How Long is a Long Temperature History?

Guest Posting from Tonyb

In comments Tonyb made these remarks. He also said:

“Hope this is in the right place please relocate as appropriate.”

Well, I think it is appropriate as a posting, so with that, take it away Tonyb:

GHCN Stations by Geography and Age

GHCN Stations by Geography and Age

Full size image at: Wikipedia

I have been analysing longer temperature data sets and am writing an article on my findings. Thought you would be intersted in this snippet as it relates to the Uppsala temperature record you carried recently.

“I want to take you on a brief journey through time to the Little Ice age thermometers that predate the CRU dataset.
Here are CRU global temperatures commencing 1850

1850 CRU Temperatures

1850 CRU Temperatures

Original full sized at: http://www.cru.uea.ac.uk/cru/data/temperature/nhshgl.gif

Amongst the longer lived records are two that I wish to highlight, as they complement each other. Stockholm commenced recording in 1756

1756 Stockholm Temperatures

1756 Stockholm Temperatures

Original at: http://www.processtrends.com/images/temp_stock_ann_trend.gif

It provides interesting information as the graph shows clear peaks and troughs, but particularly intriguing is that a couple of years ago Stockkholm recorded its mildest winter since ‘records began’ (in 1756) which heralded the start of much publicity about global warming.

However, by a delicious irony, we find that the home city of Arrhenius himself-Uppsala-had an even longer temp data set than Stockholm, from which we can see the upturn in temperatures to a period warmer than today during the 1720-1740’s which includes a series of very mild winters.

Uppsala Temperatures

Uppsala Temperatures

Original: http://www.smhi.se/sgn0102/n0205/upps_www.pdf

Intriguingly, both cities have had substantial studies made on them to identify the Urban heat island effect. Uppsala for example expanded three fold from 1850 to 1890 and continues to develop. Both data sets are due to be amended to reflect this. (Note there are lots of caveats with siting, uhi, instrument reliability etc)


http://www.statssa.gov.za/isi2009/ScientificProgramme/IPMS/0189.pdf

If we look before 1850 we can see considerable temperature variations belying the notion that today is unprecedented and that variability in the past was limited, as co2 at 280ppm was not sufficiently high to be a primary driver, in contrast to today.

If we then go to the granddaddy of them all- Central England Temperatures (CET) not only can we see the huge temperature fluctuations each year (the data is not smoothed) but a confirmation of the peaks around 1720 (when Uppsala commences) and in this case ending at a l trough in 1660.

CET Temperatures

CET Temperatures

Original full sized and readable at: http://cadenzapress.co.uk/download/beck_mencken_hadley.jpg

We know that temperature goes through other peaks and troughs-for examples the 1530’s and 40’s are known to have been very warm, as were the 1420’s and 30’s, the 1300’s generally were very cold and a peak of warmth was reached in the Medieval warm period from around 850 to 1250AD (although there were cold spells within that.)

To the surprise of no one we have got warmer since the end of the LIA, but we are able to see the modern era in a much better context as just part of a constant variation.

Natural variability has enabled our present civilisation to enjoy what appears to be a period of ‘comfortable normality’ with our age comfortably placed as instrumental temperatures meander gently somewhere between the LIA and the MWP values-despite liberal enhancement of UHI in some cases. Our equitable situation doesn’t require legislation or expensive remedies. Enjoy it while you can- until nature throws the next extreme at us

tonyb

Thanks, TonyB. My comments follow

This part started life as a comment over on WUWT, but I’ve decided to “preserve” it here, since I’ve posted similar things a few times.

E.M.Smith (16:22:05) : (in response to)
JimInIndy (14:42:51) :
“I was born in 1937. I don’t put much stock in 30 year trends. Let’s look back at the low ice levels of the pre-WWII, pre-fossil fuel exploitation, pre-CO2- increase period and explain the high temps of the 1930s, compared to the lower temps of the 2000s. A longer perspective sometimes offers a better focus.”

Truer words were never spoken. This ought to be printed out 10,000 times and sent in paper mail to the jokers looking at arctic ice. Maybe if they had to read it that many times it would sink in, just a little bit…

(Realize that times runs the other way on these very long history ice charts. You’d think climate folks would have standardized on one direction for time… Oh Well.)

Take a look at this chart:

Ice Core Isotope Temperatures

Ice Core Isotope Temperatures

Original full sized image

140,000 years of “temperatures” via proxies.

Here is a close up of the last 40,000 years in ice:

Ice Core Vostok Temperatures

Ice Core Vostok Temperatures

Original full sized image

Notice for that the entire 10-12,000 years of the Holocene we have been in a general downtrend. Slowly, inexorably, cooling. Notice that it is an incredibly flat stable time when compared to the rest of the 140,000 years. Then ask just exactly how “extreme” our “climate change” has been when it has been ’steady as a rock’ in comparison to the past…

Now look back to the LAST interglacial. Notice the “pop and start dropping” with not very long at the top? (Though the top of the “pop” was hotter than our peak this time.) We are incredibly lucky ours has been flatter; and we will need to be ever more incredibly lucky if we are to prevent that “drop” this time by any means possible.

There is some process that acts as a hard lid on temperatures just a bit above our present temperatures. (If you look at longer duration charts you see all the inter-glacials whack into it and bounce off). There is no such protection to the downside. ALL the risk is to the downside.

The cold has rapid onset, but the ice build up (bottom line) is slower:

Ice Age Ice and Temperatures

Ice Age Ice and Temperatures

Original full sized image

No, it isn’t an issue any time soon. The ice extends as a wobbly linear trend in a glacial. Take the max extent at last glacial and measure the distance to the Greenland sheet. Divide by 100,000 years. You get the ice advancing at about 800 FEET per year.

We could easily already be in the “next” glacial and the LIA might have been the start. We wobble that much, but the max extent of ice in the NEXT LIA ought to be all of “800 feet further south per year since the last LIA” (at the bottom of the next one). Not the kind of thing to get excited about in any one human lifetime… Call me in 1000 years and we’ll see if the ice is 800,000 feet or about 150 miles further south than in 1816.

You can walk south farther in a few minutes than the ice advances in a year, on average. That’s the “fun” part of the “Ice Age Is Coming!” disaster scenario. You get to have all the disaster and panic talk, but nobody gets hurt for 1000 years ;-)

The real question we ought to be asking is “Why has this interglacial been so stable and hospitable to life when prior interglacials were pop-and-drop spikes? And how do we keep this one from dropping off a cliff like the last ones?”

Update: Ellie, in comments, provided a nice UK Graph

The graph below is of the entirety of the UK data, from the first thermometer to the last, in the Global Historical Climate Network data set downloaded from NOAA at:


ftp://ftp.ncdc.noaa.gov/pub/data/ghcn/v2

The U.S. Historic Climate Network (USHCN) data are at:


ftp://ftp.ncdc.noaa.gov/pub/data/ushcn

though they are also in the GHCN data set.

I detail where to get the data along with where to get more guidance in:

https://chiefio.wordpress.com/2009/02/24/ghcn-global-historical-climate-network/

So what is this graph? You can down load MAX, MIN, or AVERAGE data. This is the average of the daily high and low temperatures for each day of the month, that are then averaged together by NOAA to make a “monthly average mean”. I have taken those “monthly average mean” data and averaged them together for the whole year and for all the thermometer records in the UK Country code. Ellie has been kind enough to turn that into a gif for me, and here it is:

GHCN UK Stations Annual Average of Monthly Means

GHCN UK Stations Annual Average of Monthly Means

Original full sized image.

We can see the more severe winters of the LIA. The “1800 and Froze To Death around 1810-1817”. The 1920 to 1930 or so warmth during the “dust bowl era” then the plummet into the 1970’s “Ice Age Is Coming” scare. Also interesting is the 1850 to 1880 or so warmth. As I remember it, that was something of a golden era of the British Empire. Looking at the red line, we have about an 80 year long “ripple”. That GIStemp chooses to “start history” in 1880 at the bottom of one of those ripples is, IMHO, no accident. The peak tends to come about 45 years after the bottom of a dip; and the last dip was about 1975, so that would make it 2020 that we fall off the top again. Right in keeping with the current “Sign your life away so we can save the planet by 2020” chant around the Copenhagen meeting. Only someone may have forgotten that a Simple Moving Average lags the reality by about 1/2 a period So call it 2005 for the “turn”. Gee, hasn’t it been getting cooler on a global basis since, oh, 200x? Another interesting feature is the tendency for the top peaks to “hit a wall” at about 10C. There is an upward trend to the data, but it is very small, and it is not the tops getting hotter, it is the bottoms moderating the severe winter lows. This also happens at the tops of ripples. Hardly anything terrible, IMHO. Especially coming out of a Little Ice Age as we were over this time.

The thermometer count rises over this time period from the one and two range, to 40s, then back down to 10. It is highly likely that the averaging of more thermometers will also have a dampening effect on the extreme events in the average, Inspection of the actual sites used for the early records might be enlightening.

I see nothing at all to be worried about in the UK temperature trends.

Advertisements

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 AGW Climate Perspective, AGW Science and Background, Favorites and tagged , . Bookmark the permalink.

76 Responses to How Long is a Long Temperature History?

  1. E.M.Smith says:

    BTW, the “peaks” on the CET graph are roughly (for my thumb and monitor) one thumb width apart. A semi-regular spacing.

    Sure looks like a quasi-periodic cyclical thing to me with about a 60 (ish) year return to 10.5 C (and what do we see at about the 1875 – 1880 point? One of the three sharpest down spikes on the record. Nice time for GIStemp to ‘begin history’…)

    When I download the PDF at:

    http://www.statssa.gov.za/isi2009/ScientificProgramme/IPMS/0189.pdf

    I get a bucket of bits, but nothing readable on my screen. Do I have the link wrong or is it new release of PDF format and I’m being a bit slow ;-)

  2. j ferguson says:

    Odd, download worked for me. link must be ok.

    Reader 9

    regards, John

  3. E.M.Smith says:

    I’m probably just using an old version of Adobe. Who knows how long ago I last updated…

    UPDATE: I think the issue is that while Adobe is doing OK, my default puts the pdf into a Safari window and this Safari is at least a half dozen years old and maybe older… Guess it’s time for a browser update…

  4. Ripper says:

    There is some unheralded great work going on here.

    With all the debate about proxies lately , it struck me that these temp records and E.M. Smiths top 10% of long lived thermometers are far better proxies than tree rings.

    With all the cramming temps into boxes et al, it struck me that the US “Proxy” and Australian “Proxy” of the 1930’s more accurately reflect the global temp than the calculated one.

  5. Tony Hansen says:

    Tambora and Krakatoa seem to show up but what is going on with 1695 and 1740?

  6. E.M.Smith says:

    @Ripper:

    BINGO! We have a winner!

    @Tony Hansen:

    I donno… anyone else have an idea?

    I’m going to add another really long temperature for perspective “in a bit”… right now I need more coffee… ;-)

  7. Tonyb says:

    Yes the link is fine so it must be your adobe version.

    I recently posted this elsewhere, but it is highly relevant to this thread so forgive me for repeating it.

    I am currently writing a piece on pre 1850 temperature datasets and also one on individual temperatures datasets from 1850. This latter is particularly interesting as it is obvious that the increasingly UHi affected urban areas (which are warming) are disguising the fact that many other places have been cooling, so the magic gas co2 is able to do both!

    This is illustrated in a study I have just done on Australia reproduced below:

    ” I have been examining climatic cycles as evidenced by thermometers, and posted the results of some of the 30 I have discovered that go back to 1701 (excluding CET)

    The cyclic effect throughout the decades is very apparent.

    The question has been posed ‘has global warming stopped?’ First of all it was never global. An examination of the temperature records from 1850 (Hadley) or 1880 (Giss) shows that the warming that is driving the rise in ‘global’ temperatures is very largely coming from Urban areas (which has a very big part in the Giss database) Other areas have been cooling-some for very many years.

    Taking into account the last decade of acknowledged (even by the Met office) cooling, and accepting that a ‘trend’ is considered to be thirty years or more, we are finding that many more individual temperature records are now showing a cooling trend. These are increasingly counterbalancing the heavily UHI affected urban areas in the ‘global’ record.

    For the benefit of our Australian friends, listed below are those places in Australia that are showing a cooling trend. (All these from Bom/Giss)

    Adelaide Airport cooling since 1881
    Brisbane-eagle Farm cooling since 1957
    Cape Otaway cooling since 1865
    Darwin airport cooling since 1905
    Dubbo cooling since 1882
    Echura (Victoria) cooling since 1881
    Willis island cooling since 1965
    Perth cooling since 1977.

    BOM seems to have a severe case of Hansenitis-in citing 2008 as Adelaides hottest ever summer they omitted to mention that 1914 was the hottest year. Other places in Australia are showing no trend or slight warming.”

    Tonyb

  8. dougie says:

    you should be recognised for your hard work on this E.M.
    i, for one, will mention you every chance i get.

    cheers
    dougie

  9. E.M.Smith says:

    dougie
    you should be recognised for your hard work on this E.M.
    i, for one, will mention you every chance i get.

    cheers
    dougie

    Thanks! Though in this case it is TonyB that has done the work and deserves the mention. That the rural areas are cooling and the urban warming, even after the shenanigans that GIStemp does to the data, is a telling point.

    Just scanning down these temperature graphs does, I think, show a very visible and very easy to spot “cherry pick” to the temperature series used by GISS, Hadley, et. al.

    But I think that is about to run into a wall of reality. The present weather reports for North America show the Cold Is Coming. In spades.

    We’ve already had a major league baseball game in Denver postponed due to cold.

    http://neighbors.denverpost.com/viewtopic.php?f=16&t=13535808&p=1083688

    The Pacific Coast is bracing for a major storm (it is very blustery and headed for icky as I type this).

    http://www.accuweather.com/news-story.asp?partner=accuweather&traveler=0&article=1

    has:

    The first major West Coast storm of the season will bring powerful winds late Monday night into Tuesday. The hardest-hit state will be California.
    South to southwest winds will average 20-40 mph with gusts in excess of 60 mph along the coast from southwestern Oregon down to Point Conception, California.
    The strong winds will lead to downed trees and power lines, which will cause some power outages.
    If you have not done so already, now is the time to make sure that you have plenty of batteries, non-perishable food, and potable water. A winter storm kit is also a great idea. Also, if you have a generator, make sure it is in proper working order,

    While
    http://www.weather.com/newscenter/nationalforecast/index.html

    says:

    A massive and very potent storm in the eastern Pacific will roll into central California like a bowling ball Tuesday bringing heavy rain and flooding with it.
    Some moisture and energy with this system was injected by former Super Typhoon Melor.
    All of this will be riding in on a jet stream that will exceed 160 mph!

    And on the CNBC business / trader news the suggested trade was “long oil and long natural gas” on the expected extreme cold snap.

    Hang onto your hats, boys and girls, it’s going to be very cold, very stormy and very much not Global Warming for the next few months in America.

    Welcome to reality…

    These folks are tracking the records as they happen:

    http://www.iceagenow.com/Record_Lows_2009.htm

    So good luck to the AGW crowd trying to sell the concept of Global Warming into a very frozen northern hemisphere…

    In a way it is kind of sad. Here they are on the cusp of achieving what they see as their finest hour in Copenhagen and the reality is that they have only 2 potential results.

    1) They fail to get anything of substance a fail outright.

    2) They do get something of substance and the whole world will either vilify them (if they believe what was done caused the next few years of intense cold) or will ridicule them (if they recognize it as just a climate shift to damned cold from normal processes such as the PDO and solar sunspot shutdown).

    I’d not want that set of choices.

    So keep your wood dry and stay warm!

  10. Tonyb says:

    EM Smith

    I have got hold of some of the Giss US weather records and just wanted to make sure I understand what they appear to be telling me. Can you please check out two datasets with your own figures (ignore the obvious imperfections in Giss and just use the ‘official’ figures)

    The two stations that will enable me to get a cross reference and see if the new source can be relied on, are two of many that appear to have been cooling since 1930.

    First up is Atlanta POS 33.65 84.42 312 metres a.s.l
    Av mean temperature is 17.28C. This started off in 1881 at 18.4 and ends in 2008 at 16.9C

    Second cross check is Amarillo POS 35.22 101.72 1098 metres a.s.l
    starts at 1892 at 14.1C peaks in 1934 at 15.7 and in 2008 was 14.2C.

    Both show a cooling trend using a linear regression.

    Are you able to confirm this information is correct as it comes from a different source to those I have previously used and I want to ensure its accuracy before carrying out any analysis.
    Thanks for your help

    tonyb

  11. E.M.Smith says:

    Tonyb: I have got hold of some of the Giss US weather records and just wanted to make sure I understand what they appear to be telling me. Can you please check out two datasets with your own figures (ignore the obvious imperfections in Giss and just use the ‘official’ figures)

    I’ll be using the non-GIStemp figures from the straight GHCN input file for these numbers. This is the GHCN input file fed into GIStemp in STEP0.

    First up is Atlanta POS 33.65 84.42 312 metres a.s.l
    Av mean temperature is 17.28C. This started off in 1881 at 18.4 and ends in 2008 at 16.9C

    From the v2.inv file we find ATLANTA at that lat/long as:

    42572219000 ATLANTA/MUN., 33.65 -84.42 315 285U 2960FLxxno-9A 1WARM FOR./FIELD C3 112

    So we have Country Code of 425, the USA. Then the full station ID is 72219000 as the “8 digit lacking modification history flag” that is 5 digits of basic station id and 3 of sub-stations (if any). The “A” means airport…

    From v2.mean for 72218000 we have:

    $ cat v2.mean | grep 42572219000
    4257221900001879 68 66 143 149 210 238 264 232 204 179 121 108
    4257221900001880 124 109 129 174 217 247 262 249 208 160 85 56
    4257221900001881 43 81 94 150 216 254 273 260 242 196 114 98
    4257221900001882 94 113 141 182 192 250 243 244 220 188 107 52
    4257221900001883 61 101 101 164 195 249 276 246 218 189 122 89
    4257221900001884 21 109 124 147 215 217 256 241 238 200 107 70
    4257221900001885 44 40 84 161 193 248 260 251 212 139 103 60
    4257221900001886 23 56 105 161 203 231 248 251 232 170 103 47
    4257221900001887 43 116 108 165 223 243 254 247 218 152 112 54
    4257221900001888 70 94 106 179 203 243 262 258 204 146 113 64
    4257221900001889 66 53 111 169 201 229 258 234 211 157 109 140
    4257221900001890 106 127 98 168 206 260 257 240 220 153 142 74
    4257221900001891 58 102 83 171 197 254 241 248 224 152 96 83
    4257221900001892 35 88 89 150 204 246 247 246 211 170 99 58
    4257221900001893 22 79 108 179 196 234 270 250 228 164 104 81
    4257221900001894 82 72 137 167 204 247 246 247 230 168 97 79
    4257221900001895 47 13 108 158 194 249 250 251 247 153 113 64
    4257221900001896 54 74 96 187 238 240 257 269 237 163 131 68
    4257221900001897 38 87 127 154 199 259 258 246 236 190 118 74
    4257221900001898 83 62 142 136 227 262 256 251 232 157 92 64
    4257221900001899 57 42 118 153 236 264 262 267 220 176 126 56
    4257221900001900 61 51 101 170 210 234 260 273 245 194 116 76
    4257221900001901 68 50 109 127 208 248 269 244 212 166 84 43
    4257221900001902 48 28 106 152 233 256 268 262 212 170 143 61
    4257221900001903 53 80 139 147 203 219 258 257 217 168 89 43
    4257221900001904 38 63 120 131 202 244 250 242 241 172 109 66
    4257221900001905 26 23 142 163 221 249 255 246 236 164 118 57
    4257221900001906 72 69 86 177 201 248 244 259 240 148 118 80
    4257221900001907 111 67 164 123 197 234 267 257 229 166 100 67
    4257221900001908 52 46 154 178 209 241 254 251 217 159 132 89
    4257221900001909 78 96 112 162 196 248 251 254 221 163 146 34
    4257221900001910 58 56 159 156 193 226 248 250 240 180 98 42
    4257221900001911 88 103 118 156 222 262 246 252 257 188 83 84
    4257221900001912 31 44 99 166 212 227 251 249 241 178 100 76
    4257221900001913 97 74 117 152 216 243 264 259 213 156 127 77
    4257221900001914 72 62 92 166 218 271 262 249 219 169 113 46
    4257221900001915 55 77 64 180 219 241 259 251 238 185 123 65
    4257221900001916 93 67 104 156 226 240 247 256 219 168 113 72
    4257221900001917 88 69 114 176 178 239 256 241 211 137 107 23
    4257221900001918 16 104 151 144 223 249 248 260 201 188 109 90
    4257221900001919 66 69 122 165 198 249 252 247 234 216 126 71
    4257221900001920 59 54 98 148 195 247 252 239 233 182 101 62
    4257221900001921 77 89 162 161 202 260 260 248 263 161 126 91
    4257221900001922 61 101 120 171 206 249 254 243 241 173 116 103
    4257221900001923 90 65 112 157 188 239 251 252 237 167 99 107
    4257221900001924 34 57 94 151 188 253 257 269 201 174 126 76
    4257221900001925 62 106 129 191 200 267 271 266 283 162 99 58
    4257221900001926 59 84 75 150 211 241 258 261 252 181 84 79
    4257221900001927 70 128 127 179 217 234 253 244 239 190 132 68
    4257221900001928 59 67 110 144 197 234 260 259 219 185 108 73
    4257221900001929 73 64 138 181 204 239 254 257 218 153 110 65
    4257221900001930 60 110 96 174 213 240 274 260 242 156 98 42
    4257221900001931 63 87 82 153 192 263 272 250 253 189 154 108
    4257221900001932 107 118 93 166 198 249 270 258 221 163 88 87
    4257221900001933 101 69 113 153 234 256 252 252 253 177 109 111
    4257221900001934 73 44 102 159 206 256 267 258 232 178 122 54
    4257221900001935 67 72 142 162 213 246 265 266 230 177 121 23
    4257221900001936 39 53 137 149 228 263 273 266 244 180 101 81
    4257221900001937 117 69 103 154 216 264 264 264 218 149 86 62
    4257221900001938 61 108 147 163 218 240 262 273 232 179 122 64
    4257221900001939 73 98 132 156 207 260 268 254 246 182 101 76
    4257221900001940 -13 56 99 152 199 252 251 258 220 180 107 87
    4257221900001941 68 43 81 176 226 252 267 267 248 209 114 88
    4257221900001942 59 46 111 177 211 254 269 254 224 176 124 66
    4257221900001943 79 82 103 163 222 273 266 272 215 161 106 69
    4257221900001944 69 102 116 157 226 268 257 251 231 171 106 51
    4257221900001945 62 93 166 176 192 256 262 257 246 163 118 36
    4257221900001946 71 89 151 178 202 241 254 251 218 168 141 91
    4257221900001947 77 37 74 176 205 242 250 262 239 191 93 74
    4257221900001948 32 90 133 187 214 258 263 252 221 159 127 89
    4257221900001949 112 109 118 149 217 241 265 250 211 189 108 77
    4257221900001950 128 102 94 147 223 247 248 242 215 185 83 44
    4257221900001951 71 84 110 154 209 251 263 273 234 182 81 84
    4257221900001952 97 94 108 162 218 272 274 251 214 149 116 63
    4257221900001953 88 90 127 151 232 252 256 260 223 179 119 64
    4257221900001954 73 101 117 186 179 256 278 281 254 181 97 65
    4257221900001955 64 87 134 188 226 227 262 267 241 166 104 63
    4257221900001956 53 107 122 161 224 249 263 269 222 182 103 124
    4257221900001957 77 126 108 179 216 254 261 264 229 147 121 82
    4257221900001958 38 32 99 169 215 250 260 261 233 167 136 59
    4257221900001959 57 87 107 172 223 248 262 270 229 182 110 79
    4257221900001960 68 66 54 176 203 249 268 261 233 187 114 48
    4257221900001961 36 97 126 136 184 226 246 242 228 162 129 64
    4257221900001962 50 101 88 143 234 242 259 248 211 171 95 46
    4257221900001963 29 38 136 172 204 235 243 254 216 184 108 19
    4257221900001964 49 46 107 159 205 254 247 244 226 149 131 81
    4257221900001965 62 66 93 174 221 221 248 252 227 157 121 71
    4257221900001966 26 62 101 159 198 230 257 241 214 152 112 63
    4257221900001967 66 54 137 183 190 225 234 233 192 152 92 87
    4257221900001968 40 35 113 161 197 243 254 261 219 165 101 46
    4257221900001969 46 59 83 171 207 252 271 245 216 169 101 54
    4257221900001970 22 66 116 180 212 237 259 262 250 184 97 86
    4257221900001971 61 68 86 160 193 251 246 248 232 193 104 113
    4257221900001972 82 59 113 163 192 223 249 254 237 162 98 91
    4257221900001973 52 60 141 142 183 242 261 252 242 182 131 68
    4257221900001974 118 77 143 162 217 225 255 248 212 163 113 68
    4257221900001975 84 84 103 154 217 241 247 255 213 174 122 63
    4257221900001976 36 108 136 165 186 232 247 244 210 134 68 43
    4257221900001977 -15 56 129 172 211 251 264 254 231 153 124 56
    4257221900001978 9 41 109 164 198 246 259 257 246 169 147 78
    4257221900001979 29 54 134 171 212 243 260 267 226 169 124 82
    4257221900001980 72 55 112 170 222 262 295 288 261 164 111 72
    4257221900011945-9999-9999-9999-9999-9999-9999 263-9999 246-9999 118-9999
    4257221900011946 70 89 151-9999-9999-9999 254 252-9999 167-9999-9999
    4257221900011947 77 37-9999 175 205 243 250-9999 239 192 93 74
    4257221900011948 30 90 131 187 215 261 267 252 219 157 128 88
    4257221900011949 110 109 118 155 221 246 266 255 216 196 107 76
    4257221900011950 131 105 101 149 227 253 253 246 219 188 87 48
    4257221900011951 74 86 117 156 212 253 262 272 233 184 86 84
    4257221900011952 97 91 109 164 222 280 278 257 219 150 116 65
    4257221900011953 84 86 126 153 229 257 260 260 224 178 115 66
    4257221900011954 73 101 117 186 179 256 279 281 255 180 97 65
    4257221900011955 65 87 134 188 226 227 262 267 241 167 104 63
    4257221900011956 53 106 122 162 225 249 263 269 223 182 103 124
    4257221900011957 78 126 109 179 215 254 261 264 230 147 121 82
    4257221900011958 38 32 99 169 215 250 260 261 233 167 136 59
    4257221900011959 57 87 107 172 223 248 262 270 229 183 110 79
    4257221900011960 68 65 54 176 203 249 269 261 233 187 115 49
    4257221900011961 36 97 126 136 184 226 247 242 228 162 129 64
    4257221900011962 50 101 88 143 234 242 259 248 211 171 95 46
    4257221900011963 29 39 136 172 204 235 244 254 216 184 108 19
    4257221900011964 49 46 107 159 205 254 247 244 226 149 131 81
    4257221900011965 63 66 93 174 221 221 248 252 227 157 121 71
    4257221900011966 26 62 101 159 198 230 257 241 214 153 112 64
    4257221900011967 67 55 137 184 190 225 234 233 192 152 92 87
    4257221900011968 40 35 113 161 197 243 254 261 220 165 101 46
    4257221900011969 46 59 83 172 206 252 271 245 216 169 100 54
    4257221900011970 22 66 116 180 212 237 259 262 250 185 97 86
    4257221900011971 60 69 87 160 193 251 246 249 232 193 104 113
    4257221900011972 82 59 113 163 191 224 249 254 237 162 98 91
    4257221900011973 52 60 141 142 184 242 261 252 242 181 132 68
    4257221900011974 118 77 143 162 217 225 255 248 212 163 114 68
    4257221900011975 84 84 103 155 217 241 247 255 213 174 122 63
    4257221900011976 36 108 136 165 186 233 247 244 210 135 68 43
    4257221900011977 -16 56 129 173 211 251 264 254 231 153 124 56
    4257221900011978 9 41 109 164 198 246 259 257 246 169 147 78
    4257221900011979 29 54 135 170 212 243 260 267 226 169 124 82
    4257221900011980 72 55 112 170 222 262 295 288 261 164 111 72
    4257221900011981 42 82 110 198 198 274 279 254 225 157 125 40
    4257221900011982 36 86 136 147 225 246 262 253 214 171 120 99
    4257221900011983 47 69 107 136 199 234 275 274 215 167 109 44
    4257221900011984 42 86 110 145 197 257 249 253 219 210 103 120
    4257221900011985 24 68 138 178 210 253 258 253 225 191 167 52
    4257221900011986 63 99 125 172 217 267 290 252 237 178 144 73
    4257221900011987 55 76 118 157 229 255 272 278 234 154 132 93
    4257221900011988 40 75 127 172 211 259 270 272 230 152 128 81
    4257221900011989 98 86 138 172 205 250 266 264 228 179 124 40
    4257221900011990 99 125 143 166 213 259 270 270 243 180 136 95
    4257221900011991 68 96 135 189 227 249 272 262 238 181 106 95
    4257221900011992 73 110 122 166 201 236 268 245 229 166 109 70
    4257221900011993 83 73 110 152 217 263 296 278 248 174 122 74
    4257221900011994 47 101 141 198 208 269 262 264 233 178 145 103
    4257221900011995 81 81 150 186 234 250-9999 270 217 167 86 61
    4257221900011996 50 84 106 161 239 262 275 264 228 175 112 92
    4257221900011997 83 106 161 145 184 220 259 247 228 164 89 62
    4257221900011998 78 84 100 150 225 262 272 256 242 189 136 98
    4257221900011999 89 97 103 184 206 239 261 278 228 167 136 84
    4257221900012000 62 103 142 147 228 256 275 264 217 178 106 28
    4257221900012001 53 106 103 175 212 236 259 262 220 161 156 103
    4257221900012002 84 75 125 181 200 244 267 267 245 183 104 66
    4257221900012003 44 81 131 164 206 236 256 262 222 175 142 62
    4257221900012004 62 67 145 167 228 242 264 248 223 194 135 69
    4257221900012005 82 93 111 156 193 240 263 264 247 180 128 55
    4257221900012006 96 67 123 188 210 251 272 272 222 164 120 100
    4257221900012007 78 73 157 156 220 264 258 298 243 188 117 103
    4257221900012008 56 90 120 162 208 265 265 256 236 166 103 92
    4257221900012009 65 86 128 159 212 265-9999-9999-9999-9999-9999-9999
    4257221900021949 109 109 118 155 222 245 266 256 217 196 108 77
    4257221900021950 132 107 103 150 228 254 253 247 220 190 88 49
    4257221900021951-9999 87 118 157 213 254 263 273 234 185 86 86
    4257221900021952 98 93 110 166 223 281 279 258 220 159 117 66
    4257221900021953 86 87 122 154 231 257 261 261 224 179 116 67
    4257221900021954 74 102 118 187 180 257 279 282 256 182 98 67
    4257221900021955 66 88 135 189 226 228 262 269 242 168 104 64
    4257221900021956 53 107 123 163 226 249 264 270 224 184 103 126
    4257221900021957 77 127 110 181 216 256 262 264 231 148 122 83
    4257221900021958 40 33 100 169 216 249 261 261 235 169 137 60
    4257221900021959 57 88 107 173 224 248 261 269 229 182 110 79
    4257221900021960 68 63 54 176 203 249 268 261 233 187 114 48
    4257221900021961 36 97 126 136 184 226 246 242 228 162 129 64
    4257221900021962 50 101 88 143 234 242 259 248 211 171 95 46
    4257221900021963 29 38 136 172 204 235 243 254 216 184 108 19
    4257221900021964 49 46 107 159 205 254 247 244 226 149 131 81
    4257221900021965 62 65 93 174 221 221 248 229 227 157 121 71
    4257221900021966 26 62 101 159 198 230 257 241 214 152 112 63
    4257221900021967 66 54 137 183 190 225 234 233 192 152 92 87
    4257221900021968 40 35 113 161 197 243 254 261 219 165 101 46
    4257221900021969 46 59 83 170 207 252 271 245 216 169 101 54
    4257221900021970 22 66 116 180 212 237 259 262 250 184 97 86
    4257221900021971 61 68 86 160 193 251 246 248 232 193 104 113
    4257221900021972 82 59 113 163 192 223 249 254 237 162 98 91
    4257221900021973 52 60 141 142 183 242 261 252 242 182 131 68
    4257221900021974 118 77 143 162 217 225 255 248 212 163 113 69
    4257221900021975 84 84 102 154 217 241 247 255 213 174 122 63
    4257221900021976 36 108 136 165-9999 232 247-9999 210 134 68 43
    4257221900021977 43 56 129 172 211 251 264 264 231 153 124 56
    4257221900021978 9 41 109 164 198 246-9999 257 246 169 147 78
    4257221900021979 29 54 134 169 212 243 260 267 226 169 124 82
    4257221900021980 72 55 112 170 222 262 295 288 260 179 111-9999
    4257221900021981 42 82 110 198 195 274 279 254 224 156 125 39
    4257221900021982 36 88 136 147 230 246 262 253 214 172 119 99
    4257221900021983 45 58 106 136 199 233 271 271 216 167 105 42
    4257221900021984 42 86 109 145 197 256 246 253 218 210 103 121
    4257221900021985 24 68 138 178 211 253 258 253 225 191 167 52
    4257221900021986 63 99 124 172 217 267 289 250 233 178 144 72
    4257221900021987 56 78 117 156 228 254 272 277 234 156 132 94
    4257221900021988 40 74 126 161 201 259 269 272 230 152 128 81
    4257221900021989 97 86 138 160 204 249 266 263 227 179 124 39
    4257221900021990 99 124 143 166 213 259 270 270 243 180 136 95
    4257221900031961 36 97 126 136 184 226 246 242 228 162 129 64
    4257221900031962 50 101 88 143 234 242 259 248 211 171 95 46
    4257221900031963 29 38 136 172 204 235 243 254 216 184 108 19
    4257221900031964 49 46 107 159 205 254 247 244 226 149 131 81
    4257221900031965 62 66 93 174 221 221 248 252 227 157 121 71
    4257221900031966 26 62 101 159 198 230 257 241 214 152 112 63
    4257221900031967 66 54 137 183 190 225 234 233 192 152 92 87
    4257221900031968 40 35 113 161 197 243 254 261 219 165 101 46
    4257221900031969 46 59 83 171 207 252 271 245 216 169 101 54
    4257221900031970 22 66 116 180 212 237 259 262 250 184 97 86
    4257221900031971 61 68 86 160 193 251 246 248 232 193 104 113
    4257221900031972 82 59 113 163 192 223 249 254 237 162 98 91
    4257221900031973 52 60 141 142 183 242 261 252 242 182 131 68
    4257221900031974 118 77 143 162 217 225 255 248 212 163 113 68
    4257221900031975 84 84 103 154 217 241 247 255 213 174 122 63
    4257221900031976 36 108 136 165 186 232 247 244 210 134 68 43
    4257221900031977 -15 56 129 172 211 251 264 254 231 153 124 56
    4257221900031978 9 41 109 164 198 246 259 257 246 169 147 78
    4257221900031979 29 54 134 171 211 243 260 267 226 169 124 82
    4257221900031980 72 55 112 170 222 262 295 288 261 164 111 72
    4257221900031981 42 82 110 198 198 274 279 254 224 157 125 39
    4257221900031982 36 86 136 147 225 246 262 253 214 171 121 99
    4257221900031983 47 69 107 136 199 233 274 274 216 167 108 44
    4257221900031984 42 86 109 145 197 257 249 253 218 210 103 121
    4257221900031985 24 68 138 178 211 253 258 253 225 191 167 52
    4257221900031986 63 99 124 172 217 267 289 252 237 178 144 73
    4257221900031987 56 78 117 156 228 254 272 277 234 156 132 94
    4257221900031988 40 74 126 171 211 259 269 272 230 152 128 81
    4257221900031989 97 86 138 171 204 249 266 263 227 179 124 39
    4257221900031990 99 124 143 166 213 259 270 270 243 180 136 95
    4257221900031991 68 96 134 188 227 248 272 262 238 181 106 95
    4257221900041984 43 87 111 146 198 258 250 254 220 211 104 122
    4257221900041985 25 69 139 179 212 254 259 255 227 192 168 53
    4257221900041986 64 99 126 172 218 268 291 253 238 179 144 74
    4257221900041987 56 77 119 159 230 256 273 278 236 154 133 95
    4257221900041988 42 76 129 173 213 261 271 273 231 153 129 82
    4257221900041989 99 88 139 172 206 251 267 264 228 180 126 41
    4257221900041990 101 126 143 168 214 260 271 271 244 181 137 96
    4257221900041991 69 97 137 190 228 249 273 263 238 182 107 96
    4257221900041992 74 111 124 167 202 237 269 246 230 167 109 71
    4257221900041993 85 74 111 153 218 264 298 279 249 174 123 75

    Ok, now the fun “Magic Decoder Ring” time… First up is Country Code and Station ID, that was our search key into the Linux utility “grep”: 42572219000

    Then there is a single digit, then a 4 digit year. So the last line of the data above has a “4” and then “1993”. This is the “modification flag” and the year of the data. This is then followed by 12 monthly averages of the daily average of high and low in 1/10 C. So the last number for the last line says that you had 7.5 C as the average of December daily “high and low averages”. Hope that was clear…

    The “modification flag” says that the data had a different modification history. My mediocre understanding of it is that It might be a different kind of equipment was installed so a different “adjustment” was applied, or it might be 2 readings per day at different times so a different TOBS was applied, or… It basically just says that the same place had more than one “history of modification”. So you can look back up the list and find 3 readings for 1990. You get to decide how to spice these series together. The method GIStemp uses is pretty bad, so anything you do will be at least that good. You could spend some time finding out exactly what the “modification history” flag means for the site and make your “splice” better, I suppose…

    Oh, and -9999 is the “missing data” flag, and yes, you get to decide how to fill in, interpolate, or ignore it. GIStemp, again, does a pretty mediocre job of “guessing” so I’d trust just about anything else just about as much or more.

    If you want the USHCN data in degrees F, I have that as well.

    Second cross check is Amarillo POS 35.22 101.72 1098 metres a.s.l
    starts at 1892 at 14.1C peaks in 1934 at 15.7 and in 2008 was 14.2C.

    Here is the v2.inv station record:

    42572363000 AMARILLO/INTL 35.22 -101.72 1098 1086U 158FLxxno-9A 3WARM GRASS/SHRUBC3 31

    Again, the “A” in “-9A” says it is an airport.

    The data:

    $ grep 42572363000 v2.mean
    4257236300001892 4 55 54 136 173 232 249 236 223 134 77 2
    4257236300001893 42 6 83 132 167 237 254 228 213 141 59 43
    4257236300001894 24 -16 80 144 183 223 249 217 193 156 89 39
    4257236300001895 8 -36 78 136 176 206 225 234 212 117 57 17
    4257236300001896 42 46 61 154 208 244 236 250 191 123 60 63
    4257236300001897 -23 42 73 129 174 224 247 231 207 147 82 11
    4257236300001898 17 55 78 133 176 215 234 233 194 136 56 -23
    4257236300001899 10 -27 72 131 183 213 232 256 201 147 90 22
    4257236300001900 47 22 84 112 169 225 231 239 203 152 89 40
    4257236300001901 44 10 72 121 164 229 251 247 198 152 90 28
    4257236300001902 19 35 77 138 191 228 239 249 189 157 83 24
    4257236300001903 37 -15 63 131 167 188 257 247 197 145 83 39
    4257236300001904 18 73 111 130 180 212 239 236 207 143 89 28
    4257236300001905 -5 -44 102 111 173 231 231 254 213 126 83 12
    4257236300001906 38 36 43 132 180 218 231 230 206 119 54 77
    4257236300001907 58 68 138 114 150 220 244 246 216 142 63 36
    4257236300001908 38 47 112 132 176 229 227 236 198 137 74 51
    4257236300001909 49 54 72 122 170 231 255 256 207 143 102 -5
    4257236300001910 43 20 136 146 164 243 263 246 232 157 94 44
    4257236300001911 76 37 108 138 187 250 241 243 240 139 53 -10
    4257236300001912 -3 20 41 126 192 213 262 247 181 146 83 24
    4257236300001913 17 -1 63 134 201 212 259 267 182 129 102 7
    4257236300001914 74 34 85 133 173 246 254 243 227 144 102 -8
    4257236300001915 11 52 29 139 164 224 237 219 204 150 97 46
    4257236300001916 18 64 121 116 194 237 261 248 199 140 69 26
    4257236300001917 26 48 79 127 144 231 263 233 208 132 104 23
    4257236300001918 -6 67 114 118 198 252 257 256 186 155 59 2
    4257236300001919 -17 33 80 125 166 205 246 253 218 143 59 28
    4257236300001920 18 47 84 106 178 224 256 220 216 159 60 39
    4257236300001921 50 53 111 128 186 212 238 247 231 171 106 64
    4257236300001922 11 49 74 127 183 230 260 276 228 158 84 54
    4257236300001923 79 23 60 133 177 224 257 251 210 112 74 11
    4257236300001924 17 48 38 128 159 256 241 257 198 156 104 0
    4257236300001925 22 81 119 162 183 251 258 238 203 101 79 31
    4257236300001926 22 80 64 106 180 228 240 251 208 164 83 28
    4257236300001927 51 70 89 159 217 232 257 241 201 174 114 14
    4257236300001928 52 41 100 124 186 223 254 237 213 166 68 42
    4257236300001929 24 -7 94 156 166 239 260 262 201 157 30 52
    4257236300001930 -33 97 71 174 181 242 264 266 232 138 90 31
    4257236300001931 44 72 56 121 163 254 263 245 249 174 83 46
    4257236300001932 18 95 67 150 196 226 268 256 205 146 84 -6
    4257236300001933 66 27 104 132 192 262 279 251 243 169 102 86
    4257236300001934 50 58 95 160 205 264 289 274 212 189 107 56
    4257236300001935 68 63 117 138 161 236 267 262 198 166 79 42
    4257236300001936 20 8 112 144 194 253 270 277 207 136 81 61
    4257236300001937 -6 47 68 142 197 236 272 278 228 166 87 44
    4257236300001938 49 61 116 137 188 239 262 276 231 182 76 55
    4257236300001939 55 19 108 146 198 248 270 254 243 172 91 67
    4257236300001940 -14 51 104 139 193 230 278 249 223 182 67 58
    4257236300001941 41 49 62 138 188 224 250 253 217 158 102 57
    4257236300001942 29 37 74 143 185 246 268 251 202 151 111 48
    4257236300001943 43 77 73 169 178 256 266 286 216 153 84 17
    4257236300001944 31 52 76 124 194 252 260 271 216 162 99 32
    4257236300001945 46 58 111 116 199 236 261 262 221 161 109 39
    4257236300001946 28 69 113 181 179 248 279 265 213 164 87 77
    4257236300001947 33 28 64 124 181 239 266 271 244 193 64 33
    4257236300001948 -6 11 46 163 188 239 261 246 214 147 58 48
    4257236300001949 -29 40 82 121 182 231 257 239 201 143 116 36
    4257236300001950 41 77 83 133 188 243 235 232 194 182 78 43
    4257236300001951 16 50 71 121 183 223 273 265 211 150 61 36
    4257236300001952 53 59 72 124 183 268 258 276 211 144 58 28
    4257236300001953 71 46 116 126 186 285 268 251 227 158 88 24
    4257236300001954 43 92 76 163 162 248 284 266 241 162 103 57
    4257236300001955 32 31 87 152 187 221 252 248 213 154 63 51
    4257236300001956 31 18 86 121 212 256 253 251 228 171 65 56
    4257236300001957 26 80 72 113 159 224 273 251 191 133 48 59
    4257236300001958 27 32 28 106 193 244 256 257 214 145 82 29
    4257236300001959 2 40 77 128 189 237 242 258 212 127 58 46
    4257236300001960 -3 -3 57 146 184 235 242 246 208 148 84 16
    4257236300001961 14 41 92 134 196 233 247 251 198 153 52 25
    4257236300001962 -2 75 82 140 221 224 256 257 213 164 88 49
    4257236300001963 -17 54 102 166 206 234 274 254 224 190 97 3
    4257236300001964 29 1 71 144 205 247 277 256 209 159 86 39
    4257236300001965 50 31 36 152 194 223 259 249 203 153 114 61
    4257236300001966 -23 12 100 125 189 234 283 231 203 143 109 19
    4257236300001967 48 48 119 159 174 232 251 239 202 161 81 19
    4257236300001968 35 33 93 129 174 238 250 250 206 167 74 27
    4257236300001969 52 47 38 155 196 227 281 268 217 125 82 36
    4257236300001970 7 62 54 133 204 236 271 262 216 122 81 63
    4257236300001971 32 38 91 132 183 244 249 223 195 140 76 31
    4257236300001972 21 49 109 153 173 231 236 233 207 144 27 5
    4257236300001973 8 42 84 102 172 239 256 260 204 167 104 37
    4257236300001974 17 57 117 157 219 239 263 231 172 150 76 23
    4257236300001975 28 18 72 127 181 229 242 248 188 162 77 47
    4257236300001976 27 88 81 138 157 223 238 239 192 99 36 31
    4257236300001977 -11 66 93 143 196 256 271 253 235 159 86 44
    4257236300001978 -17 -11 84 163 176 241 271 245 215 152 76 4
    4257236300001979 -39 44 81 125 168 216 251 231 210 156 47 38
    4257236300001980 16 32 66 113 166 257 283 258 215 137 59 52
    4257236300001981 33 57 93 173 187 258 274 236 206 137 95 45
    4257236300001982 29 21 85 122 174 223 260 259 217 141 76 23
    4257236300001983 8 23 74 105 158 213 267 272 231 158 89 -41
    4257236300001984 -2 46 67 110 194 237 242 240 187 138 83 48
    4257236300001985 -3 29 98 156 201 240 267 264 205 138 64 11
    4257236300001986 59 47 115 152 184 229 270 236 204 132 60 29
    4257236300001987 11 56 70 127 181 223 251 241 198 144 74 15
    4257236300001988 1 35 69 123 176 230 243 247 198 150 87 36
    4257236300001989 48 2 107 155 196 208 245 244 191 158 88 -4
    4257236300001990 40 49 83 133 176 274 247 246 223 143 98 7
    4257236300001991 3 77 98 146 203 237 247 246 195 148 51 39
    4257236300001992 33 72 104 148 174 216 248 231 212 156 41 12
    4257236300001993 7 22 78 124 181 233 261 239 200 125 49 37
    4257236300001994 27 31 94 128 184 263 253 242 206 145 81 53
    4257236300001995 39 67 87 123 164 214 250 256 200 147 92 36
    4257236300001996 17 64 70 139 225 242 250 236 186 145 75 42
    4257236300001997 17 36 103 95 175 220 253 239 220 145 61 9
    4257236300001998 45 50 64 120 208 253 278 256 245 156 95 34
    4257236300001999 45 86 81 125 170 223 250 256 192 145 114 31
    4257236300002000 42 84 98 145 208 222 267 278 231 153 45 6
    4257236300002001 14 50 72 161 184 247 287 258 214 158 103 45
    4257236300002002 42 42 75 156 192 256 261 256 206 117 67 20
    4257236300002003 45 31 89 150 195 208 267 261 198 167 86 50
    4257236300002004 42 31 114 128 208 228 245 231 211 149 61 44
    4257236300002005 43 62 76 127 178 237 255 243 225 141 89 25
    4257236300002006 60 40 90 160 207 248 257 238 180 144 93 34
    4257236300002007 -3 41 119 116 181 224 250 265 224 166 90 33
    4257236300002008 33 56 89 137 191 252 253 237 194 145 90 80
    4257236300002009 36 76 106 133 177 242-9999-9999-9999-9999-9999-9999
    4257236300011892 4 55 54 136 173 232 249 236 223 134 77 2
    4257236300011893 42 6 83 132 167 237 254 228 213 141 59 43
    4257236300011894 24 -17 80 144 183 223 249 217 193 156 89 39
    4257236300011895 8 -37 78 136 176 206 225 234 212 117 57 17
    4257236300011896 42 46 61 154 208 244 236 250 191 123 60 63
    4257236300011897 -24 42 73 129 174 224 247 231 207 147 82 11
    4257236300011898 17 55 78 133 176 215 234 233 194 136 56 -24
    4257236300011899 10 -28 72 131 183 213 232 256 201 147 90 22
    4257236300011900 47 22 84 112 169 225 231 239 203 152 89 40
    4257236300011901 44 10 72 121 164 229 251 247 198 152 90 28
    4257236300011902 19 35 77 138 191 228 239 249 189 157 83 24
    4257236300011903 37 -16 63 131 167 188 257 247 197 145 83 39
    4257236300011904 18 73 111 130 180 212 239 236 207 143 89 28
    4257236300011905 -6 -45 102 111 173 231 231 254 213 126 83 12
    4257236300011906 38 36 43 132 180 218 231 230 206 119 54 77
    4257236300011907 58 68 138 114 150 220 244 246 216 142 63 36
    4257236300011908 38 47 112 132 176 229 227 236 198 137 74 51
    4257236300011909 49 54 72 122 170 231 255 256 207 143 102 -6
    4257236300011910 43 20 136 146 164 243 263 246 232 157 94 44
    4257236300011911 76 37 108 138 187 250 241 243 240 139 53 -11
    4257236300011912 -4 20 41 126 192 213 262 247 181 146 83 24
    4257236300011913 17 -2 63 134 201 212 259 267 182 129 102 7
    4257236300011914 74 34 85 133 173 246 254 243 227 144 102 -9
    4257236300011915 11 52 29 139 164 224 237 219 204 150 97 46
    4257236300011916 18 64 121 116 194 237 261 248 199 140 69 26
    4257236300011917 26 48 79 127 144 231 263 233 208 132 104 23
    4257236300011918 -7 67 114 118 198 252 257 256 186 155 59 2
    4257236300011919 -18 33 80 125 166 205 246 253 218 143 59 28
    4257236300011920 18 47 84 106 178 224 256 220 216 159 60 39
    4257236300011921 50 53 111 128 186 212 238 247 231 171 106 64
    4257236300011922 11 49 74 127 183 230 260 276 228 158 84 54
    4257236300011923 79 23 60 133 177 224 257 251 210 112 74 11
    4257236300011924 17 48 38 128 159 256 241 257 198 156 104 -1
    4257236300011925 22 81 119 162 183 251 258 238 203 101 79 31
    4257236300011926 22 80 64 106 180 228 240 251 208 164 83 28
    4257236300011927 51 70 89 159 217 232 257 241 201 174 114 14
    4257236300011928 52 41 100 124 186 223 254 237 213 166 68 42
    4257236300011929 24 -8 94 156 166 239 260 262 201 157 30 52
    4257236300011930 -34 97 71 174 181 242 264 266 232 138 90 31
    4257236300011931 44 72 56 121 163 254 263 245 249 174 83 46
    4257236300011932 18 95 67 150 196 226 268 256 205 146 84 -7
    4257236300011933 66 27 104 132 192 262 279 251 243 169 102 86
    4257236300011934 50 58 95 160 205 264 289 274 212 189 107 56
    4257236300011935 68 63 117 138 161 236 267 262 198 166 79 42
    4257236300011936 20 8 112 144 194 253 270 277 207 136 81 61
    4257236300011937 -7 47 68 142 197 236 272 278 228 166 87 44
    4257236300011938 49 61 116 137 188 239 262 276 231 182 76 55
    4257236300011939 55 19 108 146 198 248 270 254 243 172 91 67
    4257236300011940 -15 51 104 139 193 230 278 249 223 182 67 58
    4257236300011941 41 49 62 138 188 224 250 253 217 158 102 57
    4257236300011942 29 37 74 143 185 246 268 251 202 151 111 48
    4257236300011943 43 77 73 169 178 256 266 286 216 153 84 17
    4257236300011944 31 52 76 124 194 252 260 271 216 162 99 32
    4257236300011945 46 58 111 116 199 236 261 262 221 161 109 39
    4257236300011946 28 69 113 181 179 248 279 265 213 164 87 77
    4257236300011947 33 28 64 124 181 239 266 271 244 193 64 49
    4257236300011948 8 21 53 172 198 249 270 255 225 160 73 63
    4257236300011949 -16 50 89 130 192 242 266 248 212 156 131 51
    4257236300011950 54 87 91 142 198 254 244 241 205 194 93 58
    4257236300011951 60 50 71 121 183 223 273 265 211 150 61 36
    4257236300011952 53 59-9999-9999-9999 268 258 276 211 154 58 28
    4257236300011953 71 46 116 125 186 285 268 250 227 158 88 24
    4257236300011954 43 92 76 163 162 248 284 266 241 162 103 57
    4257236300011955 32 31 87 152 187 220 252 248 213 154 63 51
    4257236300011956 30 18 86 121 212 256 253 251 228 170 65 56
    4257236300011957 26 80 72 113 159 224 273 251 191 133 48 59
    4257236300011958 27 32 28 106 193 244 256 256 214 145 82 29
    4257236300011959 2 40 77 128 189 237 242 258 212 127 58 46
    4257236300011960 -3 -3 57 146 184 235 242 246 208 148 84 16
    4257236300011961 14 41 92 134 196 233 247 251 198 153 52 23
    4257236300011962 -2 75 82 140 221 224 256 257 213 164 88 49
    4257236300011963 -17 54 102 166 206 234 274 254 224 190 97 3
    4257236300011964 29 1 71 144 205 247 277 256 209 159 86 39
    4257236300011965 50 31 36 152 194 223 259-9999 203 153 114 61
    4257236300011966 -23 12 100 125 189 234 283 231 203 143 109 19
    4257236300011967 48 48 119 159 174 232 251 239 202 161 81 19
    4257236300011968 35 33 93 129 174 238 250 250 206 167 74 27
    4257236300011969 52 47 38 155 196 227 281 268 217 125 82 36
    4257236300011970 7 62 54 133 204 236 271 262 216 122 81 63
    4257236300011971 32 38 91 132 183 244 249 223 195 140 76 31
    4257236300011972 21 49 109 153 173 231 236 233 207 144 27 5
    4257236300011973 8 42 84 102 172 252 256 260 204 167 103 37
    4257236300011974 17 57 117 157 219 239 263 231 172 150 76 23
    4257236300011975 28 18 72 127 181 229 242 248 188 162 77 47
    4257236300011976 27 88 81 138 157 223 238 239 191 99 36 31
    4257236300011977 -11 66 93 143 196 256 271 253 235 159 86 44
    4257236300011981 33 57 93 173 187 258 274 236 206 137 95 45
    4257236300011982 28 22 85 121 174 223 260 259 217 141 76 23
    4257236300011983 8 23 74 105 157 213 267 272 231 158 88 -41
    4257236300011991-9999-9999-9999-9999-9999-9999 247 246 195 147 51 39
    4257236300021984 -1 48 68 111 195 238 243 242 188 139 85 48
    4257236300021985 -2 31 99 157 202 241 268 266 206 140 65 12
    4257236300021986 59 48 116 154 186 231 271 237 205 133 60 29
    4257236300021987 12 58 72 128 182 224 253 242 199 145 75 16
    4257236300021988 2 37 70 124 177 231 244 248 199 152 88 37
    4257236300021989 49 3 108 157 197 209 247 246 192 159 89 -3
    4257236300021990 41 50 84 134 176 275 248 247 224 144 99 8
    4257236300021991 5 78 99 147 203 238 248 247 197 149 52 41
    4257236300021992 34 73 105 149 175 216 249 232 214 157 43 13
    4257236300021993 7 23 80 126 182 235 263 239 201 127 51 38

    Which has three “modification history” types, but the base case of “0” looks to be complete while “1” and “2” especially are shorter.

    Both show a cooling trend using a linear regression.

    Are you able to confirm this information is correct as it comes from a different source to those I have previously used and I want to ensure its accuracy before carrying out any analysis.
    Thanks for your help

    tonyb

    3 questions:

    1) Is this GHCN data a “different source” from yours?

    2) Do you want the USHCN data instead?

    3) Does this data let you do the regression test / confirmation you wanted to do?

  12. E.M.Smith says:

    Oh, and you can get the whole GHCN data set (and even choose just MAX, just MIN, adjusted, not adjusted, etc.) at:

    ftp://ftp.ncdc.noaa.gov/pub/data/ghcn/v2

    USHCN data are at:

    ftp://ftp.ncdc.noaa.gov/pub/data/ushcn

    and I detail where to get the data along with where to get more guidance in:

    https://chiefio.wordpress.com/2009/02/24/ghcn-global-historical-climate-network/

    for anyone wanting to do more of this at home ;-)

  13. Tonyb says:

    EM Smith

    I tried out rimfrost.no to obtain US data as it is very easy to use (I suspect nothing can be that easy involving Giss temperatures which is why I wanted to double check)

    Go to site and let it warm up-a globe will appear.

    1 Country/region insert required country
    2 Weather station use drop down menu and select
    3 Leave default as Jan and Warmest
    4 On drop down menu select years/trend
    5 This will create a linear regression
    6 Useful to do this from start to end of record
    7 Eyeball the graph and select a trend by inserting the relevant year in first box-leave 2008 as is
    8 Use years/trend function again to get new plot

    By playing with the year, if there is a cooling trend it can be seen and graphed. It must be at least 30 years to count as a proper climatic trend and must finish in 2008 so it is a current trend.

    I have used this a couple of times and cross checked the information elsewhere and it has worked fine.

    However the US records seem to be different. In this case the trends are still broadly right but the figures are not the same. However the two databases are close enough for me to suspect both are creating mean averages. (I have already asked rimfrost to confirm that it is mean average that is being graphed but have had no reply yet.)

    Atlanta Giss Rimfrost
    1881 16.84 17.28
    2008 16.825 16.9

    Amarillo Giss Rimfrost
    1892 13.125 14.1
    1934 16.325 15.7
    2008 14.64 14.2

    So Atlanta still trends lower -although with Giss only fractionally so. However Amarillo has become cooler at the start, much warmer in 1934 and slightly warmer in 2008, so the cooling trend from 1934 will presumably still be evident. I think this might have something to do with Steve Mcintyre tellng Hansen that he had his algorithm wrong for the 30’s -this was supposedly in the States only (although probably elsewhere as well!)

    Here is the dilemma- which one is more accurate (if that epithet can be applied to the sack of broken bones that are the giss figures) Rimfrost is much easier to use and has all the global figures ready to hand but I don’t want to have to cross check everything.

    Advice please on the quality of Rimfrosts figures which are clearly marked Giss.

    tonyb

  14. E.M.Smith says:

    If it is marked “GISS” it has all the cookage and fudge of GIStemp in it. I would allow 1/2C for the basic warming and about another 1/2 C for the homogenize step (the first one is measured, the second is an estimate from single stations).

    Best, IMHO, is to just use the “straight” GHCN. As far as I can tell, GIStemp adds nothing of merit to it and adds a great deal of trash.

    You know, the more I ponder it, the more I think my next step really ought to be “SmithTemp”. A de-junked GIStemp code package…

  15. Tonyb says:

    EM Smith

    I agree, but can you easily do Smith Temp for the whole world?

    If you can, great. If not why don’t you cherry pick your own individual locations to demonstrate there is a cooling trend in many places, so by definition we can’t have ‘global warmimg’

    USA is a very good place to start but will not be seen as necessarily representative of the whole world.

    I have now used Rimfrost to check out a number of other locations. Whilst accepting it is not perfect it does however seem to indicate the cooling trends evident in some of the giss records you quoted above, albeit at first sight the Giss ones do not seem so pronounced.

    Ones to check out for a cooling trend include the following. With Group 1 have double checked that the record reaches to 2009. The second group show the cooling trend as well but I haven’t checked if they are still reporting. I suspect not all are ‘official’ (perhaps they are not official because they are cooling?)

    Group 1
    Australia
    The ones I quoted in my earlier post.
    Plus
    Antarctica Hallery Sanae station Syowa Vostok island

    USA Abilene Amarillo Anchorage Atlanta-Ggeorgia
    Plus everything beyond ‘A’

    Group 2
    Atlantic ocean Halifax Port Stanley

    Brazil Curabo Curitibic Quixeramobim

    Canada Clyde Dawson Edmonton Frobisher bay
    Sable island Vancouver

    Chile-getting cooler
    Colombia big spike-suspect data
    Ethiopia cooling

    Finland Kuusamo-east Oulu Sodankyla
    Germany Bremen Hamburg Hannover-static
    Great Britain Aberporth
    Greenland Nuuk-Godthap Nuuk-Nasa Upernavik

    There will be many more but as you can see I started at ‘A’ and got to ‘G’ in the world list. Hope this is a useful pointer.

    As I say, if it is just as easy to sort out Smith temp for the whole world that would be preferable. If not, and you want some more pointers of individual locations to check out, I will go through Rimfrost and finish off the list.

    Tonyb

  16. E.M.Smith says:

    Tonyb
    EM Smith

    I agree, but can you easily do Smith Temp for the whole world?

    If you can, great. If not why don’t you cherry pick your own individual locations to demonstrate there is a cooling trend in many places, so by definition we can’t have ‘global warmimg’

    Well, for some really silly reasons, it is actually easier to do “SmithTemp” for the whole world than for a subset.

    Basically, GIStemp has a peculiar bad programming style issue where several files that it “matches” to each other must have exactly the same station data in them. Delete something, and it breaks. So doing a subset of the stations ‘has issues’. I could work my way around this, but it would be modestly hard and take time.

    At the same time, it is heavily parameterized for specific things (IMHO, a result of “hand tuning” to select for those that show the most “warming”) and it is exactly those things that I would set to more rational values, so it is easy to do.

    For example, the radius used for UHI “correction” and for various other “interpolations” is a parameter. All it takes to turn it from the (fairly bogus) 1000 km in one section and 1200 km in another section is to change the number to something more rational like 100 km. About 5 minutes if I was slow about it.

    Similarly, there is a “way to few” selection of 6 latitude bands for the whole planet in one of the steps. But it, too, is a parameter. Setting it to 9 would be far more reasonable. Then a thermometer would have to be within 20 degrees to be in that band, instead of 30. IMHO, 18 bands would be even better, then the thermometers in a 10 degree band could be used to adjust each other, but at least a thermometer in Anchorage Alaska would not be considered to be in the same climate band as one in Baja California!

    In an odd sort of way, the “hand tuning” of GIStemp makes it easy to “de tune” since it was left as parameters…

    I think I could have a “global SmithTemp” running in about a week; maybe half that if I really tried…

  17. Tonyb says:

    E M Smith

    I posted a reply to you yesterday but it doesn’t semn to have stuck.

    I had a reply from Rimfrost and basically I would be happy with their data as it has been well researched and despite using official information still shows many cooling trends.

    It strikes me as a useful cross reference if you do your Smith temp (registered trade mark) looking for cooling trends whist I do the same with Rimfrost.

    We will then have ‘official and ‘unofficial’ data and can see if they agree with each other.
    It would be good if you ‘really try’ so we have some sort of information in a few days :)

    Tonyb

  18. E.M.Smith says:

    I looked in both the spam and moderation queues. Nothing there. Everything in the “approved” bucket. So your prior response was not lost in the bowels of wordpress queues.

    I’ll take a crack at “SmithTemp – Phase One” Real Soon Now ;-)

    I might even be able to whack together a quick “declining trend detector program” (that would be a lot more fun than mucking about in GIStemp … speaking of lower bowels…)

    In fact, the more I think about it, just doing a quick “GHCN data only, before GIStemp mucks it up, find declining trends” program might even be better. Those same stations could then be compared with the on line official GIStemp charts and, if rising in GIStemp, well… The only way that falling real data can be turned into rising masticated data is via manipulation, computerized or otherwise.

    Dang it, now you’ve got me all spun up to start another project 8-}

  19. B.T.Builder says:

    “The real question we ought to be asking is “Why has this interglacial been so stable and hospitable to life when prior interglacials were pop-and-drop spikes? And how do we keep this one from dropping off a cliff like the last ones?” ”

    I think we’re plopping now. This is it.

  20. Tonyb says:

    E M Smith

    I have been looking through my own figures and find no less than 150 giss temperature data sets that show cooling. It is obvious that the globe is not warming equally, large parts are cooling and have been for years. This trend is apparent if you take 30 years or more back from 2009 (not the giss 1961-1990) i.e. long enough to be a climatic trend.

    Alternatively, by looking at the REALLY long data sets ( I now have 40 back as far as 1660 with a lage grouping starting in the early 1700’s) the cooling trend to the modern day kicks in in at either around 1935 or around 1880 for a significant number of individual places.

    The cyclic variability (warm cold warm cold) comes over really clearly if you look far enough back in time. These actual giss temperature falls are even without factoring in UHI, which is very notable in many urban databases (which also tend to be the longest).

    The cooling is occurring at one or more temperature data sets in some 32 countries, with some 10 still to check. The temperature increase in other places has disguised this counter cyclical cooling trend but will become more noticeable as more data sets start to fall into the real 30 year climate trend window going back from 2009.

    See what your figures say. No wonder they stopped using the term ‘global warming’ and started using Climate change.

    Tonyb

  21. E.M.Smith says:

    Yes, it is cooling. That is what the 500,000 year records and the 40,000 year records show. But it has “ripples” on top of it. The GIStemp / AGW folks are just fooling themselves with too short a time base.

    I wonder some times if they are really that stupid or if it is malice. I remember Hanlon’s razor “Never attribute to malice that which is adequately explained by stupidity”; but then I find my self asking “Can they really be that stupid?”. The answer is increasingly “I don’t think so, Tim.”

    I think we are reaching a limit to Hanlon’s razor…

    The clear “cherry pick” in the 1850 to 1880 start of history for both GIStemp and Hadley pushes the envelope of “accidental” and “stupid” a bit to far, IMHO. The ignoring of early 1700’s data puts an exclamation point on it. And the 12,000 year drift lower is the freight train I have trouble ignoring.

    So here we are. Sun on vacation. PDO in the cold phase. Long term down trend. 1500 or so years from the last Bond Event OMG cold spike. To quote somebody or other:

    “I’ve got a bad feeling about this.”

    (I think it was Harrison Ford in one of the Indiana Jones movies… though I’d heard it before then too… but they made it iconic, IIRC).

    FWIW, you’ve got me motivated enough again to jump into GIStemp and do the SmithTemp variation. Oh, and the “cold trending stations” filter program too. Probably about a week to get it done (if no “life event” blows up in my face :-0 …)

    I think I’ve reached a “stability point” with the cats (2 canned foods that work and a decent ‘bead’ on making home canned ‘chicken pate’ for them) so that I can return to GIStemp and not feel guilty about it. (That is, the cats have things they can eat that keep them healthy and I can just feed them that for a week or three. I’m sure they will appreciate it!)

    There is another “life event” that will get a posting in about a week when it is “all done” that is also resolving to the “good side”, but that mystery will need to wait a week or two for expansion. Lets just say that a decade of WTF is finally turning my way. And no, I’ll not say more than that until the final picture shows everything is fine…

    With that, back at the issue of “Climate Change”:

    I find the definition of “Climate” as the 30 year average of weather increasingly galling. We have known cycles of 60 years (PDO), 176-210 years (solar), and 1470 years (bond events). To try to claim that 30 years is anything other than “a short weather average” is just bogus. Frankly, I think it rises to the level of “flat out lie”. No smiley. No moderation. No hesitation. No reduction or diminution of any sort. Bald Faced Lie. If you wish to speak of “climate change” then you must use at a minimum a 3000 year baseline. (NO “IMHO”. This is a mathematical fact based on the Bond Event cycle being somewhere in the 1500 +/- 250 range, you simply MUST have a base line longer than the longest end of that cycle. So 1750 plus some range and 2 cycles is better. Call it 2000 year minimum, 3000 years better.)

    On that time scale, climate change is clearly and unequivocally happening. And it is a steady undeniable drift toward colder and the eventual plummet into an ice age.

    The only good news is that we have such short lives that the plummet is not perceptible to just about everyone living on the planet. Unfortunately, that is the bad news too …

  22. Ellie in Belfast says:

    If you wish to speak of “climate change” then you must use at a minimum a 3000 year baseline.

    Absolutely! The 120-160 years used currently is absurd.

  23. Tonyb says:

    EM Smith and Ellie

    Glad to see Ellie here, I liked her work on Belfast Temperatures.

    EM Smith can you fix the Uppsalla link-it appears to be broken.

    I have now been through the remaining records. You can add in 5 cooling data bases from China (from the 1920-/30’s

    Four from Russia cooling from 1880 or 1930.

    The US gets the prize-probably because it has the most stations. No Less than 36 are showing cooling on a significant trend scale of 30 years or more. Again these tend to group from 1880’s and 1930’s.

    I think the 1880s are significant as that seems to be when many of your records started. There are very few before that and most are from what are now conurbations so we don’t see the full sweep of climatic history we get in some of the European data bases.

    The arctic ones are particularly interesting. Many show warming since ‘satellite records began in 1979’. However what they don’t show is that if you step back a bit there has been a cooling trend since the 1930’s when many records were set.
    (Glad the cat is better!)

    Tonyb

  24. E.M.Smith says:

    @Ellie: Even worse, we use “30 year climate” in a lot of places.

    @TonyB: The links worked for me (all three of them!). The embedded graph, the “full sized image” and the redundant “link with the whole URL in it”. I removed the last one as a redundancy…

    The “climate history” in thermometers is astoundingly short. When you look at the migration of thermometers south and to the new world over time, it is even worse. We’ve had basically no coverage in depth of the southern half of the planet until about the 1950 range. See:

    https://chiefio.wordpress.com/2009/08/17/thermometer-years-by-latitude-warm-globe/

    DecadeLat: 1949 0 20 261 259 116 407 2412 887 37

    So in 1949 we had 20 thermometers in the entire southernmost 40 degrees of latitude (this is in 20 degree bands) excluding any in antarctica (this is the GHCN data set and you need to explicitly add in antarctic data). And those thermometers were not very evenly distributed over the space…

    Per the cat: Thanks! The cat is doing much better now, so this not only gives me more time, it removes a traumatic distraction. I still have some ‘cat food cooking’ to do, but now that I have two working canned foods, the pressure is off. (Said cat is in my lap as I type this… can’t sit down anymore without the cat installing herself. Much more pleasant now, too, since she is no longer sneezing “snork” all over…)

  25. Ellie in Belfast says:

    @Tonyb: Thanks for the endorsement. I too am glad you’ve found this site and that you, also, are in this for the long haul.

    I have been looking at the records as well – mostly the sets available from GHCN, and also see plenty of cooling. I’ve been comparing with the GISS homogenised set and see cooling changed to warming thanks to the homogenisation step. There are enough of these changes to be worrying, or at least to make my blood boil.

    Documenting and quantifying this seems necessary, but the question then is whether to put it in a form for publication (in the scientific literature – if it will be accepted) or just get it out into the public notice via the web where it would be very timely.

  26. Ellie in Belfast says:

    @EM Smith: re Smithtemp. I like it – a lot. I mean the whole idea of it. UHI probably has a huge influence (never mind AHI), but the homogenisation step in GIStemp is far from the fix it is claimed to be (see recent email). I would be very curious to see what the overall trend for the combined dataset would be without the homogenisation step, and, better, with a minimum of ‘messing around’ with the data.

    Here are the problems with the homogenisation step as I see them (and I know you are well aware of most of these).

    1. IMHO the estimation and correction of UHI too little for most UHIs.
    2. The GISS population estimate for urban areas is too low and seems anyway to have little effect on the UHI correction (homogenisation).
    3. Homogenisation relies on rural sites that are often:
    a) airports (https://chiefio.wordpress.com/2009/08/23/gistemp-fixes-uhi-using-airports-as-rural/)
    b) up to 1200km away
    c) climatically different from the adjusted site
    d) not as long as the urban site, which truncates the
    comparison, resulting on loss of often very old and
    good quality data.
    4. Homogenisation artifically cools some older data instead of warming it, which increases the warming trend.
    5. The size of the task, 10,000+ thermometers, means that the attention to detail is lost.

    In short the homogenisation step in GIStemp is meant to correct for UHI; I think it may actually increase the warming trend so what you are planning (Smithtemp) has been on my wishlist for some time.

    The quality value of temperature data for climate use is at the level of the individual data set and, despite GISS’s talk of quality control, homogenisation misses that point. I was taught in geography thirty-odd years ago that the climate in one valley (for example) can differ substantially from that of another only tens of miles away, depending on aspect, soil type and vegetation cover, rainshadow and many other things. Wine producers know this; so do farmers; and ecologists. GISS’s aspiration of producing a global temperature is OK in theory but this is one idea that should have been remained in someone’s imagination. Their execution of it sucks.

  27. Ellie in Belfast says:

    @EM Smith: Here’s another reason why you should produce Smithtemp. I just found this (several links over fromJoe d’Aleo’s recent post on WUWT):
    http://www.friendsofscience.org/assets/documents/CorrectCorrections.pdf
    It summarizes Steve McIntyre’s work on homogenisation – that 45% of the adjustments made are negative adjustments, i.e. those that increase the warming trend.

  28. E.M.Smith says:

    @Ellie

    Thanks, I’ll take a look at it. And yes, GIStemp homogenizing is an abomination. The GHCN data are already available in an “adjusted” form as are the USHCN. To think that you can “unadjust” the data and then readjust them based on a horrid method is, er, um, I’d best be quiet now…

  29. Tonyb says:

    E M Smith

    I need to do a presentation to a Govt dept and wish to provide some pertinent facts they are unlikely to know. As I want to ensure that I get my facts right can you help out with any of the answers?

    These questions apply for both Giss AND Hadcrut.

    ! How many stations are currently used to compile global temperatures?

    2) Which stations are used for the UK (I think Giss uses Bournemouth Airport- are there more?)

    3) Have these stations been constant, If so from when? If not, where and when did they change?

    4) Please explain in what circumstances the data from a station would NOT be used and the means by which a replacement stations data would be used instead.

    Thanks for any help you can provide, there ae so many ins and outs I want to ensure I get it right.

    tonyb

  30. Ellie in Belfast says:

    @Tonyb: I can’t answer for Hadrcrut as I’ve only really looked at stations GISS uses and how they are used. I guess I’m looking from the outside in, as are most people. EM Smith gets that unique view from the inside of the programme.

    2) Off the top of my head I think there are about 96 used in the UK. I have a list I can send you.

    3) The start and stop dates for each station are listed as part of the GISS record. If there are changes to the stations you have to dig a bit deeper to find and understand them. I haven’t quite got to grips with that for the UK yet.

    4) the official explanation is at: http://data.giss.nasa.gov/gistemp/sources/gistemp.html
    but for a lot of the data I have looked at the reasons are not obvious why some data is included or other is left out.

    You’re right about all the ins and outs. It turns my head sometimes.

  31. Ellie in Belfast says:

    Sorry – 58 not 96 – if I’ve caught all of them

  32. Tonyb says:

    Ellie

    Thanks for that.

    It is so complex it is easy to miss one part of the jigsaw and then you can say something that turns out to be wrong and that ruins future credibility. It would be good to get the simple story-hence my simple questions, only they’re not :)

    tonyb

  33. E.M.Smith says:

    Tonyb: I need to do a presentation to a Govt dept and wish to provide some pertinent facts they are unlikely to know. As I want to ensure that I get my facts right can you help out with any of the answers?

    Good luck!

    I’ll do what I can in terms of answers.

    These questions apply for both Giss AND Hadcrut.

    For Hadcrut, we have the “issue” that the “dog ate their homework” … they lost the raw data. So all that is available is the “value added” product. That fact alone might give someone pause… and I’m sure they won’t know it. How do you count the number of stations they “used” when they don’t have the original data? All we have is manipulated “data” and they don’t share how it was manipulated and / or “interpolated” (read fabricated…)

    But one can “speculate” with informed speculation. The stations used for each place on the planet end up in GHCN. These data have a description file for the stations (v2.inv) and that is available from NOAA at the FTP site I’ve posted a few times (including up the comment thread from here… search for ‘ftp’…)

    So you can find out what stations, where on the earth, are in the GHCN set by a simple ftp download. So all you need to do then is find out: “Does Hadcrut claim to use GHCN?” (Almost certainly yes). If so, you have your answer.

    ! How many stations are currently used to compile global temperatures?

    That depends a bit on what ‘currently’ means. It seems to have a bit of a “slow drift” and stations come and go (a point that also ought to be made… how do you measure the size of something when your ruler keeps changing…) but occasionally has massive jumps and drops.

    Also, the number used in any particular year varies from year to year.

    See:

    https://chiefio.wordpress.com/2009/08/10/well-theres-your-global-warming-problem/

    For decade long averages. I can provide annual numbers if needed (but they are probably overkill…) You have my permission to use any of the pages, charts, etc. here in your presentation.

    Realize these are “station records” and sometimes a couple of records from one location will be merged or spliced to make a single “value added” record for the final product. This total number around 7,000. Why around? Because it changes as you move through the GIStemp code… as records are combined, fabricated, homogenized…

    [chiefio@tubularbells STEP1]$ wc -l input_files/v2.inv
    7364 input_files/v2.inv

    So in STEP1, it is 7364, but by the time we exit STEP2, it is 6012:

    $ wc -l STEP2/to_next_step/Ts.GHCN.CL.PA.station.list
    6012 STEP2/to_next_step/Ts.GHCN.CL.PA.station.list
    $

    But you can’t really say those missing 1352 stations were not “used” since they may well have contributed a bit to fabrication of some missing data somewhere else by this point…

    2) Which stations are used for the UK (I think Giss uses Bournemouth Airport- are there more?)

    Well, that will depend just a bit on what “UK” means… England, Scotland, North Ireland, Wales,… and do these show up as part of the same “country code” or not. As a first cut, I looked up “LONDON” in the v2.inv file and got the country code “141”, which looked a lot more like New Zealand… so a search for LONDON/GATW found the list below with country code “651” that looks a lot more like UK to me. It also is 58 records and matches Ellies’ number. I then did a “grep” for all lines starting with country code “651”. That is these lines:

    65103005000 LERWICK 60.13 -1.18 84 52R -9HIxxCO 1x-9WATER A 0
    65103026000 STORNOWAY 58.22 -6.32 13 2R -9FLxxCO 1A-9WARM GRASS/SHRUBB 7
    65103038000 FORT WILLIAM 56.83 -5.10 20 229R -9MVxxCO 1x-9WARM GRASS/SHRUBC 13
    65103038001 BEN NEVIS UK 56.80 -5.10 1343 229R -9MVxxCO 1x-9WARM GRASS/SHRUBB 0
    65103055001 ORKNEY UK 59.10 -3.30 22 31R -9HIxxCO 3x-9WATER A 0
    65103068001 GORDON CASTLE UK 57.60 -3.10 32 101R -9HIFOCO10x-9COASTAL EDGES A 0
    65103072001 BRAEMAR UK 57.00 -3.40 339 610R -9MVxxno-9x-9HEATHS, MOORS A 0
    65103091000 ABERDEEN/DYCE 57.20 -2.22 65 98U 210HIxxCO 5A 3WARM CROPS C 26
    65103100000 TIREE 56.50 -6.88 12 4R -9HIxxCO 1x-9WATER A 0
    65103140000 GLASGOW AIRPO 55.87 -4.43 8 68U 881HIxxCO15x-9WARM CROPS C 46
    65103160000 EDINBURGH AIR 55.95 -3.35 41 51U 470HIxxCO 4A 2WARM GRASS/SHRUBC 29
    65103160001 EDINBURGH/ROYAL OBS.UK 55.90 -3.20 134 155U 470HIxxCO 9x-9WARM GRASS/SHRUBC 24
    65103162000 ESKDALEMUIR 55.32 -3.20 242 293R -9HIxxno-9x-9WARM CROPS A 0
    65103209001 DUMFRIES UK 55.10 -3.10 -999 177S 29HIxxCO10x-9WARM CROPS A 0
    65103241001 COCKLE PARK UK 55.20 -1.60 99 46R -9FLxxCO 5x-9WARM CROPS C 20
    65103242001 DURHAM UK 54.80 -1.60 102 107U 89HIxxCO15x-9WARM CROPS C 25
    65103257000 LEEMING 54.30 -1.53 40 42R -9HIxxno-9A-9WARM CROPS B 9
    65103292001 SCARBOROUGH UK 54.20 -0.40 -999 72S 43HIxxCO 1x-9COASTAL EDGES B 0
    65103302000 VALLEY 53.25 -4.53 11 33R -9HIxxCO 1A-9WARM CROPS B 7
    65103316001 BIDSTON UK 53.40 -2.90 -999 30U 540FLxxCO 3x-9WARM CROPS C 41
    65103329001 STONYHURST UK 53.80 -2.50 115 114R -9HIxxCO30x-9WARM GRASS/SHRUBC 15
    65103334000 MANCHESTER AI 53.35 -2.27 78 50U 490FLxxno-9A 1HEATHS, MOORS C 22
    65103334001 WARRINGTON 53.38 -2.65 27 35U 65FLxxCO25x-9WARM CROPS C 28
    65103345001 SHEFFIELD UK 53.40 -1.50 -999 136U 558HIxxno-9x-9HEATHS, MOORS C 52
    65103355001 YORK UK 53.90 -1.10 -999 22U 102FLxxno-9x-9HEATHS, MOORS A 7
    65103377000 WADDINGTON 53.17 -0.52 70 56U 74FLxxno-9A 3HEATHS, MOORS C 12
    65103482001 MILDENHALL 52.37 0.48 10 40R -9FLxxno-9A-9WARM CROPS B 20
    65103482002 LAKENHEATH 52.40 0.57 10 42R -9FLxxno-9A-9WARM CROPS B 9
    65103482003 SCULTHORPE 52.85 0.77 68 58R -9FLxxCO20A-9WARM CROPS A 0
    65103496001 GORLESTON 52.60 1.70 2 3U 50FLxxCO 1x-9COASTAL EDGES C 21
    65103501001 ABERYSTWYTH UK 52.40 -4.10 -999 37S 15HIxxCO 1x-9COASTAL EDGES A 13
    65103521001 ROSS-ON-WYE UK 51.90 -2.60 -999 96R -9HIxxno-9x-9WARM CROPS C 8
    65103534000 BIRMINGHAM/AI 52.45 -1.73 99 112U 1059HIxxno-9A 1WARM CROPS C 25
    65103534001 EDGBASTON UK 52.50 -1.90 -999 125U 1059HIxxno-9x-9WARM CROPS C 88
    65103560001 HUNTINGTON 52.37 -0.22 49 43S 17FLxxno-9A 2WARM CROPS B 14
    65103590001 BENTWATERS 52.13 1.43 26 51R -9FLxxCO 6A-9WARM CROPS A 0
    65103649001 FAIRFORD 51.68 -1.78 91 110R -9HIxxno-9A-9WARM CROPS A 0
    65103649002 OXFORD 51.75 -1.58 91 111U 117HIxxno-9x-9WARM CROPS C 14
    65103649003 UPPER HEYFORD 51.93 -1.25 133 110R -9HIxxno-9A-9WARM CROPS A 0
    65103657001 OXFORD UK 51.70 -1.20 63 70U 117HIxxno-9x-9WARM CROPS B 8
    65103670001 ROTHAMSTEAD UK 51.70 -0.30 128 101U12332HIxxno-9x-9WARM CROPS C 22
    65103672001 KEW UK 51.50 -0.30 5 44U12332FLxxno-9x-9WARM CROPS C 53
    65103683001 WETHERSFIELD 51.97 0.50 101 76R -9HIxxno-9A-9WARM CROPS A 6
    65103683002 CAMBRIDGE UK 52.20 0.10 12 54U 106FLxxno-9x-9WARM CROPS C 25
    65103696001 FELIXSTOWE 52.00 1.30 3 27S 19FLxxCO 1x-9WATER C 18
    65103696002 WOODBRIDGE 52.08 1.40 29 45R -9FLxxCO 5A-9WARM CROPS A 6
    65103715000 GLAMORGAN/RHOUSE AP 51.40 -3.40 67 26U 282HIxxCO 2A 3WATER B 13
    65103743000 LARKHILL 51.20 -1.80 133 109R -9HIxxno-9x-9WARM CROPS A 15
    65103761001 GREENHAM 51.38 -1.28 125 108S 25FLxxno-9A 2WARM CROPS A 9
    65103776000 LONDON/GATWIC 51.15 -0.18 62 73U12332HIxxno-9A15WARM CROPS C 29
    65103779001 GREENWICH/MARITIME MUK 51.50 0.00 7 37U12332FLxxCO30x-9WARM CROPS C 67
    65103817001 TRURO UK 50.30 -5.10 -999 73S 16HIxxCO 5x-9WARM CROPS B 7
    65103827000 PLYMOUTH WC 50.35 -4.12 50 76U 259HIxxCO 1x-9WATER C 19
    65103862000 BOURNEMOUTH A 50.78 -1.83 11 41U 144HIxxCO 5A 2WARM CROPS C 14
    65103865000 SOUTHAMPTON/ 50.90 -1.40 9 20U 214FLxxCO 3x-9WARM CROPS C 26
    65103874001 OSBORNE UK 50.80 -1.30 52 3R -9HIxxCO 1x-9WARM CROPS C 0
    65103894000 GUERNSEY AIRP 49.43 -2.60 102 10S 16HIxxCO 2A 3WATER B 13
    65103917000 BELFAST/ALDER 54.65 -6.22 81 68U 552FLxxCO20A15WARM CROPS B 13

    Someone more familiar with these places will have to tell me if I’m done, or if I need to go finding the country code for Scotland, et. al. (though the GLASGOW and EDINBURGH on the list imply Scotland is in…) I do note that BELFAST is in this list, but DUBLIN is not:

    Though DUBLIN comes up with a USA country code, DUBLIN AIRPORT is in this group that looks a bit more Irish to me:

    62103952000 ROCHES POINT 51.80 -8.25 41 16R -9HIxxCO 1x-9WARM CROPS B 0
    62103953000 VALENTIA OBSE 51.93 -10.25 14 47R -9HIxxCO 1x-9WATER A 0
    62103955000 CORK AIRPORT 51.85 -8.48 162 82U 134HIxxCO15A 3WARM CROPS C 12
    62103957000 ROSSLARE 52.25 -6.33 25 2R -9FLxxCO 1x-9WATER B 11
    62103960000 KILKENNY 52.67 -7.27 64 94S 13HIxxno-9x-9WARM CROPS C 20
    62103962000 SHANNON AIRPO 52.70 -8.92 20 14R -9FLxxCO 1A-9WARM CROPS B 12
    62103964001 GALWAY 53.28 -9.02 18 20S 29FLxxCO 1x-9WARM CROPS C 15
    62103965000 BIRR 53.08 -7.88 72 80R -9FLxxno-9A-9WARM CROPS A 0
    62103967000 CASEMENT AERO 53.30 -6.43 93 137U 680HIxxCO17A 3WARM CROPS C 24
    62103969000 DUBLIN AIRPOR 53.43 -6.25 85 52U 680FLxxCO10A 3WARM CROPS C 33
    62103970000 CLAREMORRIS 53.72 -8.98 69 78R -9FLxxno-9x-9WARM CROPS B 7
    62103971000 MULLINGAR 53.53 -7.37 104 100R -9FLxxno-9x-9WARM CROPS C 10
    62103974000 CLONES 54.18 -7.23 89 94R -9HIxxno-9x-9WARM CROPS A 0
    62103976000 BELMULLET 54.23 -10.00 10 19R -9HIxxCO 1x-9WARM CROPS A 6
    62103980000 MALIN HEAD 55.37 -7.33 25 13R -9HIxxCO 1x-9WARM CROPS A 0

    What else might be in, or not in, “UK” I’ll leave for others to decide (dodging political issue …).


    3) Have these stations been constant, If so from when? If not, where and when did they change?

    Almost certainly NOT constant. See:

    https://chiefio.wordpress.com/2009/08/17/thermometer-years-by-latitude-warm-globe/

    For a nice chart of thermometer change, by decade, over the life of GHCN. Since, in decade ending 1709 to decade ending 1729, there is exactly ONE thermometer in the record, it will depend on when they choose to ‘start history’. BTW, it then changes to 2 thermometers…

    This chart is in 20 degree bands by lattitude, so 90-70 N, then 70-50 N. Inspection of my globe shows UK as being more or less N of 50 and S of 70, so that’s the N.Cold band so in 1829 there were still not as many stations in the whole band as there are in the UK today (48 on the chart).

    If you need an exact count for the UK only, I can produce that with a couple of hours work. It would not be so much “by latitude” as it would be “UK stations by year”…

    4) Please explain in what circumstances the data from a station would NOT be used and the means by which a replacement stations data would be used instead.

    Good Luck With That One.

    I can not make heads nor tails of the reasons for inclusion nor removal of stations. Some are still gathering data, but whoever “does GHCN” decided to drop them (there is a WUWT comment with a graph showing this imparts a warming bias to the data…). Others get dropped when a station shuts down. (As a hypothetical, an airport might be converted to a shopping mall and stop reporting… I expect this to happen to was WAS the main E. German airport: Tempelhof someday. See:

    https://chiefio.wordpress.com/2009/08/26/agw-gistemp-measure-jet-age-airport-growth/

    where I discuss this.

    Then there was the Soviet Union collapse that removed a large number of cold siberian stations, and, of course, as in the above list, “Add an Airport, Add a Station” seems to be the rule… at least for major airports. But as near as I can tell, a lot of the recent changes have been for “non-technical” reasons (i.e. probably a cherry pick by someone…)

    And whenever a station is “in the v2.inv list” GIStemp will try to fill in missing data in v2.mean for it. The fact that a station leaves the set in, oh, 1990, would not stop GIStemp from making up data where there is “missing” data in the years after 1990. IMHO, this makes dropping stations in recent years a Highly Suspect Behaviour.

    Oh, and for GIStemp in particular, it takes the USHCN stations (that have duplicate records in GHCN for some of the records) and attempts to merge these. It does this by converting USHCN to degrees C, then makes a composite value out of the two records for any one place. Sometimes it keeps USHCN, sometimes it keeps GHCN, sometimes it uses something in between those two… (it depends on which of them contain data, one, the other, or both) and sometimes if there is a gap in both records, it just makes something up via “The Reference Station Method” from thermometers up to 1000 km away.

    One might question the sanity of this, given that the GHCN data came from USHCN in the first place…

    Thanks for any help you can provide, there ae so many ins and outs I want to ensure I get it right.

    Hope this has helped, and if there is something more specific that you need, I am willing to whack on it for a while.

    Oh, and one other caveat: A given station can come into the record, leave, come back with different equipment, leave, have data “made up and filled in” for any gaps (as GIStemp MUST DO given the code. Gaps ARE filled in with fabricated data from up to 1200 km away if needed and possible).

    To put a size on this, the file v2.mean holds the GHCN data. This file has:

    $ wc -l v2.mean
    595737 v2.mean

    595 k records. These are one line for each station and modification history, per year, with 12 monthly averages in that line. Those lines with “9999” in them have missing data. So if we search for those lines, then count them, we will get the number of lines with “missing data”:

    $ grep 9999 v2.mean | wc -l
    107544

    So we have 107 k out of 595 k records with missing data THAT WILL BE FILLED IN WITH FABRICATED DATA IN GIStemp!

    So about 1/6 of the “data lines” for any given year have computer fantasies in them for at least one month, and that is before we get to the really fancy data manipulation… Such as filling in those years that are entirely missing, so don’t even have a record in v2.mean… A “someday” project is to actually measure how much of the final product is real and how much is a complete fabrication.

    Hope this helps, and let me know if you need more of anything in particular…

  34. Ellie in Belfast says:

    The more I look at this stuff the more I am convinced Hanlon’s Razor is stretched to the limit. I’ve just discovered one of the Guam stations I’ve been looking at stopped reporting in 1995. GIStemp reports it up to 2004 so Filnet has been at its work for 9 years. Why stop then? It is very convenient as it strengthens the warmer data more than if it stopped at 1995.

  35. E.M.Smith says:

    @Ellie:

    Yeah. I’m “At Smith’s Limit to Hanlon’s Razor”… (“When stupidity can not explain sufficiently, while malice does, one may resort to malice as the more likely explanation.”)

    The pattern I’m seeing, that I can not yet assert with sufficient data, is that the recent dropping of stations is a deliberate cherry pick that “tunes” the results to warming.

    The “over the edge” for me was the posting on WUWT that has California with a 115 year record high September. There is just no way. Impossible. Then a commenter put in a comment that had a graph with removed vs kept stations showing the warming bias added by removing a bunch of rural stations. It’s not a cherry pick, its a God Damned Sledgehammer IMHO. Just so heavy handed, directed, and tuned that it can not be stupidity. We’ve been chilled almost all summer, and had a cool somewhat wet September. The plants are sulking in the cold. No way that’s record hot.

    And the code itself is all one big Cherry Pick. The things that are parameters are just too conveniently those things that let you tune up for max warming. Yes, it is opinion. I’m working on proving it…

  36. E.M.Smith says:

    Well this is rather interesting… Doing the “thermometers by year” on those records that have the England country code has only 10 stations being used “currently”. So I did a “grep” of the input file to double check that (as I found it hard to believe…)

    [chiefio@tubularbells analysis]$ grep 2008 v2.mean.england
    6510300500012008 42 52 37 60 91 104 131 133 120 78 57 48
    6510302600012008 52 58 49 69 112 115 140 143 118 83 68 49
    6510309100012008 43 58 46 65 104 125 149 146 125 85 57 41
    6510310000012008 63 71 61 76 120 120 142 144 127 95 78 59
    6510316200032008 37 41 40 58 114 117 147 140 110 73 51 19
    6510325700012008 61 47 60 74 118 135 162 161 134 94 65 32
    6510330200012008 78 69 70 83 139 134 160 159 140 110 87 56
    6510337700012008 66 54 60 79 131 145 168 171 139 102 71 36
    6510386200012008 73 56 71 85 142 147 165 165 137 101 80 41
    6510391700032008 55 58 58 80 129 134 154 153 129 91 67 45

    Yet there it is… Of all those UK Thermometer Records, we only have 10 being used at present… (The same 10 are used for 2009, but the year not being over yet, I chose to use 2008 as the benchmark cutoff…)

    This record is the 3 digit country code, 5 digit station ID, 3 digit sub-station, 1 digit modification code, and then the 4 digit year. Then 12 records of monthly averages of daily min/max averages in 1/10 C.

    (And yes, I know England is not the UK, I just didn’t what to use a lowercase UK in a file name and England is an easy search key while doing an “ls” in my directory…)

    Hmm…. he thinks. Mighty strange. You can put the codes into the GISS tool and find out where they are and what their profiles look like. Ellie can show you how if you need help.

    Meanwhile, this caused me to wonder what the thermometer counts looked like by year for the whole history. Here it is. not pretty… The trailing 1/100 C digit is bogus but is there to check that the program was not adding precision nor had misaligned data. It ought to always be a zero. Also, the records are long so will wrap from one line to another. The right most field is the count of active thermometers in that year. It may only have had one month’s data, but it was used for the year (and the average I compute for that year).

    Each of the other fields are the 12 monthly averages of the daily means for the data available for those thermometers. I don’t see a lot of warming trend in this series, but have not formally analyzed it. The “13th” temperature is the annual average. The 1850 or so to about 2001 or so trend is pretty flat and 1846 is particularly interesting.

    [chiefio@tubularbells analysis]$ more v2.mean.sorted.yrs.GAT
    GAT year: 1763 -0.70 6.00 5.20 8.10 9.10 14.00 15.30 15.60 13.10 8.80 6.20 6.80 8.96 1
    GAT year: 1764 4.10 4.10 4.30 7.00 11.80 13.30 15.70 14.90 11.60 8.40 4.70 2.80 8.56 2
    GAT year: 1765 5.20 0.80 5.30 7.60 10.90 13.00 14.70 14.90 12.60 9.40 3.70 1.80 8.32 2
    GAT year: 1766 1.10 2.20 4.10 8.10 8.60 13.40 15.40 16.00 13.00 9.30 7.60 3.50 8.53 2
    GAT year: 1767 0.40 5.90 5.00 6.70 9.70 11.70 13.90 15.80 13.70 9.10 7.20 3.80 8.57 2
    GAT year: 1768 1.40 5.10 4.70 8.10 11.00 13.50 14.90 15.70 11.60 9.30 5.60 4.80 8.81 2
    GAT year: 1769 3.00 3.00 5.00 7.50 10.40 12.40 15.50 14.50 12.70 7.80 5.80 5.30 8.58 2
    GAT year: 1770 4.30 5.10 2.90 5.50 9.50 12.60 15.10 15.60 13.70 8.20 5.00 4.20 8.47 2
    GAT year: 1771 1.00 3.00 2.70 5.30 11.50 13.50 15.40 14.60 11.80 9.30 6.00 6.00 8.34 2
    GAT year: 1772 0.70 1.50 3.70 6.20 9.60 15.30 16.00 15.70 12.30 10.70 6.50 4.30 8.54 2
    GAT year: 1773 3.60 2.30 5.70 7.70 9.50 13.70 15.20 16.30 12.20 9.10 4.30 3.10 8.56 2
    GAT year: 1774 -0.40 3.20 4.70 7.50 9.40 14.20 15.30 15.80 12.00 9.40 4.30 3.00 8.20 2
    GAT year: 1775 4.20 5.30 5.20 9.40 12.40 15.20 16.70 15.20 13.50 8.50 4.20 4.00 9.48 2
    GAT year: 1776 -1.80 3.60 6.10 8.60 10.30 13.80 16.40 15.00 12.00 10.00 5.60 4.30 8.66 2
    GAT year: 1777 1.80 2.00 5.80 6.70 11.50 13.20 15.40 16.20 14.10 10.10 6.40 3.20 8.87 2
    GAT year: 1778 2.70 3.40 4.80 7.60 12.50 16.10 17.90 16.30 11.40 7.00 6.10 6.30 9.34 2
    GAT year: 1779 2.80 8.10 8.10 9.50 11.80 14.70 19.00 18.70 14.70 10.30 5.30 2.40 10.45 2
    GAT year: 1780 -1.40 2.00 8.10 6.10 12.80 14.70 16.90 18.10 14.80 8.90 4.10 3.70 9.07 2
    GAT year: 1781 2.50 4.90 6.90 8.90 11.50 16.30 16.60 15.60 12.40 9.20 6.10 5.10 9.67 3
    GAT year: 1782 3.80 1.30 3.10 4.40 7.50 14.20 15.20 14.50 11.60 6.50 2.20 2.70 7.25 3
    GAT year: 1783 2.80 3.70 2.90 9.50 9.30 13.70 18.10 14.90 11.90 8.60 6.00 2.90 8.69 3
    GAT year: 1784 0.00 1.30 2.00 5.50 13.50 13.10 15.30 13.80 13.30 7.60 4.60 0.50 7.54 3
    GAT year: 1785 2.80 -0.40 0.60 9.00 10.70 15.90 15.50 13.00 12.90 7.70 5.80 2.20 7.97 3
    GAT year: 1786 2.70 2.80 1.70 7.30 10.30 14.60 14.00 14.80 10.70 6.80 3.60 2.50 7.65 3
    GAT year: 1787 4.10 6.00 6.80 6.80 10.40 13.00 15.30 15.40 12.50 9.10 3.70 3.00 8.84 3
    GAT year: 1788 3.80 3.20 3.10 9.70 12.00 14.30 16.20 15.30 13.00 8.80 5.80 0.20 8.78 3
    GAT year: 1789 1.60 4.70 1.70 7.20 11.90 13.80 15.90 16.60 12.80 8.80 4.90 6.30 8.85 3
    GAT year: 1790 4.20 6.70 6.70 6.00 11.60 14.30 14.70 15.00 11.40 9.50 4.90 3.70 9.06 3
    GAT year: 1791 4.30 4.10 6.60 8.80 10.80 13.60 15.30 15.40 13.10 8.50 5.30 0.90 8.89 3
    GAT year: 1792 2.00 4.60 5.60 10.10 9.80 12.60 15.20 16.10 11.20 8.50 7.20 4.00 8.91 3
    GAT year: 1793 3.00 4.40 3.70 5.80 10.40 13.00 16.70 14.60 11.90 11.10 5.80 5.30 8.81 3
    GAT year: 1794 2.10 6.30 6.70 9.60 10.30 15.10 17.30 14.70 12.00 8.90 5.60 3.70 9.36 4
    GAT year: 1795 -2.00 0.30 3.70 7.90 9.60 12.40 15.00 16.10 15.20 11.20 4.00 6.50 8.32 4
    GAT year: 1796 6.80 4.50 4.40 10.40 10.10 13.50 15.20 15.60 14.00 7.80 4.40 0.00 8.89 4
    GAT year: 1797 4.10 5.20 4.50 7.30 11.30 12.70 16.40 15.00 12.20 8.00 4.80 4.50 8.83 4
    GAT year: 1798 3.80 3.70 4.80 10.40 12.50 16.30 15.90 15.70 12.50 8.90 4.40 3.20 9.34 4
    GAT year: 1799 1.90 2.20 3.00 5.00 8.90 13.00 14.40 13.30 12.20 7.60 4.90 1.20 7.30 4
    GAT year: 1800 2.10 1.90 3.60 8.70 11.20 12.30 16.20 15.90 13.10 8.70 4.90 2.90 8.46 4
    GAT year: 1801 3.90 4.00 6.00 8.00 11.30 13.80 14.60 16.30 13.80 9.90 4.40 1.40 8.95 4
    GAT year: 1802 1.60 3.30 5.50 8.30 9.90 13.20 13.40 16.20 13.30 9.50 5.50 3.50 8.60 4
    GAT year: 1803 1.60 2.80 5.50 8.50 9.80 12.80 16.80 15.10 10.70 8.70 4.30 3.40 8.33 4
    GAT year: 1804 5.10 2.40 3.90 5.90 12.70 15.00 14.50 14.70 13.10 9.80 5.40 1.90 8.70 4
    GAT year: 1805 2.00 3.30 5.90 7.60 9.20 12.20 15.20 15.60 13.60 7.80 5.30 3.10 8.40 4
    GAT year: 1806 2.80 3.50 4.20 6.10 11.10 14.00 14.80 15.20 12.90 9.70 6.70 5.50 8.87 4
    GAT year: 1807 2.50 2.70 2.70 7.00 10.60 13.20 16.40 16.20 9.60 10.40 2.10 2.00 7.95 4
    GAT year: 1808 2.20 2.20 3.00 5.50 13.20 14.00 17.50 15.90 12.30 6.60 5.30 2.10 8.32 4
    GAT year: 1809 0.30 4.90 5.80 4.80 12.30 13.00 14.30 14.50 11.90 10.00 4.30 3.30 8.28 4
    GAT year: 1810 2.00 2.80 3.60 7.70 8.50 14.00 14.70 14.70 13.50 9.30 4.80 2.70 8.19 4
    GAT year: 1811 1.00 3.70 6.80 8.00 12.10 13.60 15.30 14.20 13.20 11.60 7.20 2.80 9.12 4
    GAT year: 1812 2.50 4.80 3.00 5.10 10.40 12.80 13.60 14.30 12.80 8.60 4.50 1.80 7.85 4
    GAT year: 1813 2.00 5.10 6.60 7.40 11.00 13.50 15.40 14.70 12.50 7.50 3.90 2.90 8.54 4
    GAT year: 1814 -2.90 1.30 2.80 9.30 9.00 11.90 15.40 14.50 12.60 7.60 3.90 2.80 7.35 4
    GAT year: 1815 -0.10 5.00 5.70 7.50 11.40 13.50 14.60 14.60 12.50 9.30 3.40 1.60 8.25 4
    GAT year: 1816 2.00 1.80 3.30 5.70 9.40 12.40 13.60 13.30 11.50 9.00 3.60 2.20 7.32 4
    GAT year: 1817 3.90 5.30 4.60 6.90 8.30 14.00 13.90 13.00 12.90 5.90 7.80 1.50 8.17 4
    GAT year: 1818 3.20 1.80 3.30 5.90 11.00 16.10 17.30 14.80 12.70 11.50 8.60 3.60 9.15 4
    GAT year: 1819 3.50 3.00 6.00 7.80 11.00 13.30 16.20 17.60 13.00 8.30 3.50 1.10 8.69 4
    GAT year: 1820 -0.60 3.40 4.70 8.60 11.00 12.90 15.40 14.50 12.00 7.30 5.30 4.20 8.23 4
    GAT year: 1821 3.00 3.00 5.40 9.30 9.10 12.00 14.60 15.20 14.30 9.90 7.00 5.30 9.01 4
    GAT year: 1822 4.10 5.30 7.20 8.00 12.10 16.20 15.30 15.00 11.70 9.80 7.40 1.80 9.49 4
    GAT year: 1823 -0.10 2.00 4.70 6.30 11.80 12.10 13.90 14.60 12.10 7.90 7.00 4.00 8.03 4
    GAT year: 1824 4.20 4.20 4.30 7.50 10.20 13.50 15.90 15.20 13.10 8.30 5.70 3.90 8.83 4
    GAT year: 1825 3.60 3.60 4.80 8.40 11.00 13.80 17.00 16.40 14.60 10.00 4.10 4.10 9.28 4
    GAT year: 1826 0.40 5.80 5.50 8.00 11.30 15.90 17.50 17.00 13.30 10.50 4.10 5.30 9.55 4
    GAT year: 1827 1.40 0.80 4.80 7.80 11.10 13.60 15.60 14.00 13.20 10.70 6.00 6.50 8.79 5
    GAT year: 1828 4.60 4.80 6.20 7.80 11.20 14.80 15.80 14.90 14.00 10.10 7.60 7.10 9.91 5
    GAT year: 1829 0.60 4.20 4.50 6.30 12.00 14.30 15.10 13.60 11.10 8.30 4.70 2.10 8.07 5
    GAT year: 1830 0.80 2.40 7.40 8.30 11.60 12.30 15.70 14.50 12.10 10.50 6.70 2.10 8.70 5
    GAT year: 1831 1.90 4.60 6.70 8.50 10.70 14.70 16.20 16.30 13.40 12.30 5.50 6.00 9.73 5
    GAT year: 1832 3.60 4.10 5.60 8.30 10.40 14.40 15.20 15.10 13.20 10.50 6.40 5.30 9.34 5
    GAT year: 1833 1.80 5.10 3.70 7.40 14.20 13.90 15.30 14.10 11.90 10.10 6.10 6.10 9.14 5
    GAT year: 1834 6.10 5.30 6.70 7.50 12.40 14.90 16.40 15.60 13.30 10.30 6.80 6.00 10.11 6
    GAT year: 1835 3.70 5.20 5.50 7.70 10.40 13.70 15.50 15.70 12.80 8.40 6.90 3.70 9.10 6
    GAT year: 1836 3.80 3.30 5.20 6.50 11.00 14.30 14.60 14.10 10.90 8.40 5.10 4.10 8.44 6
    GAT year: 1837 3.20 4.70 2.30 4.40 9.20 14.30 15.80 14.70 12.20 10.40 5.40 5.60 8.52 6
    GAT year: 1838 -0.40 0.10 4.80 5.40 9.70 13.20 15.10 14.50 12.20 9.30 4.80 4.60 7.78 6
    GAT year: 1839 2.70 3.70 3.80 6.30 9.60 13.40 15.00 14.10 12.40 9.20 6.90 4.10 8.43 6
    GAT year: 1840 4.10 3.40 4.40 9.40 10.60 13.30 13.90 15.30 11.00 7.90 6.00 2.30 8.47 6
    GAT year: 1841 1.10 2.80 7.50 7.40 11.80 12.00 13.00 14.00 13.00 7.80 4.90 4.30 8.30 5
    GAT year: 1842 1.50 4.40 5.90 7.60 11.00 14.60 14.00 16.10 12.70 7.60 5.30 7.20 8.99 5
    GAT year: 1843 4.30 1.50 5.40 7.60 9.50 11.90 14.40 14.90 14.20 7.80 5.90 7.60 8.75 5
    GAT year: 1844 4.40 1.90 5.00 9.60 10.30 13.40 14.40 13.20 12.40 8.80 6.40 1.60 8.45 5
    GAT year: 1845 3.50 1.90 2.50 7.90 9.10 14.00 13.50 13.10 11.40 9.50 6.80 4.30 8.13 5
    GAT year: 1846 6.20 6.50 5.70 7.30 11.60 16.80 15.80 15.90 14.60 9.50 7.30 1.50 9.89 5
    GAT year: 1847 2.30 2.10 5.20 6.20 11.20 13.10 16.60 14.80 10.80 10.00 7.60 4.80 8.73 5
    GAT year: 1848 1.40 5.10 5.30 7.00 13.10 13.00 14.80 12.80 12.10 8.90 5.40 5.30 8.68 6
    GAT year: 1849 3.70 5.50 5.70 6.00 11.00 12.70 14.50 14.50 12.60 8.50 6.00 3.60 8.69 7
    GAT year: 1850 0.80 6.10 4.60 7.90 9.20 14.20 15.10 13.80 11.60 7.60 6.70 4.60 8.52 7
    GAT year: 1851 6.00 4.90 5.80 7.20 10.20 13.80 14.40 15.30 12.80 10.70 3.90 5.50 9.21 10
    GAT year: 1852 5.30 4.90 5.20 7.80 10.50 12.90 18.20 15.80 12.90 8.60 8.00 7.70 9.82 10
    GAT year: 1853 5.40 1.10 3.60 7.60 10.40 13.80 14.90 14.50 12.40 10.20 6.50 2.70 8.59 9
    GAT year: 1854 3.90 4.60 6.90 8.60 10.10 12.70 15.10 15.30 14.30 9.60 5.50 5.30 9.33 9
    GAT year: 1855 2.70 -0.40 3.60 7.30 8.60 13.00 16.20 15.70 13.40 10.10 5.80 3.50 8.29 10
    GAT year: 1856 3.80 5.20 4.30 7.60 8.90 13.20 14.80 15.80 12.00 10.80 5.50 4.90 8.90 11
    GAT year: 1857 3.00 4.50 5.00 6.90 10.70 14.80 15.90 16.70 14.50 11.20 7.80 7.80 9.90 12
    GAT year: 1858 4.30 2.60 4.80 7.50 10.10 15.70 14.20 15.10 14.00 9.30 4.80 5.40 8.98 12
    GAT year: 1859 5.00 5.50 6.90 6.60 11.00 14.10 17.10 15.90 12.80 9.40 5.50 2.20 9.33 13
    GAT year: 1860 3.60 1.90 4.90 6.10 11.40 11.80 13.90 13.10 11.10 9.50 5.00 2.30 7.88 13
    GAT year: 1861 2.20 5.00 6.00 7.50 10.40 14.30 14.70 15.40 13.00 11.60 4.60 4.50 9.10 13
    GAT year: 1862 4.20 5.00 5.10 8.10 11.60 12.20 13.40 14.00 12.90 10.00 4.00 6.50 8.92 13
    GAT year: 1863 4.80 5.80 6.30 8.10 9.90 13.00 14.60 14.70 11.10 9.80 7.30 6.10 9.29 13
    GAT year: 1864 2.90 1.90 4.40 8.50 11.40 12.90 14.80 13.70 12.70 9.30 5.90 4.10 8.54 13
    GAT year: 1865 2.70 2.40 3.10 9.80 11.50 14.90 15.90 14.60 16.00 10.00 6.80 6.40 9.51 14
    GAT year: 1866 5.70 4.30 4.30 7.80 9.50 14.50 15.20 14.10 12.20 10.50 7.00 6.40 9.29 14
    GAT year: 1867 1.40 6.70 3.20 8.70 10.40 13.60 14.20 15.40 13.30 9.40 5.80 4.30 8.87 14
    GAT year: 1868 3.70 6.30 6.70 8.50 12.40 14.70 17.70 15.80 13.90 8.60 5.20 7.10 10.05 14
    GAT year: 1869 5.70 7.10 3.70 9.50 9.00 12.60 16.30 14.80 13.60 9.80 6.40 3.10 9.30 14
    GAT year: 1870 3.60 2.60 4.80 8.80 11.10 14.50 16.70 15.50 13.20 9.70 5.30 1.40 8.93 15
    GAT year: 1871 1.20 6.00 7.00 7.90 10.70 12.50 15.20 16.60 12.70 9.70 4.10 3.90 8.96 20
    GAT year: 1872 4.90 6.40 6.50 8.20 9.60 13.90 16.50 15.00 12.80 8.50 6.60 5.20 9.51 22
    GAT year: 1873 5.10 2.20 5.20 7.60 9.70 14.00 15.90 15.30 11.90 8.30 6.40 5.70 8.94 22
    GAT year: 1874 5.50 4.30 6.70 9.30 9.50 13.60 16.50 14.70 13.30 10.00 6.00 0.90 9.19 22
    GAT year: 1875 5.90 2.60 4.90 8.10 11.60 13.70 14.40 15.70 14.30 9.20 5.40 4.10 9.16 22
    GAT year: 1876 3.60 4.50 4.40 7.70 9.50 13.70 16.70 15.70 12.40 11.00 6.30 6.00 9.29 22
    GAT year: 1877 5.30 5.90 4.60 6.80 8.70 14.50 14.60 14.60 11.30 9.30 7.10 5.00 8.98 23
    GAT year: 1878 4.60 5.70 5.50 8.40 11.40 14.40 16.30 15.90 13.30 10.40 4.00 0.40 9.19 23
    GAT year: 1879 0.20 3.00 4.30 5.70 8.60 12.60 13.50 14.30 12.20 9.00 4.60 1.40 7.45 24
    GAT year: 1880 2.10 5.70 6.30 7.90 10.20 13.20 14.90 16.20 14.30 7.30 5.40 5.10 9.05 24
    GAT year: 1881 -0.70 2.90 4.80 6.80 11.20 13.00 15.70 13.50 12.40 7.70 8.60 4.30 8.35 24
    GAT year: 1882 5.30 6.00 7.20 7.80 11.00 12.70 14.80 14.80 11.90 9.80 5.60 3.30 9.18 25
    GAT year: 1883 4.60 5.70 2.40 7.70 10.10 13.20 14.20 14.90 12.90 9.50 6.00 4.90 8.84 25
    GAT year: 1884 5.70 4.80 5.70 6.60 10.30 12.90 15.00 15.90 13.50 9.00 5.40 4.00 9.07 27
    GAT year: 1885 2.60 5.00 4.20 7.00 8.10 12.80 15.10 13.00 11.50 6.90 5.70 3.80 7.98 27
    GAT year: 1886 1.70 1.40 3.60 6.80 9.50 12.40 14.70 14.70 12.60 10.50 6.40 1.90 8.02 27
    GAT year: 1887 2.60 4.00 3.50 5.90 9.10 14.70 16.30 14.70 11.30 7.00 4.50 2.80 8.03 28
    GAT year: 1888 3.50 1.70 2.70 5.70 9.90 12.30 12.90 13.30 11.90 8.10 7.20 5.00 7.85 28
    GAT year: 1889 3.60 2.80 4.40 6.50 12.00 14.60 14.30 14.10 12.20 8.20 6.70 3.70 8.59 29
    GAT year: 1890 5.60 3.30 5.70 6.80 11.20 13.00 13.70 13.80 14.30 9.60 5.80 0.60 8.62 29
    GAT year: 1891 1.90 5.00 3.60 5.70 8.90 13.80 14.30 13.60 13.50 9.10 5.40 4.30 8.26 29
    GAT year: 1892 2.30 3.30 2.60 6.90 10.60 12.50 13.60 14.40 11.60 6.70 6.50 2.20 7.77 29
    GAT year: 1893 2.40 4.30 6.90 9.80 12.10 14.70 15.30 16.40 12.30 9.50 4.90 4.80 9.45 29
    GAT year: 1894 3.00 4.50 6.30 8.90 8.50 12.80 15.00 13.80 11.40 8.70 7.50 5.00 8.78 29
    GAT year: 1895 0.20 -1.30 4.90 7.80 11.40 13.80 14.40 14.90 14.90 7.00 7.10 3.90 8.25 29
    GAT year: 1896 4.60 4.80 6.20 8.50 11.60 14.80 15.30 13.60 12.40 6.60 4.80 3.90 8.93 29
    GAT year: 1897 1.60 5.20 5.80 6.40 9.30 13.80 15.60 15.50 11.50 9.80 7.40 4.70 8.88 29
    GAT year: 1898 6.30 4.30 4.10 7.80 9.30 12.80 14.50 15.30 14.50 10.90 6.70 6.70 9.43 30
    GAT year: 1899 4.20 4.70 4.80 7.00 9.10 14.60 16.20 16.80 12.60 9.20 8.40 2.60 9.18 30
    GAT year: 1900 4.20 1.90 3.30 7.70 9.60 13.70 16.30 14.30 13.00 9.10 6.80 6.80 8.89 30
    GAT year: 1901 3.30 2.10 3.70 7.70 10.90 12.90 16.90 14.90 13.20 8.80 5.00 3.10 8.54 27
    GAT year: 1902 4.10 1.30 6.00 6.90 7.90 12.50 13.50 13.20 12.30 9.00 6.80 4.40 8.16 27
    GAT year: 1903 3.50 6.20 6.10 5.70 10.00 12.10 14.10 13.30 12.30 9.50 6.00 3.00 8.48 27
    GAT year: 1904 4.10 3.20 4.20 8.40 10.20 13.00 16.00 14.70 12.40 9.70 5.60 4.10 8.80 25
    GAT year: 1905 4.00 4.90 6.50 6.80 10.50 14.10 16.70 14.30 12.20 7.20 5.00 5.30 8.96 25
    GAT year: 1906 5.10 3.00 4.70 6.90 9.80 13.70 15.10 16.10 13.60 10.40 7.40 3.20 9.08 25
    GAT year: 1907 3.70 3.00 6.40 7.40 9.90 12.10 13.80 13.90 13.40 9.70 6.50 4.60 8.70 24
    GAT year: 1908 2.90 5.30 4.20 5.90 12.00 13.90 15.40 14.30 12.80 12.00 7.40 4.30 9.20 24
    GAT year: 1909 3.60 3.40 3.40 8.40 10.50 11.80 14.40 15.30 11.80 10.00 4.90 3.60 8.43 24
    GAT year: 1910 3.40 4.70 6.10 6.90 10.70 14.10 13.90 14.90 12.50 10.70 3.50 6.30 8.98 24
    GAT year: 1911 4.10 4.70 5.00 7.40 12.50 14.00 17.30 17.50 13.60 9.10 5.90 5.90 9.75 24
    GAT year: 1912 3.60 5.10 6.90 8.70 11.20 13.30 15.40 12.50 10.90 8.30 6.10 6.20 9.02 24
    GAT year: 1913 4.10 4.80 5.70 7.50 10.90 13.70 14.40 14.90 13.60 10.80 8.10 4.90 9.45 24
    GAT year: 1914 3.70 6.50 5.70 9.50 10.40 14.10 15.60 15.80 13.10 10.30 6.70 4.30 9.64 24
    GAT year: 1915 3.80 3.90 5.10 7.60 10.20 13.80 14.40 15.10 13.10 9.00 3.30 4.80 8.68 24
    GAT year: 1916 7.10 3.60 3.20 7.90 10.90 11.40 14.80 15.90 12.80 10.20 6.80 2.60 8.93 24
    GAT year: 1917 1.90 1.70 3.10 5.10 11.80 14.40 15.60 15.20 13.70 7.40 7.90 2.70 8.37 24
    GAT year: 1918 3.50 6.30 5.80 6.60 12.30 13.00 15.10 15.60 11.40 9.20 5.60 6.50 9.24 24
    GAT year: 1919 3.00 2.00 3.40 7.10 12.40 13.70 13.80 15.40 12.60 7.90 3.00 5.00 8.27 23
    GAT year: 1920 4.70 5.90 6.80 7.70 11.30 13.80 13.80 13.40 12.70 10.40 7.10 4.20 9.32 24
    GAT year: 1921 6.70 5.00 6.80 8.10 11.10 14.10 17.40 14.80 13.90 12.60 5.00 6.50 10.17 24
    GAT year: 1922 3.60 4.40 4.70 5.40 12.10 13.30 13.30 13.20 11.80 8.40 6.00 5.60 8.48 24
    GAT year: 1923 5.60 5.20 6.30 7.10 8.90 12.20 16.70 14.60 12.00 9.40 3.50 3.60 8.76 24
    GAT year: 1924 4.50 3.50 3.90 6.80 10.70 13.40 14.80 13.90 12.80 9.90 7.10 6.80 9.01 24
    GAT year: 1925 5.20 4.70 4.80 7.00 11.00 14.40 16.40 15.10 11.20 10.20 3.90 2.70 8.88 24
    GAT year: 1926 4.50 6.30 6.20 8.90 9.70 13.20 16.50 15.90 14.00 7.80 5.90 4.50 9.45 24
    GAT year: 1927 4.50 4.40 6.90 7.50 10.30 11.80 15.30 15.20 12.00 10.10 6.00 2.00 8.83 24
    GAT year: 1928 4.90 5.60 5.60 7.80 10.30 12.20 15.60 14.90 12.40 9.80 7.50 3.60 9.18 24
    GAT year: 1929 1.80 0.70 6.40 6.30 10.90 12.80 15.60 14.80 15.00 9.30 6.60 5.40 8.80 24
    GAT year: 1930 5.10 2.50 4.90 7.80 10.30 14.70 14.90 15.40 13.30 10.10 6.00 4.30 9.11 24
    GAT year: 1931 3.40 3.70 3.80 7.20 10.60 13.20 14.70 13.80 11.40 9.00 7.80 5.60 8.68 32
    GAT year: 1932 6.20 4.00 4.80 6.40 9.70 13.30 15.30 16.10 12.60 8.50 6.60 5.70 9.10 32
    GAT year: 1933 2.80 4.10 7.00 8.50 11.10 14.60 17.00 16.70 14.50 10.00 6.00 2.70 9.58 32
    GAT year: 1934 4.70 4.60 4.80 7.20 10.60 14.00 17.10 14.80 13.90 10.00 6.20 7.70 9.63 32
    GAT year: 1935 4.80 5.40 6.40 7.40 9.50 13.90 16.10 15.90 13.00 9.20 6.60 3.10 9.27 32
    GAT year: 1936 3.50 2.70 6.40 6.10 10.70 13.80 14.90 15.50 13.90 9.30 5.90 5.30 9.00 32
    GAT year: 1937 5.00 4.90 3.20 8.40 11.30 13.40 15.20 15.90 12.90 10.20 5.80 3.10 9.11 32
    GAT year: 1938 5.50 5.00 8.80 7.60 9.90 13.40 14.40 15.30 13.40 10.10 8.70 4.40 9.71 32
    GAT year: 1939 3.90 5.60 5.80 8.00 10.70 13.60 14.70 15.80 13.80 8.40 8.10 3.70 9.34 32
    GAT year: 1940 0.00 2.70 5.60 7.80 11.80 15.60 14.40 14.90 12.20 9.50 6.80 4.20 8.79 32
    GAT year: 1941 0.80 3.10 4.50 6.10 8.80 13.90 16.20 14.10 14.20 10.10 6.50 5.70 8.67 32
    GAT year: 1942 1.60 0.80 4.60 8.40 10.30 13.40 14.70 15.60 13.00 10.10 5.60 6.50 8.72 32
    GAT year: 1943 4.60 6.00 6.60 9.90 11.10 13.60 15.30 14.90 12.70 10.60 6.30 4.30 9.66 32
    GAT year: 1944 5.90 3.70 5.20 9.30 10.50 12.80 15.50 16.10 12.00 9.10 5.90 4.10 9.18 32
    GAT year: 1945 0.70 6.70 7.90 9.30 11.20 13.70 15.90 15.20 13.90 11.50 7.60 5.30 9.91 32
    GAT year: 1946 3.40 5.60 5.20 9.20 10.10 12.60 15.40 14.00 13.40 9.60 7.70 3.50 9.14 32
    GAT year: 1947 2.60 -1.10 2.90 7.90 12.00 14.40 16.00 17.80 14.20 10.70 6.80 5.20 9.12 32
    GAT year: 1948 4.60 4.60 7.90 8.60 10.60 12.90 14.90 14.20 13.20 9.80 7.40 5.70 9.53 32
    GAT year: 1949 5.20 5.50 4.90 9.30 10.40 14.00 16.30 15.90 15.50 11.30 6.60 5.40 10.03 33
    GAT year: 1950 4.50 4.80 7.30 7.00 10.60 15.10 15.30 15.10 12.40 9.50 5.50 1.40 9.04 39
    GAT year: 1951 3.50 3.30 3.80 6.20 9.30 13.00 15.50 14.40 13.60 9.90 8.00 5.30 8.82 45
    GAT year: 1952 2.40 3.40 6.20 9.00 12.50 13.60 16.00 15.40 10.70 8.80 4.10 3.10 8.77 47
    GAT year: 1953 3.70 4.40 5.70 6.90 12.00 13.60 15.00 15.50 13.60 9.90 8.20 6.80 9.61 47
    GAT year: 1954 3.10 2.40 5.50 7.30 10.80 12.90 13.80 14.10 12.20 11.30 6.70 6.20 8.86 49
    GAT year: 1955 2.40 1.10 3.20 9.00 9.20 13.10 16.80 17.20 13.90 9.00 7.30 5.30 8.96 49
    GAT year: 1956 3.30 0.20 5.80 6.30 11.50 12.50 15.10 13.00 13.70 9.40 6.20 5.80 8.57 47
    GAT year: 1957 5.20 4.80 8.60 8.40 10.00 14.40 15.70 14.90 12.00 10.50 6.50 4.60 9.63 47
    GAT year: 1958 3.20 4.30 3.30 7.10 10.50 13.10 15.30 15.30 14.60 10.70 6.70 4.50 9.05 50
    GAT year: 1959 1.70 4.50 7.00 9.00 11.90 14.50 16.60 16.80 14.70 12.30 7.00 5.70 10.14 50
    GAT year: 1960 3.90 3.70 6.00 8.70 12.20 15.30 14.70 14.60 12.90 10.20 7.00 3.90 9.43 49
    GAT year: 1961 3.50 6.60 8.00 9.30 10.40 13.80 14.60 14.90 14.50 10.50 6.00 2.40 9.54 52
    GAT year: 1962 4.10 4.40 2.60 7.20 9.70 13.00 14.10 14.00 12.20 10.40 5.50 2.40 8.30 52
    GAT year: 1963 -1.30 -0.40 5.70 8.00 9.90 14.10 14.40 13.80 12.50 10.50 7.50 2.90 8.13 51
    GAT year: 1964 3.50 4.30 4.20 8.20 12.40 13.20 15.30 14.80 13.50 8.70 7.20 3.50 9.07 51
    GAT year: 1965 3.20 3.30 4.90 7.60 10.80 13.60 13.20 14.20 12.00 10.60 4.40 4.00 8.48 48
    GAT year: 1966 2.90 4.90 6.30 6.60 10.50 14.60 14.30 14.10 13.60 9.80 5.30 4.80 8.98 47
    GAT year: 1967 4.30 5.30 6.70 7.50 9.70 13.60 16.00 15.10 13.30 10.30 5.60 4.20 9.30 48
    GAT year: 1968 4.20 1.80 6.00 7.50 9.00 14.00 14.20 14.80 13.30 11.90 6.40 3.40 8.88 41
    GAT year: 1969 5.10 0.90 3.20 7.00 10.30 13.30 15.90 15.70 13.40 12.40 4.80 3.60 8.80 40
    GAT year: 1970 3.80 2.70 3.80 6.30 11.80 15.40 14.20 15.30 13.60 10.30 7.20 4.40 9.07 30
    GAT year: 1971 4.70 4.90 5.30 7.50 10.90 11.80 15.70 14.80 13.60 11.10 6.10 6.80 9.43 31
    GAT year: 1972 4.00 4.40 6.10 8.10 10.00 11.30 14.80 14.20 11.30 10.30 5.90 5.70 8.84 30
    GAT year: 1973 4.70 4.10 6.10 6.50 10.40 14.00 14.90 15.40 13.40 8.90 5.70 4.70 9.07 30
    GAT year: 1974 5.90 5.40 5.30 7.70 10.40 12.90 14.20 14.40 11.40 7.60 6.40 7.20 9.07 30
    GAT year: 1975 6.10 4.70 4.60 7.80 9.20 13.70 16.20 17.20 12.40 9.90 6.10 5.30 9.43 30
    GAT year: 1976 5.30 4.70 4.70 7.70 10.90 15.40 17.10 16.30 12.70 10.10 6.00 2.10 9.42 27
    GAT year: 1977 2.70 4.40 6.50 6.80 9.80 11.80 15.20 14.50 12.30 11.30 5.80 5.90 8.92 27
    GAT year: 1978 3.00 2.20 6.20 6.10 11.00 12.90 13.90 14.20 13.20 11.40 8.10 4.10 8.86 27
    GAT year: 1979 0.40 1.60 4.00 7.00 8.70 13.10 14.90 13.80 12.30 10.80 6.40 5.20 8.18 24
    GAT year: 1980 2.40 5.20 4.40 8.30 10.30 12.90 13.70 14.80 13.80 8.50 6.40 5.30 8.83 24
    GAT year: 1981 4.50 3.40 6.80 7.30 10.50 12.40 14.40 15.10 13.60 7.40 7.10 1.00 8.62 19
    GAT year: 1982 3.00 5.10 5.70 8.20 10.40 13.90 15.40 14.60 13.10 9.90 7.20 4.20 9.22 18
    GAT year: 1983 5.90 2.40 6.20 5.90 9.30 12.90 17.30 16.20 12.70 9.80 7.40 5.90 9.32 18
    GAT year: 1984 3.00 4.00 4.60 7.60 9.40 13.40 15.50 16.10 12.60 10.30 7.50 5.50 9.12 18
    GAT year: 1985 1.30 3.00 4.40 7.70 9.90 11.80 14.80 13.70 13.20 10.60 3.90 5.90 8.35 18
    GAT year: 1986 3.30 0.00 4.90 5.30 10.20 13.50 14.50 12.60 10.80 10.20 7.30 5.60 8.18 18
    GAT year: 1987 1.70 3.80 4.20 9.00 9.50 11.70 14.90 14.50 12.70 9.10 6.60 5.70 8.62 34
    GAT year: 1988 4.90 4.60 5.80 7.60 10.90 13.70 14.00 14.30 12.60 10.10 5.90 7.40 9.32 34
    GAT year: 1989 6.70 5.70 6.70 6.10 11.70 13.30 16.60 15.20 13.20 11.10 6.70 4.50 9.79 34
    GAT year: 1990 6.20 6.50 7.70 7.50 11.70 12.80 15.50 16.40 12.20 11.10 6.70 4.50 9.90 34
    GAT year: 1991 3.50 2.50 7.20 7.40 10.30 11.30 16.10 15.80 13.70 9.70 6.60 5.20 9.11 32
    GAT year: 1992 4.20 5.60 6.80 8.00 12.20 14.60 15.00 14.30 12.40 7.50 7.00 4.00 9.30 16
    GAT year: 1993 5.50 5.50 6.20 8.60 10.40 13.40 14.10 13.60 11.80 8.10 5.20 4.70 8.92 16
    GAT year: 1994 4.70 3.40 6.60 7.40 9.80 13.10 16.20 14.80 12.20 9.70 9.70 6.10 9.47 16
    GAT year: 1995 4.50 5.90 5.00 8.10 10.60 13.40 16.70 17.50 12.90 12.20 7.70 2.60 9.76 16
    GAT year: 1996 4.90 3.10 4.50 7.90 8.40 13.10 15.00 15.40 13.10 11.00 5.50 3.40 8.78 15
    GAT year: 1997 3.20 6.00 7.70 8.40 10.50 12.90 15.60 17.30 13.10 9.90 8.50 6.00 9.92 15
    GAT year: 1998 5.20 7.40 7.20 7.00 11.60 12.60 14.10 14.40 13.50 9.40 6.00 5.50 9.49 14
    GAT year: 1999 4.90 4.70 6.60 8.50 11.20 12.40 15.50 14.70 14.30 10.30 7.50 4.20 9.57 12
    GAT year: 2000 5.00 5.60 6.80 6.90 10.80 12.70 14.10 15.00 13.60 9.80 6.60 5.30 9.35 11
    GAT year: 2001 3.70 3.80 4.50 6.90 11.30 12.40 14.90 14.90 12.50 12.50 7.40 4.10 9.07 11
    GAT year: 2002 5.80 5.90 6.70 8.40 10.90 13.30 14.20 15.40 13.40 9.10 8.10 5.50 9.72 11
    GAT year: 2003 4.60 4.40 7.10 9.10 10.70 14.40 16.10 16.20 13.20 8.70 8.00 5.10 9.80 11
    GAT year: 2004 5.00 5.00 6.20 8.70 11.00 13.80 14.30 16.00 13.60 9.80 7.70 5.90 9.75 11
    GAT year: 2005 6.00 4.40 6.70 8.00 9.80 13.60 14.90 14.60 13.60 11.70 6.40 5.30 9.58 10
    GAT year: 2006 5.00 4.50 4.20 7.40 10.30 13.90 17.10 14.90 15.00 12.00 7.80 6.20 9.86 10
    GAT year: 2007 6.20 5.60 6.70 10.10 10.50 13.30 14.20 14.20 12.60 10.80 7.50 5.30 9.75 10
    GAT year: 2008 5.70 5.60 5.50 7.20 12.00 12.70 15.10 15.10 12.70 9.10 6.80 4.20 9.31 10

  37. E.M.Smith says:

    Got me worked up a bit here… so I wanted to know what the stations were. A quick little script pulled these out of the v2.inv file:

    $ UK.2008 ../STEP0/input_files/v2.inv
    65103005000 LERWICK 60.13 -1.18 84 52R -9HIxxCO 1x-9WATER A 0
    65103026000 STORNOWAY 58.22 -6.32 13 2R -9FLxxCO 1A-9WARM GRASS/SHRUBB 7
    65103091000 ABERDEEN/DYCE 57.20 -2.22 65 98U 210HIxxCO 5A 3WARM CROPS C 26
    65103100000 TIREE 56.50 -6.88 12 4R -9HIxxCO 1x-9WATER A 0
    65103162000 ESKDALEMUIR 55.32 -3.20 242 293R -9HIxxno-9x-9WARM CROPS A 0
    65103257000 LEEMING 54.30 -1.53 40 42R -9HIxxno-9A-9WARM CROPS B 9
    65103302000 VALLEY 53.25 -4.53 11 33R -9HIxxCO 1A-9WARM CROPS B 7
    65103377000 WADDINGTON 53.17 -0.52 70 56U 74FLxxno-9A 3HEATHS, MOORS C 12
    65103862000 BOURNEMOUTH A 50.78 -1.83 11 41U 144HIxxCO 5A 2WARM CROPS C 14
    65103917000 BELFAST/ALDER 54.65 -6.22 81 68U 552FLxxCO20A15WARM CROPS B 13

    So, are those representative or not?

  38. E.M.Smith says:

    Curiouser and curiouser… Looking back to 1991 to see what is the difference between the larger number of station records then and the smaller number now, we find a large number that are duplicates in all but the “modification history” flag. that is the digit just before the year. See how it goes 0-1-0-1-0-1 etc down the list? While the first of the number stays the same for each 0-1 set? That is the exact same place, but a different “modification history”. Basically, they changed how they manipulated that data. Wonder what the change was… didn’t change the temperatures much. But at least they kept a record overlap…

    [chiefio@tubularbells analysis]$ grep 1991 v2.mean.england
    6510300500001991 41 29 56 59 78 87 132 131 103 83 60 52
    6510300500011991 41 29 56 59 78 87 132 131 103 83 60 52
    6510302600001991 45 37 66 69 93 97 148 143 115 87 66 67
    6510302600011991 45 37 66 69 93 97 148 143 115 87 66 67
    6510309100001991 24 23 59 69 100 107 151 154 120 91 57 40
    6510309100011991 24 23 59 69 100 107 151 154-9999 91 57-9999
    6510310000001991 53 43 71 77 100 107 152 146 127 98 74 75
    6510310000011991 53 43 71 77 100 107 152 146-9999 98 74-9999
    6510316000001991 24 25 64 75 110 115 158 157 128 93 61 43
    6510316000011991 24 25 64 75 110 115 158 157-9999 93 61-9999
    6510316200001991 9 8 56 56 91 95 151 139 111 77 45 26
    6510316200031991 9 8 56 56 91 95 151 139-9999 77 45-9999
    6510325700001991 26 22 76 77 106 118 172 168 140 102 64 44
    6510325700011991 26 22 76 77 106 118 172 168 140 102 64 44
    6510330200001991 48 37 77 86 110 121 167 163 151 113 86 71
    6510330200011991 48 37 77 86 110 121 167 163 151 113 86 71
    6510333400001991 29 24 80 81 110 118 176 168 147 105 68 53
    6510333400011991 29 24 80 81 110 118 176 168 147 105 68 53
    6510337700001991 28 18 79 77 106 120 180 177 148 102 64 44
    6510337700011991 28 18 79 77 106 120 180 177 148 102 64 44
    6510353400001991 30 16 80 77 110 121 174 170 147 101 65 40
    6510353400011991 30 16 80 77 110 121-9999 170 147 101 65 40
    6510371500001991 40 22 79 82 113 123 166 168 154 103 76 54
    6510371500011991 40 22 79 82 113 123-9999 168 154 103 76 54
    6510377600001991 43 14 84 82 105 127 176 179 152 102 70 41
    6510377600011991 43 14 84 82 105 127 176-9999 152-9999-9999 41
    6510382700001991 58 38-9999-9999-9999-9999-9999 166 159 112 86 73
    6510382700011991 58 38-9999-9999-9999-9999-9999-9999-9999-9999-9999-9999
    6510386200001991 44 18 80 79 113 126 167 172 151 105 74 55
    6510386200011991 44 18 80 79 113 126 167-9999-9999-9999-9999-9999
    6510391700001991 33 30 74 77 109 117 165 160 135 96 64 57
    6510391700031991 33 30 74 77 109 117 165 160-9999 96 64-9999

    The stations are (with dups left in from the mod. flag):

    65103005000 LERWICK 60.13 -1.18 84 52R -9HIxxCO 1x-9WATER A 0
    65103005000 LERWICK 60.13 -1.18 84 52R -9HIxxCO 1x-9WATER A 0
    65103026000 STORNOWAY 58.22 -6.32 13 2R -9FLxxCO 1A-9WARM GRASS/SHRUBB 7
    65103026000 STORNOWAY 58.22 -6.32 13 2R -9FLxxCO 1A-9WARM GRASS/SHRUBB 7
    65103091000 ABERDEEN/DYCE 57.20 -2.22 65 98U 210HIxxCO 5A 3WARM CROPS C 26
    65103091000 ABERDEEN/DYCE 57.20 -2.22 65 98U 210HIxxCO 5A 3WARM CROPS C 26
    65103100000 TIREE 56.50 -6.88 12 4R -9HIxxCO 1x-9WATER A 0
    65103100000 TIREE 56.50 -6.88 12 4R -9HIxxCO 1x-9WATER A 0
    65103160000 EDINBURGH AIR 55.95 -3.35 41 51U 470HIxxCO 4A 2WARM GRASS/SHRUBC 29
    65103160000 EDINBURGH AIR 55.95 -3.35 41 51U 470HIxxCO 4A 2WARM GRASS/SHRUBC 29
    65103162000 ESKDALEMUIR 55.32 -3.20 242 293R -9HIxxno-9x-9WARM CROPS A 0
    65103162000 ESKDALEMUIR 55.32 -3.20 242 293R -9HIxxno-9x-9WARM CROPS A 0
    65103257000 LEEMING 54.30 -1.53 40 42R -9HIxxno-9A-9WARM CROPS B 9
    65103257000 LEEMING 54.30 -1.53 40 42R -9HIxxno-9A-9WARM CROPS B 9
    65103302000 VALLEY 53.25 -4.53 11 33R -9HIxxCO 1A-9WARM CROPS B 7
    65103302000 VALLEY 53.25 -4.53 11 33R -9HIxxCO 1A-9WARM CROPS B 7
    65103334000 MANCHESTER AI 53.35 -2.27 78 50U 490FLxxno-9A 1HEATHS, MOORS C 22
    65103334000 MANCHESTER AI 53.35 -2.27 78 50U 490FLxxno-9A 1HEATHS, MOORS C 22
    65103377000 WADDINGTON 53.17 -0.52 70 56U 74FLxxno-9A 3HEATHS, MOORS C 12
    65103377000 WADDINGTON 53.17 -0.52 70 56U 74FLxxno-9A 3HEATHS, MOORS C 12
    65103534000 BIRMINGHAM/AI 52.45 -1.73 99 112U 1059HIxxno-9A 1WARM CROPS C 25
    65103534000 BIRMINGHAM/AI 52.45 -1.73 99 112U 1059HIxxno-9A 1WARM CROPS C 25
    65103715000 GLAMORGAN/RHOUSE AP 51.40 -3.40 67 26U 282HIxxCO 2A 3WATER B 13
    65103715000 GLAMORGAN/RHOUSE AP 51.40 -3.40 67 26U 282HIxxCO 2A 3WATER B 13
    65103776000 LONDON/GATWIC 51.15 -0.18 62 73U12332HIxxno-9A15WARM CROPS C 29
    65103776000 LONDON/GATWIC 51.15 -0.18 62 73U12332HIxxno-9A15WARM CROPS C 29
    65103827000 PLYMOUTH WC 50.35 -4.12 50 76U 259HIxxCO 1x-9WATER C 19
    65103827000 PLYMOUTH WC 50.35 -4.12 50 76U 259HIxxCO 1x-9WATER C 19
    65103862000 BOURNEMOUTH A 50.78 -1.83 11 41U 144HIxxCO 5A 2WARM CROPS C 14
    65103862000 BOURNEMOUTH A 50.78 -1.83 11 41U 144HIxxCO 5A 2WARM CROPS C 14
    65103917000 BELFAST/ALDER 54.65 -6.22 81 68U 552FLxxCO20A15WARM CROPS B 13
    65103917000 BELFAST/ALDER 54.65 -6.22 81 68U 552FLxxCO20A15WARM CROPS B 13

    I suppose they changed the kind of thermometer maybe? Moved from glass to electronic?

  39. E.M.Smith says:

    Hacking together a new bit of code… this totals the number of stations (location and substation) by year regardless of modification flag. Here is the result for the UK. The format is YEAR, 12 monthly averages of daily min-max averages, the “annual average” of all that data, and the count of thermometer locations used for that data:

    1763 -0.7  6.0  5.2  8.1  9.1 14.0 15.3 15.6 13.1  8.8  6.2  6.8  9.0  1
    1764  4.1  4.1  4.3  7.0 11.8 13.3 15.7 14.9 11.6  8.4  4.7  2.8  8.6  2
    1765  5.2  0.8  5.3  7.6 10.9 13.0 14.7 14.9 12.6  9.4  3.7  1.8  8.3  2
    1766  1.1  2.2  4.1  8.1  8.6 13.4 15.4 16.0 13.0  9.3  7.6  3.5  8.5  2
    1767  0.4  5.9  5.0  6.7  9.7 11.7 13.9 15.8 13.7  9.1  7.2  3.8  8.6  2
    1768  1.4  5.1  4.7  8.1 11.0 13.5 14.9 15.7 11.6  9.3  5.6  4.8  8.8  2
    1769  3.0  3.0  5.0  7.5 10.4 12.4 15.5 14.5 12.7  7.8  5.8  5.3  8.6  2
    1770  4.3  5.1  2.9  5.5  9.5 12.6 15.1 15.6 13.7  8.2  5.0  4.2  8.5  2
    1771  1.0  3.0  2.7  5.3 11.5 13.5 15.4 14.6 11.8  9.3  6.0  6.0  8.3  2
    1772  0.7  1.5  3.7  6.2  9.6 15.3 16.0 15.7 12.3 10.7  6.5  4.3  8.5  2
    1773  3.6  2.3  5.7  7.7  9.5 13.7 15.2 16.3 12.2  9.1  4.3  3.1  8.6  2
    1774 -0.4  3.2  4.7  7.5  9.4 14.2 15.3 15.8 12.0  9.4  4.3  3.0  8.2  2
    1775  4.2  5.3  5.2  9.4 12.4 15.2 16.7 15.2 13.5  8.5  4.2  4.0  9.5  2
    1776 -1.8  3.6  6.1  8.6 10.3 13.8 16.4 15.0 12.0 10.0  5.6  4.3  8.7  2
    1777  1.8  2.0  5.8  6.7 11.5 13.2 15.4 16.2 14.1 10.1  6.4  3.2  8.9  2
    1778  2.7  3.4  4.8  7.6 12.5 16.1 17.9 16.3 11.4  7.0  6.1  6.3  9.3  2
    1779  2.8  8.1  8.1  9.5 11.8 14.7 19.0 18.7 14.7 10.3  5.3  2.4 10.4  2
    1780 -1.4  2.0  8.1  6.1 12.8 14.7 16.9 18.1 14.8  8.9  4.1  3.7  9.1  2
    1781  2.5  4.9  6.9  8.9 11.5 16.3 16.6 15.6 12.4  9.2  6.1  5.1  9.7  3
    1782  3.8  1.3  3.1  4.4  7.5 14.2 15.2 14.5 11.6  6.5  2.2  2.7  7.2  3
    1783  2.8  3.7  2.9  9.5  9.3 13.7 18.1 14.9 11.9  8.6  6.0  2.9  8.7  3
    1784  0.0  1.3  2.0  5.5 13.5 13.1 15.3 13.8 13.3  7.6  4.6  0.5  7.5  3
    1785  2.8 -0.4  0.6  9.0 10.7 15.9 15.5 13.0 12.9  7.7  5.8  2.2  8.0  3
    1786  2.7  2.8  1.7  7.3 10.3 14.6 14.0 14.8 10.7  6.8  3.6  2.5  7.7  3
    1787  4.1  6.0  6.8  6.8 10.4 13.0 15.3 15.4 12.5  9.1  3.7  3.0  8.8  3
    1788  3.8  3.2  3.1  9.7 12.0 14.3 16.2 15.3 13.0  8.8  5.8  0.2  8.8  3
    1789  1.6  4.7  1.7  7.2 11.9 13.8 15.9 16.6 12.8  8.8  4.9  6.3  8.9  3
    1790  4.2  6.7  6.7  6.0 11.6 14.3 14.7 15.0 11.4  9.5  4.9  3.7  9.1  3
    1791  4.3  4.1  6.6  8.8 10.8 13.6 15.3 15.4 13.1  8.5  5.3  0.9  8.9  3
    1792  2.0  4.6  5.6 10.1  9.8 12.6 15.2 16.1 11.2  8.5  7.2  4.0  8.9  3
    1793  3.0  4.4  3.7  5.8 10.4 13.0 16.7 14.6 11.9 11.1  5.8  5.3  8.8  3
    1794  2.1  6.3  6.7  9.6 10.3 15.1 17.3 14.7 12.0  8.9  5.6  3.7  9.4  4
    1795 -2.0  0.3  3.7  7.9  9.6 12.4 15.0 16.1 15.2 11.2  4.0  6.5  8.3  4
    1796  6.8  4.5  4.4 10.4 10.1 13.5 15.2 15.6 14.0  7.8  4.4  0.0  8.9  4
    1797  4.1  5.2  4.5  7.3 11.3 12.7 16.4 15.0 12.2  8.0  4.8  4.5  8.8  4
    1798  3.8  3.7  4.8 10.4 12.5 16.3 15.9 15.7 12.5  8.9  4.4  3.2  9.3  4
    1799  1.9  2.2  3.0  5.0  8.9 13.0 14.4 13.3 12.2  7.6  4.9  1.2  7.3  4
    1800  2.1  1.9  3.6  8.7 11.2 12.3 16.2 15.9 13.1  8.7  4.9  2.9  8.5  4
    1801  3.9  4.0  6.0  8.0 11.3 13.8 14.6 16.3 13.8  9.9  4.4  1.4  8.9  4
    1802  1.6  3.3  5.5  8.3  9.9 13.2 13.4 16.2 13.3  9.5  5.5  3.5  8.6  4
    1803  1.6  2.8  5.5  8.5  9.8 12.8 16.8 15.1 10.7  8.7  4.3  3.4  8.3  4
    1804  5.1  2.4  3.9  5.9 12.7 15.0 14.5 14.7 13.1  9.8  5.4  1.9  8.7  4
    1805  2.0  3.3  5.9  7.6  9.2 12.2 15.2 15.6 13.6  7.8  5.3  3.1  8.4  4
    1806  2.8  3.5  4.2  6.1 11.1 14.0 14.8 15.2 12.9  9.7  6.7  5.5  8.9  4
    1807  2.5  2.7  2.7  7.0 10.6 13.2 16.4 16.2  9.6 10.4  2.1  2.0  8.0  4
    1808  2.2  2.2  3.0  5.5 13.2 14.0 17.5 15.9 12.3  6.6  5.3  2.1  8.3  4
    1809  0.3  4.9  5.8  4.8 12.3 13.0 14.3 14.5 11.9 10.0  4.3  3.3  8.3  4
    1810  2.0  2.8  3.6  7.7  8.5 14.0 14.7 14.7 13.5  9.3  4.8  2.7  8.2  4
    1811  1.0  3.7  6.8  8.0 12.1 13.6 15.3 14.2 13.2 11.6  7.2  2.8  9.1  4
    1812  2.5  4.8  3.0  5.1 10.4 12.8 13.6 14.3 12.8  8.6  4.5  1.8  7.9  4
    1813  2.0  5.1  6.6  7.4 11.0 13.5 15.4 14.7 12.5  7.5  3.9  2.9  8.5  4
    1814 -2.9  1.3  2.8  9.3  9.0 11.9 15.4 14.5 12.6  7.6  3.9  2.8  7.4  4
    1815 -0.1  5.0  5.7  7.5 11.4 13.5 14.6 14.6 12.5  9.3  3.4  1.6  8.2  4
    1816  2.0  1.8  3.3  5.7  9.4 12.4 13.6 13.3 11.5  9.0  3.6  2.2  7.3  4
    1817  3.9  5.3  4.6  6.9  8.3 14.0 13.9 13.0 12.9  5.9  7.8  1.5  8.2  4
    1818  3.2  1.8  3.3  5.9 11.0 16.1 17.3 14.8 12.7 11.5  8.6  3.6  9.1  4
    1819  3.5  3.0  6.0  7.8 11.0 13.3 16.2 17.6 13.0  8.3  3.5  1.1  8.7  4
    1820 -0.6  3.4  4.7  8.6 11.0 12.9 15.4 14.5 12.0  7.3  5.3  4.2  8.2  4
    1821  3.0  3.0  5.4  9.3  9.1 12.0 14.6 15.2 14.3  9.9  7.0  5.3  9.0  4
    1822  4.1  5.3  7.2  8.0 12.1 16.2 15.3 15.0 11.7  9.8  7.4  1.8  9.5  4
    1823 -0.1  2.0  4.7  6.3 11.8 12.1 13.9 14.6 12.1  7.9  7.0  4.0  8.0  4
    1824  4.2  4.2  4.3  7.5 10.2 13.5 15.9 15.2 13.1  8.3  5.7  3.9  8.8  4
    1825  3.6  3.6  4.8  8.4 11.0 13.8 17.0 16.4 14.6 10.0  4.1  4.1  9.3  4
    1826  0.4  5.8  5.5  8.0 11.3 15.9 17.5 17.0 13.3 10.5  4.1  5.3  9.6  4
    1827  1.4  0.8  4.8  7.8 11.1 13.6 15.6 14.0 13.2 10.7  6.0  6.5  8.8  5
    1828  4.6  4.8  6.2  7.8 11.2 14.8 15.8 14.9 14.0 10.1  7.6  7.1  9.9  5
    1829  0.6  4.2  4.5  6.3 12.0 14.3 15.1 13.6 11.1  8.3  4.7  2.1  8.1  5
    1830  0.8  2.4  7.4  8.3 11.6 12.3 15.7 14.5 12.1 10.5  6.7  2.1  8.7  5
    1831  1.9  4.6  6.7  8.5 10.7 14.7 16.2 16.3 13.4 12.3  5.5  6.0  9.7  5
    1832  3.6  4.1  5.6  8.3 10.4 14.4 15.2 15.1 13.2 10.5  6.4  5.3  9.3  5
    1833  1.8  5.1  3.7  7.4 14.2 13.9 15.3 14.1 11.9 10.1  6.1  6.1  9.1  5
    1834  6.1  5.3  6.7  7.5 12.4 14.9 16.4 15.6 13.3 10.3  6.8  6.0 10.1  6
    1835  3.7  5.2  5.5  7.7 10.4 13.7 15.5 15.7 12.8  8.4  6.9  3.7  9.1  6
    1836  3.8  3.3  5.2  6.5 11.0 14.3 14.6 14.1 10.9  8.4  5.1  4.1  8.4  6
    1837  3.2  4.7  2.3  4.4  9.2 14.3 15.8 14.7 12.2 10.4  5.4  5.6  8.5  6
    1838 -0.4  0.1  4.8  5.4  9.7 13.2 15.1 14.5 12.2  9.3  4.8  4.6  7.8  6
    1839  2.7  3.7  3.8  6.3  9.6 13.4 15.0 14.1 12.4  9.2  6.9  4.1  8.4  6
    1840  4.1  3.4  4.4  9.4 10.6 13.3 13.9 15.3 11.0  7.9  6.0  2.3  8.5  6
    1841  1.1  2.8  7.5  7.4 11.8 12.0 13.0 14.0 13.0  7.8  4.9  4.3  8.3  5
    1842  1.5  4.4  5.9  7.6 11.0 14.6 14.0 16.1 12.7  7.6  5.3  7.2  9.0  5
    1843  4.3  1.5  5.4  7.6  9.5 11.9 14.4 14.9 14.2  7.8  5.9  7.6  8.8  5
    1844  4.4  1.9  5.0  9.6 10.3 13.4 14.4 13.2 12.4  8.8  6.4  1.6  8.4  5
    1845  3.5  1.9  2.5  7.9  9.1 14.0 13.5 13.1 11.4  9.5  6.8  4.3  8.1  5
    1846  6.2  6.5  5.7  7.3 11.6 16.8 15.8 15.9 14.6  9.5  7.3  1.5  9.9  5
    1847  2.3  2.1  5.2  6.2 11.2 13.1 16.6 14.8 10.8 10.0  7.6  4.8  8.7  5
    1848  1.4  5.1  5.3  7.0 13.1 13.0 14.8 12.8 12.1  8.9  5.4  5.3  8.7  6
    1849  3.7  5.5  5.7  6.0 11.0 12.7 14.5 14.5 12.6  8.5  6.0  3.6  8.7  7
    1850  0.8  6.1  4.6  7.9  9.2 14.2 15.1 13.8 11.6  7.6  6.7  4.6  8.5  7
    1851  6.0  4.9  5.8  7.2 10.2 13.8 14.4 15.3 12.8 10.7  3.9  5.5  9.2  10
    1852  5.3  4.9  5.2  7.8 10.5 12.9 18.2 15.8 12.9  8.6  8.0  7.7  9.8  10
    1853  5.4  1.1  3.6  7.6 10.4 13.8 14.9 14.5 12.4 10.2  6.5  2.7  8.6  9
    1854  3.9  4.6  6.9  8.6 10.1 12.7 15.1 15.3 14.3  9.6  5.5  5.3  9.3  9
    1855  2.7 -0.4  3.6  7.3  8.6 13.0 16.2 15.7 13.4 10.1  5.8  3.5  8.3  10
    1856  3.8  5.2  4.3  7.6  8.9 13.2 14.8 15.8 12.0 10.8  5.5  4.9  8.9  11
    1857  3.0  4.5  5.0  6.9 10.7 14.8 15.9 16.7 14.5 11.2  7.8  7.8  9.9  12
    1858  4.3  2.6  4.8  7.5 10.1 15.7 14.2 15.1 14.0  9.3  4.8  5.4  9.0  12
    1859  5.0  5.5  6.9  6.6 11.0 14.1 17.1 15.9 12.8  9.4  5.5  2.2  9.3  13
    1860  3.6  1.9  4.9  6.1 11.4 11.8 13.9 13.1 11.1  9.5  5.0  2.3  7.9  13
    1861  2.2  5.0  6.0  7.5 10.4 14.3 14.7 15.4 13.0 11.6  4.6  4.5  9.1  13
    1862  4.2  5.0  5.1  8.1 11.6 12.2 13.4 14.0 12.9 10.0  4.0  6.5  8.9  13
    1863  4.8  5.8  6.3  8.1  9.9 13.0 14.6 14.7 11.1  9.8  7.3  6.1  9.3  13
    1864  2.9  1.9  4.4  8.5 11.4 12.9 14.8 13.7 12.7  9.3  5.9  4.1  8.5  13
    1865  2.7  2.4  3.1  9.8 11.5 14.9 15.9 14.6 16.0 10.0  6.8  6.4  9.5  14
    1866  5.7  4.3  4.3  7.8  9.5 14.5 15.2 14.1 12.2 10.5  7.0  6.4  9.3  14
    1867  1.4  6.7  3.2  8.7 10.4 13.6 14.2 15.4 13.3  9.4  5.8  4.3  8.9  14
    1868  3.7  6.3  6.7  8.5 12.4 14.7 17.7 15.8 13.9  8.6  5.2  7.1 10.1  14
    1869  5.7  7.1  3.7  9.5  9.0 12.6 16.3 14.8 13.6  9.8  6.4  3.1  9.3  14
    1870  3.6  2.6  4.8  8.8 11.1 14.5 16.7 15.5 13.2  9.7  5.3  1.4  8.9  15
    1871  1.2  6.0  7.0  7.9 10.7 12.5 15.2 16.6 12.7  9.7  4.1  3.9  9.0  20
    1872  4.9  6.4  6.5  8.2  9.6 13.9 16.5 15.0 12.8  8.5  6.6  5.2  9.5  22
    1873  5.1  2.2  5.2  7.6  9.7 14.0 15.9 15.3 11.9  8.3  6.4  5.7  8.9  22
    1874  5.5  4.3  6.7  9.3  9.5 13.6 16.5 14.7 13.3 10.0  6.0  0.9  9.2  22
    1875  5.9  2.6  4.9  8.1 11.6 13.7 14.4 15.7 14.3  9.2  5.4  4.1  9.2  22
    1876  3.6  4.5  4.4  7.7  9.5 13.7 16.7 15.7 12.4 11.0  6.3  6.0  9.3  22
    1877  5.3  5.9  4.6  6.8  8.7 14.5 14.6 14.6 11.3  9.3  7.1  5.0  9.0  23
    1878  4.6  5.7  5.5  8.4 11.4 14.4 16.3 15.9 13.3 10.4  4.0  0.4  9.2  23
    1879  0.2  3.0  4.3  5.7  8.6 12.6 13.5 14.3 12.2  9.0  4.6  1.4  7.5  24
    1880  2.1  5.7  6.3  7.9 10.2 13.2 14.9 16.2 14.3  7.3  5.4  5.1  9.1  24
    1881 -0.7  2.9  4.8  6.8 11.2 13.0 15.7 13.5 12.4  7.7  8.6  4.3  8.3  24
    1882  5.3  6.0  7.2  7.8 11.0 12.7 14.8 14.8 11.9  9.8  5.6  3.3  9.2  25
    1883  4.6  5.7  2.4  7.7 10.1 13.2 14.2 14.9 12.9  9.5  6.0  4.9  8.8  25
    1884  5.7  4.8  5.7  6.6 10.3 12.9 15.0 15.9 13.5  9.0  5.4  4.0  9.1  27
    1885  2.6  5.0  4.2  7.0  8.1 12.8 15.1 13.0 11.5  6.9  5.7  3.8  8.0  27
    1886  1.7  1.4  3.6  6.8  9.5 12.4 14.7 14.7 12.6 10.5  6.4  1.9  8.0  27
    1887  2.6  4.0  3.5  5.9  9.1 14.7 16.3 14.7 11.3  7.0  4.5  2.8  8.0  28
    1888  3.5  1.7  2.7  5.7  9.9 12.3 12.9 13.3 11.9  8.1  7.2  5.0  7.8  28
    1889  3.6  2.8  4.4  6.5 12.0 14.6 14.3 14.1 12.2  8.2  6.7  3.7  8.6  29
    1890  5.6  3.3  5.7  6.8 11.2 13.0 13.7 13.8 14.3  9.6  5.8  0.6  8.6  29
    1891  1.9  5.0  3.6  5.7  8.9 13.8 14.3 13.6 13.5  9.1  5.4  4.3  8.3  29
    1892  2.3  3.3  2.6  6.9 10.6 12.5 13.6 14.4 11.6  6.7  6.5  2.2  7.8  29
    1893  2.4  4.3  6.9  9.8 12.1 14.7 15.3 16.4 12.3  9.5  4.9  4.8  9.5  29
    1894  3.0  4.5  6.3  8.9  8.5 12.8 15.0 13.8 11.4  8.7  7.5  5.0  8.8  29
    1895  0.2 -1.3  4.9  7.8 11.4 13.8 14.4 14.9 14.9  7.0  7.1  3.9  8.2  29
    1896  4.6  4.8  6.2  8.5 11.6 14.8 15.3 13.6 12.4  6.6  4.8  3.9  8.9  29
    1897  1.6  5.2  5.8  6.4  9.3 13.8 15.6 15.5 11.5  9.8  7.4  4.7  8.9  29
    1898  6.3  4.3  4.1  7.8  9.3 12.8 14.5 15.3 14.5 10.9  6.7  6.7  9.4  30
    1899  4.2  4.7  4.8  7.0  9.1 14.6 16.2 16.8 12.6  9.2  8.4  2.6  9.2  30
    1900  4.2  1.9  3.3  7.7  9.6 13.7 16.3 14.3 13.0  9.1  6.8  6.8  8.9  30
    1901  3.3  2.1  3.7  7.7 10.9 12.9 16.9 14.9 13.2  8.8  5.0  3.1  8.5  27
    1902  4.1  1.3  6.0  6.9  7.9 12.5 13.5 13.2 12.3  9.0  6.8  4.4  8.2  27
    1903  3.5  6.2  6.1  5.7 10.0 12.1 14.1 13.3 12.3  9.5  6.0  3.0  8.5  27
    1904  4.1  3.2  4.2  8.4 10.2 13.0 16.0 14.7 12.4  9.7  5.6  4.1  8.8  25
    1905  4.0  4.9  6.5  6.8 10.5 14.1 16.7 14.3 12.2  7.2  5.0  5.3  9.0  25
    1906  5.1  3.0  4.7  6.9  9.8 13.7 15.1 16.1 13.6 10.4  7.4  3.2  9.1  25
    1907  3.7  3.0  6.4  7.4  9.9 12.1 13.8 13.9 13.4  9.7  6.5  4.6  8.7  24
    1908  2.9  5.3  4.2  5.9 12.0 13.9 15.4 14.3 12.8 12.0  7.4  4.3  9.2  24
    1909  3.6  3.4  3.4  8.4 10.5 11.8 14.4 15.3 11.8 10.0  4.9  3.6  8.4  24
    1910  3.4  4.7  6.1  6.9 10.7 14.1 13.9 14.9 12.5 10.7  3.5  6.3  9.0  24
    1911  4.1  4.7  5.0  7.4 12.5 14.0 17.3 17.5 13.6  9.1  5.9  5.9  9.8  24
    1912  3.6  5.1  6.9  8.7 11.2 13.3 15.4 12.5 10.9  8.3  6.1  6.2  9.0  24
    1913  4.1  4.8  5.7  7.5 10.9 13.7 14.4 14.9 13.6 10.8  8.1  4.9  9.4  24
    1914  3.7  6.5  5.7  9.5 10.4 14.1 15.6 15.8 13.1 10.3  6.7  4.3  9.6  24
    1915  3.8  3.9  5.1  7.6 10.2 13.8 14.4 15.1 13.1  9.0  3.3  4.8  8.7  24
    1916  7.1  3.6  3.2  7.9 10.9 11.4 14.8 15.9 12.8 10.2  6.8  2.6  8.9  24
    1917  1.9  1.7  3.1  5.1 11.8 14.4 15.6 15.2 13.7  7.4  7.9  2.7  8.4  24
    1918  3.5  6.3  5.8  6.6 12.3 13.0 15.1 15.6 11.4  9.2  5.6  6.5  9.2  24
    1919  3.0  2.0  3.4  7.1 12.4 13.7 13.8 15.4 12.6  7.9  3.0  5.0  8.3  23
    1920  4.7  5.9  6.8  7.7 11.3 13.8 13.8 13.4 12.7 10.4  7.1  4.2  9.3  24
    1921  6.7  5.0  6.8  8.1 11.1 14.1 17.4 14.8 13.9 12.6  5.0  6.5 10.2  24
    1922  3.6  4.4  4.7  5.4 12.1 13.3 13.3 13.2 11.8  8.4  6.0  5.6  8.5  24
    1923  5.6  5.2  6.3  7.1  8.9 12.2 16.7 14.6 12.0  9.4  3.5  3.6  8.8  24
    1924  4.5  3.5  3.9  6.8 10.7 13.4 14.8 13.9 12.8  9.9  7.1  6.8  9.0  24
    1925  5.2  4.7  4.8  7.0 11.0 14.4 16.4 15.1 11.2 10.2  3.9  2.7  8.9  24
    1926  4.5  6.3  6.2  8.9  9.7 13.2 16.5 15.9 14.0  7.8  5.9  4.5  9.5  24
    1927  4.5  4.4  6.9  7.5 10.3 11.8 15.3 15.2 12.0 10.1  6.0  2.0  8.8  24
    1928  4.9  5.6  5.6  7.8 10.3 12.2 15.6 14.9 12.4  9.8  7.5  3.6  9.2  24
    1929  1.8  0.7  6.4  6.3 10.9 12.8 15.6 14.8 15.0  9.3  6.6  5.4  8.8  24
    1930  5.1  2.5  4.9  7.8 10.3 14.7 14.9 15.4 13.3 10.1  6.0  4.3  9.1  24
    1931  3.4  3.7  3.8  7.2 10.6 13.2 14.7 13.8 11.4  9.0  7.8  5.6  8.7  32
    1932  6.2  4.0  4.8  6.4  9.7 13.3 15.3 16.1 12.6  8.5  6.6  5.7  9.1  32
    1933  2.8  4.1  7.0  8.5 11.1 14.6 17.0 16.7 14.5 10.0  6.0  2.7  9.6  32
    1934  4.7  4.6  4.8  7.2 10.6 14.0 17.1 14.8 13.9 10.0  6.2  7.7  9.6  32
    1935  4.8  5.4  6.4  7.4  9.5 13.9 16.1 15.9 13.0  9.2  6.6  3.1  9.3  32
    1936  3.5  2.7  6.4  6.1 10.7 13.8 14.9 15.5 13.9  9.3  5.9  5.3  9.0  32
    1937  5.0  4.9  3.2  8.4 11.3 13.4 15.2 15.9 12.9 10.2  5.8  3.1  9.1  32
    1938  5.5  5.0  8.8  7.6  9.9 13.4 14.4 15.3 13.4 10.1  8.7  4.4  9.7  32
    1939  3.9  5.6  5.8  8.0 10.7 13.6 14.7 15.8 13.8  8.4  8.1  3.7  9.3  32
    1940  0.0  2.7  5.6  7.8 11.8 15.6 14.4 14.9 12.2  9.5  6.8  4.2  8.8  32
    1941  0.8  3.1  4.5  6.1  8.8 13.9 16.2 14.1 14.2 10.1  6.5  5.7  8.7  32
    1942  1.6  0.8  4.6  8.4 10.3 13.4 14.7 15.6 13.0 10.1  5.6  6.5  8.7  32
    1943  4.6  6.0  6.6  9.9 11.1 13.6 15.3 14.9 12.7 10.6  6.3  4.3  9.7  32
    1944  5.9  3.7  5.2  9.3 10.5 12.8 15.5 16.1 12.0  9.1  5.9  4.1  9.2  32
    1945  0.7  6.7  7.9  9.3 11.2 13.7 15.9 15.2 13.9 11.5  7.6  5.3  9.9  32
    1946  3.4  5.6  5.2  9.2 10.1 12.6 15.4 14.0 13.4  9.6  7.7  3.5  9.1  32
    1947  2.6 -1.1  2.9  7.9 12.0 14.4 16.0 17.8 14.2 10.7  6.8  5.2  9.1  32
    1948  4.6  4.6  7.9  8.6 10.6 12.9 14.9 14.2 13.2  9.8  7.4  5.7  9.5  32
    1949  5.2  5.5  4.9  9.3 10.4 14.0 16.3 15.9 15.5 11.3  6.6  5.4 10.0  31
    1950  4.5  4.8  7.3  7.0 10.6 15.1 15.3 15.1 12.4  9.5  5.5  1.4  9.0  37
    1951  3.5  3.3  3.8  6.2  9.3 13.0 15.5 14.4 13.6  9.9  8.0  5.3  8.8  43
    1952  2.4  3.4  6.2  9.0 12.5 13.6 16.0 15.4 10.7  8.8  4.1  3.1  8.8  45
    1953  3.7  4.4  5.7  6.9 12.0 13.6 15.0 15.5 13.6  9.9  8.2  6.8  9.6  45
    1954  3.1  2.4  5.5  7.3 10.8 12.9 13.8 14.1 12.2 11.3  6.7  6.2  8.9  47
    1955  2.4  1.1  3.2  9.0  9.2 13.1 16.8 17.2 13.9  9.0  7.3  5.3  9.0  47
    1956  3.3  0.2  5.8  6.3 11.5 12.5 15.1 13.0 13.7  9.4  6.2  5.8  8.6  45
    1957  5.2  4.8  8.6  8.4 10.0 14.4 15.7 14.9 12.0 10.5  6.5  4.6  9.6  45
    1958  3.2  4.3  3.3  7.1 10.5 13.1 15.3 15.3 14.6 10.7  6.7  4.5  9.0  48
    1959  1.7  4.5  7.0  9.0 11.9 14.5 16.6 16.8 14.7 12.3  7.0  5.7 10.1  48
    1960  3.9  3.7  6.0  8.7 12.2 15.3 14.7 14.6 12.9 10.2  7.0  3.9  9.4  47
    1961  3.5  6.6  8.0  9.3 10.4 13.8 14.6 14.9 14.5 10.5  6.0  2.4  9.5  47
    1962  4.1  4.4  2.6  7.2  9.7 13.0 14.1 14.0 12.2 10.4  5.5  2.4  8.3  47
    1963 -1.3 -0.4  5.7  8.0  9.9 14.1 14.4 13.8 12.5 10.5  7.5  2.9  8.1  46
    1964  3.5  4.3  4.2  8.2 12.4 13.2 15.3 14.8 13.5  8.7  7.2  3.5  9.1  46
    1965  3.2  3.3  4.9  7.6 10.8 13.6 13.2 14.2 12.0 10.6  4.4  4.0  8.5  43
    1966  2.9  4.9  6.3  6.6 10.5 14.6 14.3 14.1 13.6  9.8  5.3  4.8  9.0  42
    1967  4.3  5.3  6.7  7.5  9.7 13.6 16.0 15.1 13.3 10.3  5.6  4.2  9.3  43
    1968  4.2  1.8  6.0  7.5  9.0 14.0 14.2 14.8 13.3 11.9  6.4  3.4  8.9  36
    1969  5.1  0.9  3.2  7.0 10.3 13.3 15.9 15.7 13.4 12.4  4.8  3.6  8.8  35
    1970  3.8  2.7  3.8  6.3 11.8 15.4 14.2 15.3 13.6 10.3  7.2  4.4  9.1  25
    1971  4.7  4.9  5.3  7.5 10.9 11.8 15.7 14.8 13.6 11.1  6.1  6.8  9.4  26
    1972  4.0  4.4  6.1  8.1 10.0 11.3 14.8 14.2 11.3 10.3  5.9  5.7  8.8  25
    1973  4.7  4.1  6.1  6.5 10.4 14.0 14.9 15.4 13.4  8.9  5.7  4.7  9.1  25
    1974  5.9  5.4  5.3  7.7 10.4 12.9 14.2 14.4 11.4  7.6  6.4  7.2  9.1  25
    1975  6.1  4.7  4.6  7.8  9.2 13.7 16.2 17.2 12.4  9.9  6.1  5.3  9.4  25
    1976  5.3  4.7  4.7  7.7 10.9 15.4 17.1 16.3 12.7 10.1  6.0  2.1  9.4  22
    1977  2.7  4.4  6.5  6.8  9.8 11.8 15.2 14.5 12.3 11.3  5.8  5.9  8.9  22
    1978  3.0  2.2  6.2  6.1 11.0 12.9 13.9 14.2 13.2 11.4  8.1  4.1  8.9  22
    1979  0.4  1.6  4.0  7.0  8.7 13.1 14.9 13.8 12.3 10.8  6.4  5.2  8.2  19
    1980  2.4  5.2  4.4  8.3 10.3 12.9 13.7 14.8 13.8  8.5  6.4  5.3  8.8  19
    1981  4.5  3.4  6.8  7.3 10.5 12.4 14.4 15.1 13.6  7.4  7.1  1.0  8.6  17
    1982  3.0  5.1  5.7  8.2 10.4 13.9 15.4 14.6 13.1  9.9  7.2  4.2  9.2  16
    1983  5.9  2.4  6.2  5.9  9.3 12.9 17.3 16.2 12.7  9.8  7.4  5.9  9.3  16
    1984  3.0  4.0  4.6  7.6  9.4 13.4 15.5 16.1 12.6 10.3  7.5  5.5  9.1  16
    1985  1.3  3.0  4.4  7.7  9.9 11.8 14.8 13.7 13.2 10.6  3.9  5.9  8.3  16
    1986  3.3  0.0  4.9  5.3 10.2 13.5 14.5 12.6 10.8 10.2  7.3  5.6  8.2  16
    1987  1.7  3.8  4.2  9.0  9.5 11.7 14.9 14.5 12.7  9.1  6.6  5.7  8.6  16
    1988  4.9  4.6  5.8  7.6 10.9 13.7 14.0 14.3 12.6 10.1  5.9  7.4  9.3  16
    1989  6.7  5.7  6.7  6.1 11.7 13.3 16.6 15.2 13.2 11.1  6.7  4.5  9.8  16
    1990  6.2  6.5  7.7  7.5 11.7 12.8 15.5 16.4 12.2 11.1  6.7  4.5  9.9  16
    1991  3.5  2.5  7.2  7.4 10.3 11.3 16.1 15.8 13.7  9.7  6.6  5.2  9.1  16
    1992  4.2  5.6  6.8  8.0 12.2 14.6 15.0 14.3 12.4  7.5  7.0  4.0  9.3  16
    1993  5.5  5.5  6.2  8.6 10.4 13.4 14.1 13.6 11.8  8.1  5.2  4.7  8.9  16
    1994  4.7  3.4  6.6  7.4  9.8 13.1 16.2 14.8 12.2  9.7  9.7  6.1  9.5  16
    1995  4.5  5.9  5.0  8.1 10.6 13.4 16.7 17.5 12.9 12.2  7.7  2.6  9.8  16
    1996  4.9  3.1  4.5  7.9  8.4 13.1 15.0 15.4 13.1 11.0  5.5  3.4  8.8  15
    1997  3.2  6.0  7.7  8.4 10.5 12.9 15.6 17.3 13.1  9.9  8.5  6.0  9.9  15
    1998  5.2  7.4  7.2  7.0 11.6 12.6 14.1 14.4 13.5  9.4  6.0  5.5  9.5  14
    1999  4.9  4.7  6.6  8.5 11.2 12.4 15.5 14.7 14.3 10.3  7.5  4.2  9.6  12
    2000  5.0  5.6  6.8  6.9 10.8 12.7 14.1 15.0 13.6  9.8  6.6  5.3  9.4  11
    2001  3.7  3.8  4.5  6.9 11.3 12.4 14.9 14.9 12.5 12.5  7.4  4.1  9.1  11
    2002  5.8  5.9  6.7  8.4 10.9 13.3 14.2 15.4 13.4  9.1  8.1  5.5  9.7  11
    2003  4.6  4.4  7.1  9.1 10.7 14.4 16.1 16.2 13.2  8.7  8.0  5.1  9.8  11
    2004  5.0  5.0  6.2  8.7 11.0 13.8 14.3 16.0 13.6  9.8  7.7  5.9  9.8  11
    2005  6.0  4.4  6.7  8.0  9.8 13.6 14.9 14.6 13.6 11.7  6.4  5.3  9.6  10
    2006  5.0  4.5  4.2  7.4 10.3 13.9 17.1 14.9 15.0 12.0  7.8  6.2  9.9  10
    2007  6.2  5.6  6.7 10.1 10.5 13.3 14.2 14.2 12.6 10.8  7.5  5.3  9.8  10
    2008  5.7  5.6  5.5  7.2 12.0 12.7 15.1 15.1 12.7  9.1  6.8  4.2  9.3  10
    
    

    I know, down in the weeds of detail, but it might matter. So now we’ve suppressed the effect of putting 2 thermometers in the same place on the “thermometer count” by counting the actual locations. We still keep the “substation” differences, so the thermometer over the tarmac and the one on the roof at the same airport will count as 2 locations… But that isn’t very often. Mostly its the first 5 digits that change, then three zeros, then the mod flag changes…

    Try as I might, I don’t see any warming trend in this data for the UK.

    Ellie, can you “graph it up” and report if my “eyeball” of it comes through in a real graph? If it does, I promise I’ll learn to move graphs from PC land into postings ;-) (I’m really close to doing it, I just had a cat to save and some other “life issues” that needed tending 8-}

    Earlier you had said:

    Ellie in Belfast: Copy. Paste using Paste special… (below Paste menu) options include pictures – gif and jpg files. You can paste them back into MS Excel as pictures and lift them from there.

    Well, I got it to “paste as a gif” (it took pasting as a special graphic object of some MS type, then doing another copy / paste special) and I’ve got it pasted into a Word doc as a gif, but for the life of me I can’t see any way to save it as a gif nor to “lift it” as a gif…

    (It really ought not to be this hard to make a line graph out of a set of data points and load it into a web page… Tonight I’ve explored wigits, html, K-office on Linux RH 7.2, Appleworks, MS Excel and Word under Win-98, and how to get them to talk to each other. Heck, it would be faster and easier to point the Nikon at the screen and have a gif with one snap of the shutter… I really don’t want to hear that I need to go buy a few hundred dollars of software and a newer PC to make a line graph in a web page.)

    If this data shows no real warming pattern for the UK, that would be “very interesting” and “very useful”. This exercise has given me an idea:

    Take specific places, like the UK, Australia, etc.; and do individual historical graphs for their data. If, as I suspect, that data without all the jiggery-pokery of GIStemp or Hadley, shows flat or cyclical “trends”, they would be rather highly incriminating… The only way all the pieces can be flat, but the globe “warming”, is if the “merging the pieces” step fabricates it… as I think I’ve shown it does in the aggregate analysis.

    I could easily see a series of postings on UK, Aussy, Canada, Brazil, Argentina/Chili, India, etc. being dead flat, as being a bit of a “wedge issue” in some places ;-0

  40. E.M.Smith says:

    So the real peak of “number of locations” was in 1958-59 at 48 distinct locations.

    That gives us:

    [chiefio@tubularbells analysis]$ UK.1958 ../STEP0/input_files/v2.inv
    65103005000 LERWICK 60.13 -1.18 84 52R -9HIxxCO 1x-9WATER A 0
    65103026000 STORNOWAY 58.22 -6.32 13 2R -9FLxxCO 1A-9WARM GRASS/SHRUBB 7
    65103068001 GORDON CASTLE UK 57.60 -3.10 32 101R -9HIFOCO10x-9COASTAL EDGES A 0
    65103072001 BRAEMAR UK 57.00 -3.40 339 610R -9MVxxno-9x-9HEATHS, MOORS A 0
    65103091000 ABERDEEN/DYCE 57.20 -2.22 65 98U 210HIxxCO 5A 3WARM CROPS C 26
    65103100000 TIREE 56.50 -6.88 12 4R -9HIxxCO 1x-9WATER A 0
    65103140000 GLASGOW AIRPO 55.87 -4.43 8 68U 881HIxxCO15x-9WARM CROPS C 46
    65103160000 EDINBURGH AIR 55.95 -3.35 41 51U 470HIxxCO 4A 2WARM GRASS/SHRUBC 29
    65103160001 EDINBURGH/ROYAL OBS.UK 55.90 -3.20 134 155U 470HIxxCO 9x-9WARM GRASS/SHRUBC 24
    65103162000 ESKDALEMUIR 55.32 -3.20 242 293R -9HIxxno-9x-9WARM CROPS A 0
    65103209001 DUMFRIES UK 55.10 -3.10 -999 177S 29HIxxCO10x-9WARM CROPS A 0
    65103241001 COCKLE PARK UK 55.20 -1.60 99 46R -9FLxxCO 5x-9WARM CROPS C 20
    65103242001 DURHAM UK 54.80 -1.60 102 107U 89HIxxCO15x-9WARM CROPS C 25
    65103292001 SCARBOROUGH UK 54.20 -0.40 -999 72S 43HIxxCO 1x-9COASTAL EDGES B 0
    65103302000 VALLEY 53.25 -4.53 11 33R -9HIxxCO 1A-9WARM CROPS B 7
    65103316001 BIDSTON UK 53.40 -2.90 -999 30U 540FLxxCO 3x-9WARM CROPS C 41
    65103329001 STONYHURST UK 53.80 -2.50 115 114R -9HIxxCO30x-9WARM GRASS/SHRUBC 15
    65103334000 MANCHESTER AI 53.35 -2.27 78 50U 490FLxxno-9A 1HEATHS, MOORS C 22
    65103334001 WARRINGTON 53.38 -2.65 27 35U 65FLxxCO25x-9WARM CROPS C 28
    65103345001 SHEFFIELD UK 53.40 -1.50 -999 136U 558HIxxno-9x-9HEATHS, MOORS C 52
    65103355001 YORK UK 53.90 -1.10 -999 22U 102FLxxno-9x-9HEATHS, MOORS A 7
    65103377000 WADDINGTON 53.17 -0.52 70 56U 74FLxxno-9A 3HEATHS, MOORS C 12
    65103482001 MILDENHALL 52.37 0.48 10 40R -9FLxxno-9A-9WARM CROPS B 20
    65103482002 LAKENHEATH 52.40 0.57 10 42R -9FLxxno-9A-9WARM CROPS B 9
    65103482003 SCULTHORPE 52.85 0.77 68 58R -9FLxxCO20A-9WARM CROPS A 0
    65103496001 GORLESTON 52.60 1.70 2 3U 50FLxxCO 1x-9COASTAL EDGES C 21
    65103501001 ABERYSTWYTH UK 52.40 -4.10 -999 37S 15HIxxCO 1x-9COASTAL EDGES A 13
    65103521001 ROSS-ON-WYE UK 51.90 -2.60 -999 96R -9HIxxno-9x-9WARM CROPS C 8
    65103534000 BIRMINGHAM/AI 52.45 -1.73 99 112U 1059HIxxno-9A 1WARM CROPS C 25
    65103534001 EDGBASTON UK 52.50 -1.90 -999 125U 1059HIxxno-9x-9WARM CROPS C 88
    65103560001 HUNTINGTON 52.37 -0.22 49 43S 17FLxxno-9A 2WARM CROPS B 14
    65103590001 BENTWATERS 52.13 1.43 26 51R -9FLxxCO 6A-9WARM CROPS A 0
    65103649001 FAIRFORD 51.68 -1.78 91 110R -9HIxxno-9A-9WARM CROPS A 0
    65103649002 OXFORD 51.75 -1.58 91 111U 117HIxxno-9x-9WARM CROPS C 14
    65103649003 UPPER HEYFORD 51.93 -1.25 133 110R -9HIxxno-9A-9WARM CROPS A 0
    65103657001 OXFORD UK 51.70 -1.20 63 70U 117HIxxno-9x-9WARM CROPS B 8
    65103670001 ROTHAMSTEAD UK 51.70 -0.30 128 101U12332HIxxno-9x-9WARM CROPS C 22
    65103672001 KEW UK 51.50 -0.30 5 44U12332FLxxno-9x-9WARM CROPS C 53
    65103683001 WETHERSFIELD 51.97 0.50 101 76R -9HIxxno-9A-9WARM CROPS A 6
    65103683002 CAMBRIDGE UK 52.20 0.10 12 54U 106FLxxno-9x-9WARM CROPS C 25
    65103696001 FELIXSTOWE 52.00 1.30 3 27S 19FLxxCO 1x-9WATER C 18
    65103696002 WOODBRIDGE 52.08 1.40 29 45R -9FLxxCO 5A-9WARM CROPS A 6
    65103743000 LARKHILL 51.20 -1.80 133 109R -9HIxxno-9x-9WARM CROPS A 15
    65103761001 GREENHAM 51.38 -1.28 125 108S 25FLxxno-9A 2WARM CROPS A 9
    65103779001 GREENWICH/MARITIME MUK 51.50 0.00 7 37U12332FLxxCO30x-9WARM CROPS C 67
    65103827000 PLYMOUTH WC 50.35 -4.12 50 76U 259HIxxCO 1x-9WATER C 19
    65103865000 SOUTHAMPTON/ 50.90 -1.40 9 20U 214FLxxCO 3x-9WARM CROPS C 26
    65103917000 BELFAST/ALDER 54.65 -6.22 81 68U 552FLxxCO20A15WARM CROPS B 13

    An interesting example of the “substation” is the EDINBURGH AIR vs the EDINBURGH/ROYAL OBS.

    At any rate, you can compare this list with the other list of present 10 and decide if it looks like a “cherry pick” of warming stations that were kept or a representative sample. Alas, I would not what or where a “PLYMOUTH WC” was even if I was in desperate need of a “WC” ;-) and would certainly not know if it was warming…

  41. Ellie in Belfast says:

    @EM Smith 08:39 post: there is warming – at a rate of 0.3C per 100 years. Graph to follow.

  42. Ellie in Belfast says:

    http://sites.google.com/site/elliesgraphs/uk-temperature-graph

    Taken me several goes to work out how to put this up.

    To post graphs as GIF files or other picture formats – here’s how I found it could do it:

    Lucy Skywaker had sent me some GIf files. I opened one of them (with Microsoft Picture Manager), copied the graph and tried to paste it into the open file. No joy, or so I thought, until I went to close it and it told me there were unsaved changes.. so I was able to save the graph directly as a picture file. After that uploading it was easy.

  43. Tonyb says:

    EM Smith

    Great work.

    I think we all look at far too big a picture and like to create highly complex theories to explain the ‘catastrophic changes’ man is making, when the devil is in the detail. It is the station measuring system that is the problem, and that is because they no longer measure the same micro climate they started off recording perhaps decades before.

    I believe that micro climates are exactly what they say. They represent a very small area that can be totally different to another area close by.

    I know that our micro climate and that of another micro climate some 8 miles away are going in different directions.

    If you have thousands of thermometers in a single country you can perhaps track this. Multiply that by each country and you might have a good stab at the changing global climate if ALL the thermometers and methodology remain constant. But they don’t.

    Its where and how we record the information that is not constant and consequently data no longer represents the local climate they started off in, because the measurement point has moved and ends up recording another micro climate hundreds of miles away that will be completely different!

    Looking at the fine detail shows us that HUNDREDS of individual locations around the world have been cooling-not warming-for decades.

    The few you cite as being the max number of stations (your 8,53 pm) in Britain during 1958-9 are reasonably representative if you take that huge caveat above

    I would be horrified if those stations you quoted in your post at 7.02 are correct. It is far too few and doesn’t begin to represent the topgraphy of this country. I also suspect the identification code is not right. I have located each of those that I think should have a letter next to their name. I do promise you it is worth opening each link.

    aberdeen/dyce
    http://www.autoinsider.co.uk/mapping/map.php?pc=AB21+7AB

    There are several Leemings. Can you be more specific. This is the most famous one
    http://www.raf.mod.uk/rafleeming/

    I know of only one Valley
    http://www.raf.mod.uk/rafvalley/

    There are several waddingtons. This is the most famous
    http://www.waddingtonairshow.co.uk/

    This is Londonb/gatwic (i think the rest of this was missed off.) Here it is.
    http://www.gatwickairport.com/

    Now your super human powers of deduction will have realised by now that most of the UK sites are…Well you tell me.

    Recording one micro climate then switching to a regional one is not a good idea and completely invalidates the data. It would be interesting to do a study of another country with few stations with someone who has knowledge of that country, to see if the same applies.

    Tonyb

  44. E.M.Smith says:

    TonyB: The two numbers right after the text name are the LAT and LONG for each location. You can take, for example, LEEMING and look at the record:

    65103257000 LEEMING 54.30 -1.53 40 42R -9HIxxno-9A-9WARM CROPS B 9

    This can be broken down into meaningful bits as:

    651 country code – UK
    03257 Station Identification
    000 Sub-Station of all zeros means the main site
    LEEMING the text name
    54.30 Lattitude
    -1.53 Longitude
    42R The R means Rural. Others are U for Urban and S for suburban.

    And then some specifics about the nature of the location, the Meta-Data (that are all documented elsewhere in the descriptions in the links I provided). The -9A- is where we see, due to the A, that it is an airport. (Airports are often described as “CROPS”… don’t ask why…)

    So you can exactly identify each of these locations (and even get satellite views if you like by putting the LAT and LONG into Google Earth).

    Also, the reason i’ve provided the “ftp” links is exactly so you can validate all this for yourself. The v2.inv file is a simple text file. Just click on the ftp link and download v2.inv. You can then open it with “Word” or whatever. It is sorted by the first field (country code-station ID) so you can simply scroll through it and see what stations are in the record. Similarly, the v2.mean file is a simple text file. You can do a “find” for any station ID and see what records exist (it is sorted by countrycode-station and year, so a given stations records all are together in one part of the file by year).

    Basically, you can check all this stuff yourself with a download and an editor.

    And yes, we are being overrun with airports as ‘rural’…

  45. Tonyb says:

    E M Smith

    I have asked Ellie to check this out. Basically it appears there are far fewer ‘official’ stations than we thought. They are in areas thast do not begin to represent the UK’s many and varied climatic regions, and many seem to be airports.

    Apar from that they are no problem :)

    Tonyb

  46. Ellie in Belfast says:

    Tonyb

    It just that GIStemp rejects 12 of the stations due to intermittancy of data, and only 46 are used in the final output of the programme.

    If you think the UK is poorly served don’t dig any further ;-)

  47. E.M.Smith says:

    FWIW, realize that the results about number of stations are based entirely on NOAA GHCN data with nothing to do with the GIStemp program. I have no idea how much feedback there is from GISS to NOAA as to what ends up in GHCN, but this is not a result of the GIStemp code. Only the decisions made by the same people who lead the organizations involved. …

    AND to the extent that GHCN data are all that is used by anyone (as it looks to be) then ALL “data series” are corrupted in this same way. Be they GIStemp, HadCrut, or any other.

    Control what thermometers are “official” and control the “official global climate”. Yes, IMHO, that is exactly what best fits the patterns of thermometer change.

  48. E.M.Smith says:

    Tonyb: I have asked Ellie to check this out. Basically it appears there are far fewer ‘official’ stations than we thought. They are in areas thast do not begin to represent the UK’s many and varied climatic regions, and many seem to be airports.

    “Spot on!” covers it, I think. The “problem” is that, in the past, one ruler is used, and today the ruler has 1/2 as many “tick marks” on it. So what is the actual thing you are measuring and how much is the “rubber ruler”?

    At present, the UK ruler has 10 tick marks. A while ago it had 48. Taken over the entire temperature history, there are 59? or so; but what do they mean since they keep moving around? It is somewhat like trying to measure sobriety by how many drinks were consumed, but the individuals in the bar doing the consuming are constantly changing…

    Apart from that they are no problem :)

    Well, once the basics are all screwed up, the finesse at the top does become kind of pointless ;-)

  49. rob r says:

    This has evolved into a facinating project.

    It would be well worth a full guest posting on WUWT.

    It fits in nicely with the surfacestations project.

    I suspect you could enlist some assistance if you increased the profile of the project.

  50. Tonyb says:

    E M Smith

    I will ask you some questions which I have also put to Ellie so am hoping that collectively you can give me some answers.

    These are relevant to the study I am carrying out on pre 1850 data sets, whereby the UHI effect in the modern era is starting to come out. (Practically all the stations have become very urbanised over the last 250 years)

    1) How many official Giss stations (i.e. those that provide information that goes into their global temperature report) were there at its maximum extent and today?

    2) Can the number classified as ‘Urban’ (or should be treated as having become Urban over the years) be identified by any of the coding? Whilst ‘A’ for airport can reasonably be taken as such, is their one that denotes the size of the urban area now around the station?

    Logically there should be, otherwise how is UHI factored in to the end results? I get the impression that a disproportionate number of stations on the GISS records are now urbanised and that the tiny adjustments that IPCC-as an example- factor in is totally inappropriate.

    3) Various studies seem to suggest researchers have ‘refined’ existing data that was compiled since the relative database was set up. As an example, in 2002 Phil Jones of CRU did a report on seven pre 1850 temperature data sets and put forward ‘corrections.’ Do studies like this get incorporated into the official record? (I seem to remember a long debate on CA where US 1930’s figures had been revised just a few years ago)

    4) Why the concentration on Giss? Hadcrut is frequently used as an information source by the IPCC. Is there an equivalent to yourself and Ellie beavering away on deconstructing Hadley? ( I appreciate the difficulty in obtaining data from CRU)

    Thanks for any light you can shed on this deep black hole.

    Tonyb

  51. E.M.Smith says:

    Tonyb: I will ask you some questions which I have also put to Ellie so am hoping that collectively you can give me some answers.

    I will provide what I can.

    1) How many official Giss stations (i.e. those that provide information that goes into their global temperature report) were there at its maximum extent and today?

    This is a bit hard to say. There really isn’t such a thing as a GISS station. There is a USHCN station record from NOAA. There is a GHCN station record (that includes the USHCN stations but in C instead of F and with a different modification history) that comes from NOAA for the USA and the Rest Of World via a means unknown to me (somebody gathers the data and glues it together, but I’ve not investigated by whom it is done).

    And then GISS, via the GIStemp program, takes that data, adds in the Antarctic data (from three distinct source files), deletes anything older than 1880, replaces Hohenpeisenburg (hope I spelt that right ;-) with a “longer private copy”, merges the USHCN into the GHCN along with a slightly dodgy F to C conversion and a bit of splice-o-matic filling in of missing bits and smoothing the spice if both GHCN and USHCN records exist. The data then go off to STEP1 that does more splicing of records ( IIRC, it is merging, splicing, and averaging to replace the different modification flags on different records at one location with one single longer spliced record).

    At the end of this you have, I believe, one “station record” for each “station”, but it’s a composite beast. Some fact, some fiction, some splice, some made up via the “reference station method” from “nearby” 1000 km away…

    So to properly answer this, I would need to know if you think 2 different “modification flags” count as two different “official stations”. I have taken the simple approach of generally treating any difference as a different station. This can cause “issues” as we saw with the UK where there was a climb to 48? stations that dropped back down, when on closer inspection is looked like new equipment installed, parallel run for a few years, old equipment removed.

    Choices: By “GISS Stations”, do you mean USHCN, USA in GHCN, or the composite with whole globe in it that makes it to the STEP1 part of GIStemp (glued together bits with Antarctica) or what make it out of STEP2 (mod flag erased, some stations tossed, etc.)

    The change is from about 15,000 “station records ids” to about 6,000 max (for whole globe) depending on what choices you make. So tell me which ones you want and I’ll count them up that way.

    2) Can the number classified as ‘Urban’ (or should be treated as having become Urban over the years) be identified by any of the coding? Whilst ‘A’ for airport can reasonably be taken as such, is their one that denotes the size of the urban area now around the station?

    There is a fundamental flaw in how ALL the data is presented in GHCN (and USHCN). There is a single flag for “Urban”, vs “Rural”, vs “Suburban” for this exact moment in time. So a station may be classed as “Urban” because it is a city today and the record from 1890 when it was a cow field will still be classed as “Urban”! This “meta data” really needs to be unique by year, but only the present value is available.

    Given that limitation, the v2.inv file does mark each record with that flag. If you look here:

    https://chiefio.wordpress.com/2009/02/24/ghcn-global-historical-climate-network/

    about 1/2 way down is the “magic decoder ring” for the v2.inv type records. It has entries like:


    c ic=3 digit country code; the first digit represents WMO region/continent
    c iwmo=5 digit WMO station number
    c imod=3 digit modifier; 000 means the station is probably the WMO
    c station; 001, etc. mean the station is near that WMO station
    c name=30 character station name
    c rlat=latitude in degrees.hundredths of degrees, negative = South of Eq.
    c rlong=longitude in degrees.hundredths of degrees, – = West
    c ielevs=station elevation in meters, missing is -999
    c ielevg=station elevation interpolated from TerrainBase gridded data set
    c pop=1 character population assessment: R = rural (not associated
    c with a town of >10,000 population), S = associated with a small
    c town (10,000-50,000), U = associated with an urban area (>50,000)
    c ipop=population of the small town or urban area (needs to be multiplied
    c by 1,000). If rural, no analysis: -9.

    And more. Notice the next to last line “ipop” is the population if you need finer grain than the U,R,S flag.

    So, extending the example from above:

    65103257000 LEEMING 54.30 -1.53 40 42R -9HIxxno-9A-9WARM CROPS B 9

    This can be broken down into meaningful bits as:

    651 country code – UK
    03257 Station Identification
    000 Sub-Station of all zeros means the main site
    LEEMING the text name
    54.30 Lattitude
    -1.53 Longitude
    42R The R means Rural. Others are U for Urban and S for suburban.

    We would add:

    40 meters elevation, 42 meters elevation from Terrainbase grid.
    R for Rural, -9 no population count done. HI says it is “hilly”
    The xx says we don’t know the general vegetation type from navigation charts (MA marsh, FO forest, IC ice, DE desert, CL clear or open)

    The “no” says it is not on an Island smaller than 100 km**2 or narrower than 10 km where the station is located, NOR is it within 30 km of the coast, NOR is it next to a lake larger than 25 km**2 and the -9 says it is not coastal.
    “A” so it is an Airport, the next -9 says we don’t have km to nearest town since it is a rural Airport (or so it claims…) while the 1/2 by 1/2 degree “grid” it is in is generally “Warm crops” and I don’t know what the B9 stands for.

    As an example with a population count we have:

    42572394002 SAN LUIS OBISPO POLY 35.30 -120.67 96 227S 42HIxxCO12x-9MED. GRAZING C3 32

    which is Suburban and has a 42,000 population estimate.

    Logically there should be, otherwise how is UHI factored in to the end results? I get the impression that a disproportionate number of stations on the GISS records are now urbanised and that the tiny adjustments that IPCC-as an example- factor in is totally inappropriate.

    That is an entire discussion in itself. There are already UHI adjustments in some of the data (you can chose what adjustment history you like at the time you ftp the data from NOAA. For some inexplicable reason, GIStemp takes the unadjusted GHCN data and the adjusted USHCN data and then does a bizarre ‘unadjustment’ step, and then ‘readjusts’ it in STEP2 via a (IMHO) broken method using “The Reference Station Method” (see the ‘Pisa’ posting for details).

    So, NOAA has adjusted, and not. GIStemp has ‘quasi-adjusted’ that then gets mal-adjusted. You can download adjusted or not. So many choices for how to cook your past…

    3) Various studies seem to suggest researchers have ‘refined’ existing data that was compiled since the relative database was set up. As an example, in 2002 Phil Jones of CRU did a report on seven pre 1850 temperature data sets and put forward ‘corrections.’ Do studies like this get incorporated into the official record?

    There are changes incorporated. I don’t know exactly which ones. This link:

    http://www.ncdc.noaa.gov/oa/climate/ghcn-monthly/index.php

    Claims to have papers that describe the method. I’ve not read them.

    The data and some descriptions are here:

    ftp://ftp.ncdc.noaa.gov/pub/data/ghcn/v2

    which for me, on a Mac, I just click on the link and a window opens on my Mac with the files clickable documents like any other on my machine. Don’t know what happens on a PC.

    There are several files with “read” or “readme” in them that are likely to contain more detail. There is a similar set for USHCN if you want to inspect them:

    ftp://ftp.ncdc.noaa.gov/pub/data/ushcn

    4) Why the concentration on Giss? Hadcrut is frequently used as an information source by the IPCC. Is there an equivalent to yourself and Ellie beavering away on deconstructing Hadley? ( I appreciate the difficulty in obtaining data from CRU)

    Well, I had to start somewhere and my initial bias was to think the Brits had likely done things right (they used to; and Mum, being a Brit from the old times had raised me to expect to do it the right way as that was what a Proper British Gentleman Ought To Do.) Later I found out that while GIStemp is published, the Hadley folks will not release their code and methods. Lately we’ve found out that they lost their original raw data too. So we don’t really have anything to work with from Hadley CRU.

    If there were CRU data and code available, I’d do the same inspection of it. Any copy of code or pointers to copies cheerily accepted!

    But as near as I can tell, Hadley CRU is a sealed locked black box with (now) no input data and only an undefined (no methods) output. Don’t know how you could even begin to validate that…

    But I presume that GHCN data from the UK are “reasonably clean”… one hopes… maybe…


    Thanks for any light you can shed on this deep black hole.

    Tonyb

    Well, don’t know if I made it better or worse, or if you will still be thanking me now that I’ve lit a tiny candle and you can see some of the rats and spiders scurry about ;-)

  52. Ellie in Belfast says:

    @EM Smith. Good explanation. The only thing I can add is that the Rural/Urban/Suburban thing is decided by brightness from a nightime satellite photo. see here:

    http://data.giss.nasa.gov/gistemp/station_data/station_list.txt

    Hmmm. Many smaller airports are only operational up to ~9-10pm local time. After that I would assume lights off. So lots of airports in the rural set.

  53. Tonyb says:

    E M SMith

    Yes. Well. Simple. Unfortunately I’ve just remembered I’ve got something urgent to do for the next month…

    Seriously, this is great work. I just need to read it nineeteen or twenty times.

    The future prosperity and direction of the Westen world lies in this nonsense. Scary isnt it?

    Tonyb

  54. E.M.Smith says:

    Yes, scary. This POS (Piece Of …) is the foundation on which will be built the economic future of the planet. I think we’re gonna need a new economy…

    What I need from you, to give you a count of thermometers over time, is a fairly simple set of decisions:

    Do you want US data counted, or the “whole world” counted?

    (“GISS Stations” doesn’t mean anything to me. I don’t know if that means US Historical Climate Network, or the stuff used inside GIStemp, or what… GISS is the Goddard Institute of Space Stuff ;-) and does lots of stuff, GIStemp being one of them, and NOAA are the folks with the “semi-raw” data, but I don’t know if NOAA are part of GISS. And if they are, we still have to decide if you want NOAA stuff or GIStemp stuff… So if you want US only, I count one thing, if ROW, I count another.)

    Do you want each LOCATION counted, or each MODIFICATION of the record from each location?

    (That is, do you want to count 2 thermometers at London Gatwick during the time they had 2 sets of equipment in the cutover from old gear to new gear, or just 1 because it is the same place?)

    Do you want it as it goes in to GIStemp, or as it comes out when GIStemp uses it to make the anomaly maps?

    (That is, do you want unmolested count,or the count after GIStemp has “done it’s thing” and you have left, determining the fate of the planet, what they chose to keep…)

    Answer those, and I’ll give you a count of change over time.

  55. Tonyb says:

    E M Smith

    The whole world

    1 location-not multiple instruments

    As it comes out (i.e the stuff that will be used to compile the official record that will bring down the western world etc…

    Thanks a lot. Its 12.45 am . I’m off to bed.

    Tonyb

  56. E.M.Smith says:

    @Tonyb

    OK. So in STEP2 in GIStemp. By then it is “whole world” with multiple instruments spliced into one record. STEP3 does “Grids, Boxes, and zonal anomalies” so it is no longer dealing with thermometers at its output, only at the input to STEP3 from STEP2.

    This will be a nominally “Land only” dataset (which is all GHCN claims to be) but it does include a couple of locations at sea reported by various ships as they reach those LAT LONG locations. GIStemp also uses these “land” data to fill in “Grids and Boxes” up to 1200 km out to sea from any land station. So look at all those islands in the Pacific and put a 1200 km radius around them… It’s pretty much the whole globe, minus a few holes with no land within 1200 km in any direction.

    STEP4_5 blends in a Hadley CRU “Sea Surface Anomaly Map” that is based on it’s own zoo of some “in situ” measurements by ships and buoys; a dash of satellites; and a large quantity of “Optimal interpolations” seasoned with a bit of “simulations”… so there is no way to have any idea how many “thermometers” went into making it. Hadley having lost it’s data (“The Dogs Lunch”)… GIStemp does say in the code that running STEP4_5 is optional; so I think it is a reasonable “break point” to claim that GISS product ends at STEP3 grids and boxes and that Hadley enters into the following step, so it is outside the scope of a “GISS Product” description.

    FWIW, the total number of “world stations going in to STEP2” is 7364 per the V2.inv file size. All I need to do is figure out how to instrument that code so that it gives me the count by year. In STEP 2 there are two interesting log files:

    [chiefio@tubularbells STEP2]$ wc -l Ts.GHCN.CL.station.list
    6289 Ts.GHCN.CL.station.list
    [chiefio@tubularbells STEP2]$ wc -l short.station.list
    1341 short.station.list
    [chiefio@tubularbells STEP2]$

    These add up to 7630 stations, so we have a bit of a mystery where the added 266 stations come from since as I understand it, the first log is the stations that are used and the second log is the stations that are dropped as too short.

    I ended up spending today on “zones” (Ellie asked a Very Interesting Question …) and the result is the https://chiefio.wordpress.com/2009/10/22/gistemp-send-in-the-zones/ posting. But as a consequence it is nearing 11pm and I’m somewhat fried… So if it’s OK, this may take until Saturday PST for me to get it sorted out.

  57. Tonyb says:

    E M Smith

    No problem on timing. Thanks for your efforts.

    Your homework will be ‘Explain Gistemp methodology in no more than 100 words’ :)

    Tonyb

  58. E.M.Smith says:

    GIStemp glues together Antarctic, GHCN and USHCN data, and smoothes the join. This combined data is preened and missing data is, where possible, filled in with estimates from stations up to 1000 km away. The merged data has a UHI adjustment done by applying “corrections” from stations up to 1000 km away in different micro-climates. The result is divided into zones by latitude, then split into a set of “Grid Boxes” ) covering the whole globe. In some cases a single island can be used to “fill in” “boxes” of ocean up to 1200 km away. These are then used to make “anomaly maps” of changed temperatures over time.

    In a final step, the Hadley SST anomaly map can be merged if desired.

    Ok, so it’s closer to 120 than 100… but it’s close ;-) If you don’t count acronyms, “km”, and numbers it’s even closer 8-} And the Hadley merger is an optional step…

  59. E.M.Smith says:

    @Tony:

    The added 266 stations now make sense. If you look at “the Curious Case of Calcutta”, it shows that a “location” can be both “in” and “not in” at the same time. So while there are 7364 “StationIDs” some of the “mod flags” can be in the “joined accepted record” station list and some can end up in the “short.station.list” and have the same Station ID.

    OK, this tells me how to get the count you want. I need to take the set that ends up in the Ts.GHCN.CL.station.list and map them by years. Now it’s just a small matter of writing the computer program to do it ;-)

  60. Tonyb says:

    E M Smith

    Is there any way of factoring in the relative numbers from the NH and SH at various stages of the instrumental record, as used by gistemp?

    Say pre 1900 pre 1950 and Now.

    The reason I ask is that on another thread (the Science museum poll you were on) someone mentioned about their only being 20 stations in the SH prior to 1958.

    James Hansen mentions a very small figure for 1880 in the SH, which is one reason he chose to go with 1880 as a start point and not 1850.

    20 would seem an incredibly small number on which to base a ‘global’ teperature estimate especially as many of them have probably upped sticks and moved somewhere else by now. Surely it can’t be right?

    tonyb

  61. E.M.Smith says:

    Tony, see:

    https://chiefio.wordpress.com/2009/08/17/thermometer-years-by-latitude-warm-globe/

    That is just GHCN (and does not include the Antarctic records) but you can see the number by latitude band by year.

    It’s not pretty…

    And those are “station records” so the actual station count will be smaller by the degree to which there are multiple modification flags per station.

  62. Tonyb says:

    EM Smith

    Have read it. So everything to the right of the equator column ( NW on) is therefore NH. Everything ( excluding equator) is SH.

    Excellent!

    May I have your permision to reproduce it?

    Also do you already have a SHORT and SIMPLE explanation as to how Gistemp take a piece of raw temperature data from a specific micro climate, ‘adjust’ it, use it for their global record, then drop stations so the same micro climate is not being measured any longer.

    Adjustments for ‘urban’ areas, what is termed an urban area, the method of reaching across ‘cells to acquire information-could all usefully be mentioned.

    A sort of step by step guide entitled ‘What you see is not what you get’

    By simple I mean something even a policymaker could understand. This is something I (and many others) could make great use of. I have looked through your site but can not see such a basic guide.

    Thanks

    Tonyb

  63. E.M.Smith says:

    TonyB, you have my permission to reproduce from here anything that is of use to you in your efforts to reduce the AGW hysteria.

    I’ve been planning a “Layman’s Guide To GIStemp” but have not done it yet… How soon do you need it and how many pages?

    (I take it the 100 words were not enough ;-)

    BTW, I chose an odd number of bands deliberately so that the EQ EQUATOR band would actually be the whole equator. Both a N and S component. It really is a unique place that ought not be cut in half and apportioned some to N and some to S in climate programs / models. I can fairly easily (about 1/2 day) make a new version with different bands in it if you would rather have a divide AT the equator. That chart is by 20 degree bands (9 of them) so the EQ bucket gets the equator +/- 10 degrees. A fairly specific place.

    It would be a reasonable approximation to allot 1/2 of the equatorial band thermometers to each of NH and SH, if you need something fast.

    (The program that made the data is hard coded with 9 bands, but it is just a few lines to change to make that more, or less, bands. It will take me longer to find it and remember what all I wrote than it will take to actually change the band count to 8…)

  64. Tonyb says:

    E M Smith

    I think the table is fine for my purposes. Thanks.

    Re; The Layman’s guide.

    I am currently writing an article covering the pre 1850 thermometers. It shows man has got his fingers all over the warming, but it is a UHI one not carbon.

    We can’t do an analysis of Hadley as the dog ate it, so Gistemp will have to take the honours. This is a core element to get over to people as they need to see how messed up the temperature data is. It fails to identify UHI adequately, assigns data from a cell, Airports are rural and the global temperature concept covers up trends that demonstrate many places are cooling not warming. (A piece in the 500 words about the pointlessness of a ‘global’ figure would be useful)

    It needs to be Step by step without jargon or anything the layman could not get to grips with. Say ten bullet points. 500 words maximum. Can you do this? This would be useful so people like me can incorporate it into articles and could be forwarded to the media. Hadley are 15 miles from me and we often use the local paper to snipe at them!

    I suspect that as well as a succinct non technical summary there is also a need (in due course) for a fuller version with jargon, that could include some of your studies. This second version would make a good guest post on WUWT.

    Tonyb

  65. Tonyb says:

    E M Smith

    Going back to your March of the thermometers and the numbers of each in speciific zones.

    I find that;

    Land masses on earth total 57.9 million square miles, only 29.8% of the total surface area. They are:

    1 Asia – (44,579,000 sq km)
    #2 Africa – (30,065,000 sq km)
    #3 North America – (24,256,000 sq km)
    #4 South America – (17,819,000 sq km)
    #5 Antarctica – (13,209,000 sq km)
    #6 Europe – (9,938,000 sq km)
    #7 Australia/Oceania – (7,687,000 sq km)

    In sq miles
    – Eurasia, 21.2
    – Africa, 11.7
    – North America, 9.4
    – South America, 6.9
    – Antarctica, 5.4
    – Australia/Oceania, 3.3

    It is quite difficult to relate the various zones in your table to actual continents, as each continent could contain several categories of zones.

    Can you tell me how many stations fall into each continent? I am trying to establish how many sq miles of land surface there is for each thermometer in a given continent.

    Total in 2009 2333 stations 1 per 248178 sq miles
    Max in 1979 decade 9191 1 per 62996 sq miles

    Some country sizes in Sq miles

    UK 94,600 sq miles
    Italy 114610 sq miles
    Portrugal 34028
    France 204177
    Belgium 11,313

    Of course there is a preponderance of thermometers in some places, so they will have a disproportionate number to other less well served areas, but simplistically these days there is one thermometer for each area of land the size of France.

    In 1979 there were more like four. France has various distinct climatic zones and micro climates within these.

    Makes you think doesn’t it. How can you grab local temperatures with this sort of numbers? Of course it was even worse pre 1900

    Can you easily translate your numbers data into continents as per my first table which separates out Asia and Europe?

    Thanks

    Tonyb.

  66. E.M.Smith says:

    @TonyB

    Well, “your report” has been running for a while… just “did the math” and it will take 33 hours to run. I had used a “brute force” approach for the selection of records that is “less than ideal”. It works OK on batches of 40 or even 100 records, like smaller countries, but selecting the 6800 or so records, well, “brute force” meets ‘compute wall’… I had originally designed this bit of code for doing small batches of a single country where a broad search of the file could be done.

    I’m going to leave this one running while I cook up a more efficient method. (i.e. sort the records, then to a “line by line” match and print of the selected records. What I’m doing now is 6800 searches of the file. So read the file one time instead of 6800 times. This is the kind of stuff programmers deal with all the time). But i thought I ought to let you know, I AM working on it!

    Also you said:

    simplistically these days there is one thermometer for each area of land the size of France.

    In 1979 there were more like four. France has various distinct climatic zones and micro climates within these.

    Makes you think doesn’t it. How can you grab local temperatures with this sort of numbers? Of course it was even worse pre 1900

    Take a look at the “California” posting. 4 thermometers for California and they are all on the beach…

    Can you easily translate your numbers data into continents as per my first table which separates out Asia and Europe?

    Strange you should mention that… I’ve been taking my zones code and turning it into “by country” code. So as of right now I can give you a latitude migration chart for any country code on the planet.

    Next modification is to be able to feed it a set of country codes as a group. (Europe kind of needs this, all those dinky countries ;-) and it ought not to be too hard. I just finished the latitude programs for the poles region down to 45N. I needed one for the N.polar region to handle places like Canada and Russia… Now it’s just a matter of changing the “on country code” to “batch”…

    As soon as that bit is done, I can make any area you want as long as it is bounded / defined as a set of country codes.

  67. Tonyb says:

    E M Smith

    I think knowing the number of thermometers in each individual country through the decades would be an interesting start.

    If you can readily do this then yes please.

    Do I also understand correctly that you will do a 500 word ‘dummies guide to GISS?’

    If not 500 words, then as few as is humanly possible whilst still getting to grips with this monster and its many issues.

    Tonyb

  68. Harold Vance says:

    E.M., what kind of report are you running that would take 33 hours? I read the above threads but couldn’t piece it together. Is this some sort of analysis of v2.mean?

  69. E.M.Smith says:

    Tonyb I think knowing the number of thermometers in each individual country through the decades would be an interesting start. If you can readily do this then yes please.

    OK, i’ve got the extract of the data for “past STEP2, only what really makes it into GIStemp output” (that 33 hour run due to a slow technique…). And I’ve got programs that now let me make various collections of countries. AND I’ve figured out how to make “continent” reports (nearly trivial once you figure out that the first digit of “country code” is in fact “continent code”. Makes me wonder if the 2nd digit is ‘region in continent’…)

    So I’m about to crank out 7 reports of “Actual GIStemp accepted thermometer records change over time with thermometer counts”. Would you like the tables put here as a series of comments, or should I put them up as a new article, or in email or what? If I don’t hear anything in time, I’ll just post them as comments here. I can always do the added work of making it into a posting later.

    Do I also understand correctly that you will do a 500 word ‘dummies guide to GISS?’

    Yes.

    If not 500 words, then as few as is humanly possible whilst still getting to grips with this monster and its many issues.

    So, was my “100 words” version above the right “tone” just needing a bit more elaboration? I can make a description fit in any length. What changes is the degree of precision and detail:

    “GIStemp is a temperature reporting program that merges thermometer records from around the world to create a single temperature history, even if some parts have poor thermometer coverage, when it makes up data as needed.”

    vs. the total body of postings here under the GIStemp tab…

    Oh, and a bit of “schedule” information would be nice. Not needing exact dates, just are we talking days, months?

  70. E.M.Smith says:

    Harold Vance
    E.M., what kind of report are you running that would take 33 hours? I read the above threads but couldn’t piece it together. Is this some sort of analysis of v2.mean?

    I had a bit of code that did a ‘brute force’ extract and match of station data for a given station. Designed for 1 to a dozen scale of searches (i.e. it would do a “grep” of the v2.mean file once for each station). That does not scale well…

    TonyB wanted a version of the “thermometer count change over time” reports that used only the v2.mean type data after GIStemp STEP2 (with Antarctica added, but with some more deletions as seen in The Curious Case of Calcutta).

    While I started work on that report, I just “turned loose what I had” to let it do the selecting of records. (For the list of “post step 2 stations”, grep v2.mean and extract their records). That involves over 6000 readings, end to end, of the v2.mean file. Terribly inefficient. But it would work (and in fact has worked and run to completion). So while I was contemplating how best to make a FORTRAN program to do this data matching (and making new postings, and feeding the cat, and…) that crude approach could just crank away in a “race condition” with the elegant.

    And “brute force but slow” won the race with “elegant, fast, efficient, but not written yet, I’m working on it.”. (I estimate a “sort / match / extract one read only” approach would have taken less than a minute.)

    On a modern desktop machine with decent memory and a newer faster disk, the brute force would likely have been closer to 6 hours, FWIW. (On a really fast machine, brute force would have been minutes… Sigh, sometimes I miss “My Cray”. It really was fun having one. “Cray means never needing to wait for a result.” (At least on the scale of things I used it for. At the time, UNIX passwords had publicly readable “salt” for the encryption. We actually figured out that we could pre-compute all possible passwords and encryptions and just do a “salt lookup” into a large TB scale data set. Instant password decryption. That is why the “crypt text” of passwords is no longer visible in the /etc/passwd file… )

    Hope that lets you scratch that particular itch…

  71. Tonyb says:

    E M Smith

    Emailing the stuff to me would be fine (provided the tables don’t wander all over the place-perhaps a ‘Word’ attachment might be better?

    If it seems worthy of further discussion (and I’ve no idea yet what it will show) I might suggest it is worthy of a post

    500 words should enable you to fit in some detail without getting into the micro detail that will turn away the policymaker. Think ‘Credible, authoritative, measured.’ With a little bit of restrained ‘you won’t believe what they do’ thrown in for good measure, taking the cue from your comment “even if some parts have poor thermometer coverage, when it makes up data as needed.”

    Tonyb

  72. Tonyb says:

    EM Smith

    I thought you would enjoy this link that I discovered which interested me not only because of the numerous bits of evidence concerning climate change affecting ancient civilisations, but also for this gem;

    http://www.truthout.org/article/the-climate-man-the-curse-akkad

    The author comments that
    “… the discovery that large and sophisticated cultures have already been undone by climate change, presents what can only be called an uncomfortable precedent.” (for our own civilisation)

    And goes on concerning a visit to none other than James Hansen ;

    “ GISS’s director, James Hansen, occupies a spacious, almost comically cluttered office on the institute’s seventh floor. (I must have expressed some uneasiness the first time I visited him, because the following day I received an e-mail assuring me that the office was “a lot better organized than it used to be.”)

    Now perhaps the office tells us a great deal about the cluttered nature of the Giss database. The article goes into details I hadn’t seen before so thought you would enjoy having a read of it.

    Tonyb

    REPLY: “Thanks, I’ll take a look. Sounds interesting. -ems”

  73. Tonyb says:

    E M Smith

    I am recapping on the mass of information contained in this thread to ensure I am expecting to receive what you are expecting to supply!

    1) You are writing a Dummies guide to Giss of 500 words or so.

    2) You will be advising me of the number of thermometers per country (and if that falls easily the number by continent) and this will be by decade-like your March of the thermometers

    3) There is one additional study which may have got lost in the complexity of this subject.

    Put simply; How many stations globally are classified as urban and how many are classified as rural AND what UHI factor is used in the urban ones? Is it the same no matter the size of the urban area, and when does it start from? (The met office aply uhi from 1974)

    Following on from this, are we able to tell if stations classified as urban are genuinely so-for example a very small rural airport may be classed as urban but by any reasonable definition exhibit more rural charateristics. The opposite of course (and much more likely) is that rural stations may still be classified as such even though they have grown and now exhibit many of the characteristics of an urban area but have not had the UHI factor applied.

    My thinking on this is that whilst the urbanised area of the globe is very tiny, the number of thermometers recording urban temperatures is disproptionately very large. We are in effect mostly recording an urban temperarure from .1% of the globe rather than the global temperature of the other 99.9%

    Hope this all makes sense. These are all needed for the study I am just finishing on Historic temperature datasets.

    Thanks for your help

    PS I still laugh at the notion of Hansens comically cluttered office (which is better than it used to be) and whether this lack of organisation impacted on his highly complex Giss studies!

    Tonyb

  74. Ellie in Belfast says:

    Tonyb,
    many airports are classified as rural; many are not as they are in a major metropolitan area.

    Example – Milan/Linate is Urban, Milan/Malpensa is Rural.

    I would say Urban stations are urban – not much change there, but many many of the Rural stations I have looked at are certainly not pristine rural. When and by how much human structures and activities start to impact upon local temperatures will be a feature of the local microclimate.

    As for how much UHI adjustment is applied. Firstly UHI adjustments in the US and the rest of the world are different. I have seen that mentioned in several sites and am only starting to read up about it.

    As for how much adjustment and when it starts, that is a good question. I am sure there are rules in GIStemp, but looking at the output data it is very hard to see how they are applied – they certainly do not seem to be applied consistently.

  75. Pingback: A short anthology of changing climate | Watts Up With That?

  76. Pingback: A short anthology of changing climate | TaJnB | TheAverageJoeNewsBlogg

Comments are closed.