shores of large lakes. Santis is a mountain in the northern part of Switzerland and Lugarno is in the warmer Canton Ticino

    Anyway, if you want to know about the stations, their Lat/Long and Alt values are all here:

    http://www.meteoswiss.ch/web/en/climate/climate_since_1864/homogeneous_data.html

    So you can plot them on the maps if you have the time.

    [Response: More data! Excellent!]

  • inel // Apr 4th 2007 at 6:55 pm

    I have plotted the five Swiss stations on Google Earth using the lat/lomg/alts in the link provided by Alan, if you’d like to see them in relation to rural/urban locations, ngs.

    It is worth taking a look at the aerial images here, as well as reading the summary in English taken from MeteoSwiss’s Introduction and Methodology document (einfuehrungundmethodischeaspekte.pdf)

    [Thanks for doing some of the legwork to track down the information. Readers are great!]

  • inel // Apr 4th 2007 at 10:15 pm

    tamino, I can shed some light on your thoughts here too:

    The data from meteoswiss don’t show this precipitous drop; unfortunately their website is only available in German and French, so I don’t know what corrections thay may have applied. My best guess is that yes, there was a station move which caused the discontinuity in the data, and that the meteoswiss data correct for the station move while the ECA data do not.

    The MeteoSwiss page that Adam referenced is in German, but I can understand enough to tell you this:

    for all five sites you are interested in, data was homogenized in December 2003, and up to that date, data is definitely not provisional (MeteoSwiss explain on that page that they would indicate if any data is provisional … and I have to add that the Swiss are sticklers for sticking to their own precise rules).

    After December 2003, some data may contain inhomogeneities. Homogenization, MeteoSwiss clarify, is understood to be the adjustment of earlier measured values to today’s measurement conditions (standards?). (There is no mention in any of the pages I skimmed of Geneva station being relocated, but I am sure MeteoSwiss would clarify if I asked …)

    By the way, I was in a hurry to get somewhere (late) when I wrote my previous comment, and have been aware that I need to apologise to Adam for calling him Alan, and replacing long with lomg! (I was more careful plotting on GE than I was typing my hurried comment.) Sorry, Adam.

  • Kelly O'Day // Apr 5th 2007 at 1:45 am

    Tamino:

    I’ve put together an Excel workbook for those like me who would like to play with the data for the 5 Swiss stations.

    Here’s the link.

    I’ve plotted the mean annual temperatures rather than your wavelet transformations.
    The Saentis series in your original chart showed a downward trend about 1985 which is not apparent when you plot the mean values.

    I recalcualted the rate of temperature change (in deg C per century) and got comarable results to your values. This makes me ask what the wavelet analysis is doing that using annual mean values doesn’t do? It seems to me that calcualting the annual mean temperature handles the seasonality. How is wavelet analysis improving on the mean annual value?

    Can you discuss your thinking on why you would use the wavelet transformation.

    I’ve built in some interactive capability to my chart so that I can toggle series on and off and I can move my cursor line to display exact values for specific years.

    Kelly O’Day
    http://processtrends.com

    [Response: When I calculated the trends, I didn’t use the wavelet-smoothed data, rather I used the temperature anomaly (defined as the difference between the raw data and the seasonal trend).

    Your graphs look a *lot* like mine, except they have much more year-to-year variation. In fact, using a wavelet smooth only accomplishes one thing: to smooth the data. This makes trends a little easier to see, and for some people makes the graphs more “friendly.” Taking 1-year averages will also take care of removing the seasonal pattern, and although it’s not as “smooth” as other methods, it is perhaps the single most *reliable* method — it never introduces extraneous information. In fact I only use wavelets (in these graphs) for their smoothing effect (I’m not really looking for periodicity in these data).

    For determining trends, it best to use the raw data and subtract the seasonal pattern. Smoothing introduces false statistical confidence in trends, and averaging reduces the number of data (and therefore the statistical power of trend analysis).]

  • nanny_govt_sucks // Apr 5th 2007 at 8:03 am

    My best guess is that yes, there was a station move which caused the discontinuity in the data, and that the meteoswiss data correct for the station move while the ECA data do not.

    What do you base your guess on?

    [Response: I based this guess on the fact that I’ve studied a lot of temperature time series, and I’ve seen similar patterns before — associated with station moves. I received a message from Eli Rabbett indicating that he had looked into this, and yes indeed, in 1962 they moved the recording station to the airport. He also indicates that they computed the transformation from one series to the other by maintaining records at both stations for a number of years.]

    The meteoswiss data has me completely mystified. It appears to be an adjustment for a warm bias prior to 1960, then an adjustment for a cool bias from 1960 to 1990, then no adjustment after that. What could possibly explain this bizarre adjustment? Everyone was reading the thermometer wrong prior to 1960 by 1/2 degree then suddenly in 1960 everyone was reading the thermometer wrong the other way by 1/8 degree, then suddenly in 1990 everyone got it right?

    Were there 2 station moves, one in 1960 and one in1990? If so, how did they decide which one of the 3 station locations had the “right” temperature? It looks like they went with the post-1990 location temps, but why?

    [Response: There was a station move in 1962, and that necessitates a transformation of the pre-1962 data to make the series “on the same scale.” As for the difference 1960-1990, I honestly don’t know. But I have studied a lot of temperature data, and my *guess* would be: an instrument change. It’s also well to bear in mind that I haven’t plotted the raw data, but a wavelet smooth, so they’re bound to come out differently since the ECA raw data are daily measurements while the meteoswiss raw data are monthly averages.]

  • Adam // Apr 5th 2007 at 11:19 am

    inel, no problem about the name. I’d have plotted the stations myself if I’d have the time so I appreciate what you’ve done more than a minor slip.

    Having spent some time in Zuerich I guessed that the station was outside the centre of town as the height is noticeably higher than the lakeside height, but I didn’t know where MeteoSwiss are located. Incidentally there are some stations in the city centre (esp. at the lakeside) and at Kloten (the airport) so comparisons could be made (though I wouldn’t want to claim that exposure was ideal though).

    As for Geneva, there is a good chance the station moved at some point as it is located on the airfield (Cointrin) and the record start precedes the Kittyhawk. Whether it moved in the ’60s is unknown.

  • Adam // Apr 5th 2007 at 11:34 am

    Re Lugano:

    Using this resource: http://gis.swissinfo.org/
    and comparing the station given height (273m) with the spot heights of the lakeside I’d say it is in the Parc Civico but the sat. images aren’t clear enough to make the station out. ;)

    Finally with Geneva, Cointrin is at ~45m higher than the lakeside, for what it’s worth.

  • inel // Apr 5th 2007 at 2:56 pm

    tamino,

    Would half-year trends be just as useful as annual trends to help people understand the local impacts of climate change better?

    Here’s why I ask …

    MeteoSwiss report in their summary:

    “The climatological analysis indicates a marked difference in the long-term climatic development of the winter and summer half-year. The main long-term changes concerning mean temperatures and precipitation totals took place within the winter half-year.”

    Similar emphasis on winters in particular becoming more extreme is often heard for the UK, even though Britain has a maritime climate and Switzerland’s is continental.

    I guess this polarisation of winter half-years versus summer half-years applies primarily to temperate zones (in the Northern Hemisphere) which have erratic weather patterns anyway.

    I am wondering whether 1-year averages (of temperatures and precipitation) may be smoothing out an important aspect of climate change that gets attention from scientists but could be better explained to the public. Half-yearly plots could help people get the picture more quickly.

    [Response: There is indeed value in studying the changes during different seasons. One of the consequences of temperature changing faster in winter than in summer is a change in the annual temperature range. I’ve looked at the annual range for the five Swiss stations, and I don’t see any obvious trend — so I’m beginning to be a bit skeptical about the statement that most of the change has occured in the winter season. But this deserves more study; let me look at the data in more detail, and I’ll let ya know what I find.

    And yes, taking 1-year averages will eliminate information about the relative strength of changes during different seasons. Fortunately, the ECA data are daily temperatures, so all the seasonal information is there.]

  • the Grit // Apr 5th 2007 at 9:39 pm

    Hi tamino,

    Switzerland covers an area of 16,105 square miles. One measurement per 3,000 square miles is hardly sufficient to draw any conclusions from.

    the Grit

    [Response: This is a very foolish comment. 3000 square miles is about 55 miles squared, so the “average” distance between stations will be on the order of 55 miles. The “correlation length scale” for temperature change is on the order of about 800 miles — so one station per 3000 square miles is actually rather good coverage. In fact, the difference in *altitude* of the stations is likely to have much greater impact on the differences between them, than location changes of less than a few hundred miles. And by taking temperature *anomaly* rather than raw temperature, we isolate the changes from the absolute values; then we see that there’s very strong correlation between the changes, even for stations with dramatically different temperature due to very different altitude. Note from the first graph that Saentis, which is so much colder than the other Swiss stations that I plotted it on a different axis, nonetheless shows most of the same “ups and downs” as the others.

    I strongly suspect that you want to discredit global warming because you consider it to be a “socialist plot” to dismember our economic prosperity. But consider this: if it’s really happening (and nature doesn’t care whether you’re democrat or republican, socialist or fascist, christian or jew), it probably represents the greatest threat to our economic prosperity in all the world. Try to consider the scientific case on its own merits. I suggest you download a lot temperature data (from ECA, or from GHCN) and take a good bit of time to study it in detail. You really need greater familiarity with how temperature changes over time, and across space.]

  • Eli Rabett // Apr 6th 2007 at 2:14 am

    Hi, you win. There is a long discussion of the measurements and the corrections in two Adobe Acrobat documents . Quoting from one of them in translation:

    “Geneva lies in western central Switzerland where the Rhone flows out of Lake Geneva. The conventional (they mean that people took the readings) was in the city center near the observatory (405 m above sea level) until 1962 when it was moved about 5.5 km to the Geneva Cointrin airport (430 m above sea level) at the NNW city boundary, and automated in 1980. The conventional measurements were continued in parallel until the end of 1995 since the introduction of ANETZ (the automated net of measuring stations) . In addition since 1929 measurements were made at Geneva-Cointrinfor the synotpic measurement network. For some of the parameters (they have measurements of lots of things besides temperature) that data was used as the official measurement.”

  • nanny_govt_sucks // Apr 6th 2007 at 7:32 am

    It appears that most of the recent measured warming for most of the sites mentioned occured around 1986/87. Did something else (third variable) besides climate change at that time?

    [Response: As far as I know, there was no non-climatological change at that time which could have affected temperature measurements. The really interesting thing is that *most* of the sites show this behavior. It’s unbelievably improbable that they all experienced a station move, or instrument change, or other non-climate factor, which caused a change of about the same amount, in the same direction, at the same time. The only realistic explanation, as far as I can see, is that Switzerland got hotter.]

    Where is the Basel station located? Urban or rural? It appears to show a flat trend until 1986 where there is a sharp jump up one degree. That’s odd, don’t you think?

    [I don’t know whether it’s urban or rural, but the GHCN (Global Historical Climate Network) data list the Basel station location landform as of type “cool crops.” I’ve studied enough temperature time series to know that a sharp jump in temperature at a single location is not really odd. Many locations show a sharp jump at one time or another; a steady rise is the exception. However, different locations will show jumps at different times, so the *global average* tends to show more steady trends than single locations.]

  • Kelly O'Day // Apr 7th 2007 at 1:15 am

    Tamino:

    You may remember that I wrote to you a while back asking your thoughts on the use of CuSum and Change Point Analysis on temperature trend data.

    I’ve applied CuSum/Change point analysis to the 5 Swiss stations to see if it could shed any light on the discussion. The results look pretty good to me.

    Here’s the link to my CuSum charts and change point analysis results.

    All 5 stations had a change point in 1988! 2 stations had change points in 1920, 1 station in 1919. 2 stations had change points in 1943, 1 in 1940.

    To me, CuSum charts seem to offer a relatively sensitive tool to compare temperature trends. By stacking the CuSum charts, you can compare change points by series.

    Looking forward to hearing your thoughts on this.

    Kelly

    [Response: Very impressive! If bloggers were judged by the quality and ingenuity of their readership, I’d rank very high on the list.

    I’ve only just glanced at your results, but they seem both sound and meaningful. I’ll have to study them in greater detail before I can offer much in the way of opinion. You seem to have a fondness for CUSUM analysis (and rightfully so), so you might be interested in a variant called “SCUSUM” (for “scaled CUSUM); it’s described in this paper.]

  • nanny_govt_sucks // Apr 10th 2007 at 6:50 am

    This is a very foolish comment. 3000 square miles is about 55 miles squared, so the “average” distance between stations will be on the order of 55 miles. The “correlation length scale” for temperature change is on the order of about 800 miles — so one station per 3000 square miles is actually rather good coverage. In fact, the difference in *altitude* of the stations is likely to have much greater impact on the differences between them, than location changes of less than a few hundred miles.

    But in the Geneva series we see that a station move to the outskirts of town (I assume the airport was on the outskirts in 1962) is reflected by a 1.5 degree C drop in temperature. Certainly that move was less than 55 miles. Certainly there is not much of an altitude difference between the two locations. Obviously there are some temperature variations on smaller scales that need to be accounted for before drawing conclusions from such sparse data.

    [Response: First of all, the difference between the two records (ECA and meteoswiss) is most definitely *not* 1.5 deg.C; it’s more like 0.5 deg.C. Most of the dip in 1962 isn’t due to the station move, it’s a genuine (albeit temporary) temperature drop, as also recorded in Basel, Lugano, and Zuerich (note how strongly the records from those different stations are *correlated*). For you to state otherwise is disingenuous.

    Second, anyone who looks at the graphs can see that nearly every little rise and fall in the Geneve series is mirrored at Basel, Lugano, and Zuerich — the *correlation* between the temperature time series from different Swiss locations, separated by considerably more than 55 miles, is nothing short of astounding.

    Third, this much is abundantly clear: you are far less interested in what’s been happening with temperature in Switzerland, than in casting any aspersion you can get your hands on. Why else would you deny correlation between different locations, when the correlation is so clearly *visible* in the very first graph? Why else would you ask *three times* whether the increase around 1988 was due to a non-climatological factor, even after readers had worked to dig up the history of the record-keeping (but you had not bothered to do so)? Why else, when faced with incontrovertible evidence of a change in 1988, would you state that there’s still no temperature change before or after (when it’s obvious you didn’t run the numbers)?

    I’ll certainly give you credit for persistence. But shame on you for casting so many doubts, without bothering to do any of the *work* required to find the answers. This suggests that you don’t really want answers — you just want to muddy the waters.]

  • Adam // Apr 10th 2007 at 9:27 am

    As far as I’m aware, the only instrument change that has brought about a “real” change in what is recorded is in sunshine recording. Thus long-term sunshine records need to be treated with caution, but other instrumental changes should have no or little effect (esp. if they are sited in the same Stevenson Screen or enclosure).

  • nanny_govt_sucks // Apr 11th 2007 at 8:36 am

    Tamino,

    Where have I denied a correlation between the temperature records of the Swiss stations?

    I was simply pointing out a huge drop in temps for Geneva at the time of the station move that is unmatched in the other series. Eyeballing the difference… looks to be about a .75 degree change simply because of a station move to a nearby location. Don’t you think that’s significant?

    And now that I look more into it, I think there’s something FAR more interesting going on…

    Just follow the Basel and Geneva temperatures for a moment. They start out the same, then gradually around 1920 Geneva edges above Basel. That difference increases gradually until there’s arount 1 degree C between them. The suddenly the Geneva station moves to the outskirts of town and ZAP! Basel and Geneva line up again - but not exactly. Geneva is a bit cooler than Basel for 20 years or so, then around 1980 the two are in sync. Then in 1990 Geneva edges above Basel again!

    What could be causing Geneva to gradually warm in relation to Basel, more so when measurements were taken at the Geneva observatory, and less so when taken at the outskirts of town? Tamino, what I think you’ve got here is a rather exellent example of an urban warming effect. I’ll just take a guess here, perhaps someone can verify for me: Geneva grew faster than Basel over the last 100 years or so, correct?

    [Response: Interesting idea. But the data contradict it. See UPDATE #3 to this post.]

  • Andrew Dodds // Apr 11th 2007 at 3:07 pm

    NGS -

    http://travelguide.all-about-switzerland.info/major-swiss-cities-population-map-distances.html

    The population of Geneve has been constant since 1970; Basel has been dropping. Per capita CO2 emissions, used as a proxy for energy usage, has been roughly constant in that time suggesting that total Swiss UHI would be constant for this period.

    So no, this is not UHI you are looking at.

  • Adam // Apr 11th 2007 at 3:29 pm

    All but Lugano had a drop at about the same time. It’s hard to see how the Geneva enclosure move could have an effect on the temp. drops elsewhere.

    Plus, as has been said there was an overlap in the station usage at the two sites.

  • Dano // Apr 11th 2007 at 6:28 pm

    The tactic is to throw out a bunch of stuff to make it look like there is doubt.

    Better to make the FUD purveyor produce the numbers to back their assertion. What usu. happens when you make someone back their assertion is either silence or hand-waving, both of which are beneficial outcomes.

    Best,

    D

  • the Grit // Apr 11th 2007 at 8:15 pm

    Hi tamino,

    While I do see it as a plot, although not socialist in nature, I also give due consideration to the facts. In this case, being a farmer, I pay close attention to our local weather. It is most common to see a three to five degree reported temperature variance across Shelby County, which is a smaller area than you mention. This difference dwarfs not only the “global warming” claims, but the local Swiss data.

    As to one station per 3,000 square miles being good coverage, you have just made my point about the fundamental lack of data to support Global Warming. Really, if you truly believe that there is not a higher normal variation across that area than the claimed average temperature increase, it is your political motives that are in question.

    the Grit

    [Response: There can indeed be large differences between stations much closer than 55 miles, just as there are very large difference between Swiss stations. Saentis is, on average, cooler than Lugano by about 14 deg.C (over 25 deg.F). And changes in *altitude* can have even more dramatic effect than changes of several hundred miles in location.

    But the *changes over time* are very strongly correlated — if you need to know the absolute temperature at a given location, at a given altitude, you’d be well advised to put a thermometer right there. But if you want to know how the temperature is changing over time, whether it’s warming or cooling, over the long term (annual averages), the correlation is so strong that spacing your instruments 100 miles apart is plenty good enough to do the job. That’s why climatologists studying global warming tend to look at *temperature anomaly* (the difference between the temperature, and its average at the same location during a “reference period”) rather than absolute temperature.

    Global warming is about changes in the global average temperature on decadal time scales. For that, Switzerland has plenty of data.

    I’m not at all surprised that one side of Shelby county (?) is hotter than the other. But I’ll bet that when one side of the county warms up, the other does too. And if you look at annual averages, I’ll bet the *time changes* between different sides of Shelby county are even more strongly correlated. For more information, see the latest post.]

  • Andrew Dodds // Apr 12th 2007 at 12:18 pm

    Mr Grit -

    Currently, the single greatest strategic threat to the US is the massive dependance on imported oil. In the relatively near future, you may add a dependancy on imported LNG - and you can ask Iran and Russia about the ‘Gas OPEC’ they are planning.

    Large scale action to reduce CO2 emissions would by nessicity reduce or remove these problems for the US. This should be extremely obvious.

    So if there is any anti-US plot going on here, it’s the global warming deniers who are plotting, the Exxons of this world who will make billions whilst selling the US to the highest bidder.

  • Dano // Apr 13th 2007 at 4:22 am

    Andrew,

    Currently, delusionists think the single greatest strategic threat to the US is liberals and scientists.

    They will hear nothing else.

    Best,

    D

  • the Grit // Apr 13th 2007 at 11:10 pm

    Hi tamino,

    Actually, around here, none of that is true. Where the local warming and cooling is depends on minor variations in the general West to East wind pattern. Then, on occasion, there are the strong storms that come down from the north-east. It is not at all uncommon to have thunderstorms and/or tornadoes, with their associated temperature change, ten miles to the north or south, while having clear skies and stable temperature here. The idea that one temperature reading in a 100 mile radius is “plenty good enough” is exactly why I, and others, are so skeptical of the Global Warming stuff.

    I try to keep an open mind (while mocking the politics of it) but, the the potential canceling of the Boston Marathon due to the cold weather for the first time in 110 years makes it difficult to keep a straight face :)

    Seriously, I appreciate your reasoning and input on this subject. Thus, while we disagree on how many data points are needed to get an accurate picture of global temperature changes, I am also most interested in data on CO2 output, so, if you stumble across it, kindly give me a pointer.

    Hi Andrew,

    I have been, since junior high, which was way too long ago to mention, a supporter of alternative fuels. I wrote an essay, 30 something years ago advocating that we shift to a hydrogen fuel source economy, saving our oil supply for making plastics, My theory was that by building nuclear reactors along the coasts, the power could be used to produce hydrogen from sea water, with the bonus of collecting the dissolved mineral content of the separated water. Just a dreamer I guess.

    the Grit

    [Response: I think you’re falling into a classic trap. I’m sure all the things you mention really happened, but I very much doubt that you took note of all the times that temperature changes, or other weather changes, were the *same* on different sides of the county. In fact, the things that people notice and remember, tend to be the *unusual* things, while the normal course of events quickly escapes recollection.

    Rather than rely on anecdotal evidence, let’s look at real data for your location. Remember we’re not talking about “not at all uncommon,” we’re talking about annual-to-decadal trends in temperature change. If you’ll tell me what state Shelby county is in, I’ll track down the data for the nearest several meteorological stations, analyze it, and report the results, regardless how they may come out.]

  • tamino // Apr 14th 2007 at 3:09 pm

    Re: request for CO2 data.

    This from the U.S. Geological survey gives information about the emissions from volcanoes, and how they compare to those from human activity:

    http://volcanoes.usgs.gov/Hazards/What/VolGas/volgas.html#co2

    This from Wikipedia gives essentially the same information of CO2 emissions from volcanoes:

    http://en.wikipedia.org/wiki/Volcano#Effects_of_volcanoes

    Data for CO2 emissions from Mt. St. Helens can be accessed here (CO2 emissions per day are in the final column, in tonnes per day):

    http://pubs.usgs.gov/of/1994/of94-212/Core/raw/table.txt

    As for emissions due to human activity, this site gives data for various regions of the world:

    http://cdiac.ornl.gov/trends/emis/em_cont.htm

    The EPA also provides information for the U.S.:

    http://yosemite.epa.gov/oar/globalwarming.nsf/content/ResourceCenterPublicationsGHGEmissions.html

    Finally, the most comprehensive source of data for CO2 concentration in the atmosphere is the World Data Center for Greenhouse Gases:

    http://gaw.kishou.go.jp/wdcgg.html

  • the Grit // Apr 14th 2007 at 8:49 pm

    Hi tamino,

    Shelby County is in Tennessee. Much of it is taken up by Memphis, home of the King. I look forward to seeing your results.

    Also, thanks for the CO2 output references.

    the Grit

  • Gaudenz Mischol // Jun 6th 2007 at 6:59 am

    Just some information from a Swiss guy
    Population in cities have been stable or declining, just because suburban areas have been developing very fast, people are moving out of town. Traffic jam on the highways into the cities have become more and more a problem.
    All stations except Saentis are “big” cities for Switzerland, even if in Zurich there is only a population of 1m, the whole metropolitan area spreads far out in what used to be once countryside. Along the highway between Zurich an Berne ( the capital) there is a massive development and soon can’t be regarded as urban.
    I’m puzzled to see, that only one station out of four used by GISS is really urban (Saentis) all other are fast developing metropolitan areas.
    No wonder you see fast warming there and a decline in the urban station since 1990

  • Gaudenz Mischol // Jun 6th 2007 at 7:04 am

    Correction: one station out of FIVE is urban

Leave a Comment