Some China Comparisons

Here are a few plots of Jones et al 1990 China urban vs rural sites. Maybe one of our more computer-oriented people could make a little applet to yield 42 comparisons from the small data set which is now collated. For now, here are some very quick and non-prettied up comparisons. I started with a couple of sites in western China because of the recent discussion of Dulan junipers. This is also useful because the leverage on these remote sites is very great because there are few observations and they affect large areas through the Jones gridcell methodology. (One thing to keep in mind with the Jones et al 1990 comparison is that it is only from 1954-1983, when there was much turmoil in China and that there is little trend in many stations. Chinese station temperatures took off more recently.)

First here is Xining (on the road from Lanzhou to Dulan) versus unidentified “rural” WMO station 53161. In this case, there is a pronounced trend in the urban centre while the rural centre is unchanged. Does this mean that there is a UHI trend in Xining? Note that the rural station is warmer than the urban station here.

chinas12.gif

Next here is a comparison of the two most remote western sites in the J90 comparison. There is somewhat of a trend in the urban site; whether the “rural” site has a trend or merely multidecadal volatility cannot be determined on the graph, but there is a difference between the two.

chinas13.gif

Steve Sadlov mentioned Guangzhou. Here’s a plot of the Guangzhou comparison to Heyuan. In this case, the “rural” location increased while Guangzhou remained relatively unchanged. Is Heyuan an example of exurban development?

chinas14.gif

Some of the comparisons look like there must have been some site change that is unaccounted for. For example, consider this graph where two series cross over. In this case, the “rural” station appears to warm while the city cools. Does this make sense?

chinas16.gif

Here’s another comparison where the urban center shows cooling, while the rural comparandum remains unchanged. It looks as though there might be some non-homogeneity in the urban series in the early part, where it moved to a cooler location without being adjustment. In this particular case, this results in a reduction of the urban-rural UHI differential.

chinas17.gif

Having looked at all these series, my overall impression is one that this is very thin gruel to be using in big reports. This is supposed to be the best and the brightest?

45 Comments

  1. Posted Apr 11, 2007 at 2:58 PM | Permalink

    Regarding the Chengdu cooling, I commented about this on another thread. The station, #56294, was located near the center of the city, but was then relocated, in 1971, to a western suburb, i.e. upwind from the center. Hence the cooling that you observed not only makes sense, it was predictable (even predicted—by me).

    I would like to know what Jones et al. have to say about all this (including the other threads).

  2. fFreddy
    Posted Apr 12, 2007 at 6:02 AM | Permalink

    Maybe one of our more computer-oriented people could make a little applet to yield 42 comparisons from the small data set which is now collated.

    Sorry, where is the collation ?

  3. KevinUK
    Posted Apr 12, 2007 at 7:50 AM | Permalink

    Steve,

    On what basis does Jones consider a site to be urban or rural? Why was China used rather than say the US or USSR? Based on what you’ve presented so far i.e. relative warming in some cooling in others I can were Jones probably gets his 0.05C UHI correction from. Perhaps we should be consider a UCI effect as well as a UHI effect? My general impression thus far is that the temperature measures here look just as problematic as the dendro data.

    KevinUK

  4. Steve McIntyre
    Posted Apr 12, 2007 at 9:41 AM | Permalink

    #2 fFreddy, the collated Jones 1990 station information together with names is at
    http://data.climateaudit.org/data/jones90/jones90.location.dat . This gives Jones #, wmo, lat, long (minutes converted to fractions of a degree) and name identification.

    The temparature time series from 1954 to 1983 are in
    http://data.climateaudit.org/data/jones90/jones90.data.dat with the header in each column being the wmo number, the columns are arranged in the same order as the location data (unlike the original).

    Rural stations are the first 42, urban stations are the second 42; the orders match.

  5. Douglas Hoyt
    Posted Apr 12, 2007 at 10:30 AM | Permalink

    Qitai has a population over 200,000 (see http://www.xist.org/cntry/subs/cn-65.aspx), so it is rather urban. Wulumuqi has a population of nearly 1.8 million. Both cities seem to be growing fast if one searches on them using google.

  6. Douglas Hoyt
    Posted Apr 12, 2007 at 10:36 AM | Permalink

    Also Chengdu has a decreasing population (http://www.xist.org/cntry/subs/cn-51.aspx) so perhaps that is why it is cooling.

  7. Douglas Hoyt
    Posted Apr 12, 2007 at 10:42 AM | Permalink

    Finally “rural” Heyuan has a population of 544,846 (http://www.xist.org/cntry/china.aspx). I guess a tabulation of populations of the rural and urban sites is needed.

  8. Steve McIntyre
    Posted Apr 12, 2007 at 11:34 AM | Permalink

    If someone sends me a list of populations for the 42 “rural” sites (And the urban sites for that matter, I’ll incorporate the information into the ASCII reference file. If people send them in piecemeal, insert the Jones number as well to save me looking it up. Maybe some people can volunteer to do 10 each.

  9. Steve Sadlov
    Posted Apr 12, 2007 at 12:44 PM | Permalink

    Some personal observation comparing Western countries with NE Asian ones with regards to development patterns. In the West, including even in crowded Europe, there is a reasonably good amount of arable land. In the countryside, as a result, there is a highly dispersed development pattern. Towns are small and even the regional service hubs tend to be small cities in the 5 – 20 thousand population range. This trend is at its most pronounced in the Americas, where rural areas can be truly devoid of substantial population. The trend toward larger properties and mechanization has only exacerbated it.

    In NE Asia, in contrast, arable land is at a premium. In places like Japan and South Korea, it’s due to topography. In China, it’s due to the fact that most of the arable land is concentrated in the eastern quarter of the country, as is the lion’s share of the Billion plus population. Imagine 700,000 people living east of the Mississippi. As a result, you don’t see dispersed settlement patterns. Even “rural” areas have densepack. “Rural small cities” would qualify as major metro areas by Western standards.

  10. Steve Sadlov
    Posted Apr 12, 2007 at 12:44 PM | Permalink

    Oops, missed a few zeros there. I meant “700,000,000 people.”

  11. Posted Apr 12, 2007 at 2:34 PM | Permalink

    The meteorological data spans 1954–1983. So if you are interested in population estimates, they should be from that time, not 2000+. I am sceptical that reliable population data for that time exists—for the same reason that I was/am sceptical that reliable meteorological data exists. A quick search seems to confirm that, e.g.

    H. Yuan Tien (1980),
    “Age-Sex Statistics for China: What Do Recent National Disclosures and Local Figures Reveal?”,
    Population and Development Review, 6: 651–662.
    doi:10.2307/1972931

    There are, though, some UN estimates. E.g. Chengdu had estimated populations of 958000 (1955), 1835000 (1970), 2639000 (1985). Note that the estimate for 2000 (3294000) is substantially different than the 2000 census data cited in Douglas Hoyt’s comments (5268000), presumably because of a different demarcation of Chengdu.

  12. Bob Koss
    Posted Apr 12, 2007 at 7:04 PM | Permalink

    Here are the elevations in meters for 80 of the 84 sites you listed.

    50854 150
    50949 136
    54135 180
    54497 14
    50963 110
    54324 176
    50756 240
    54616 11
    54725 12
    54476 N/A
    54936 109
    54852 72
    54916 53
    57193 53
    57378 66
    57584 52
    57662 35
    58144 19
    58150 7
    57799 78
    58314 68
    58251 5
    58754 38
    59293 41
    59417 129
    58477 37
    58646 60
    59501 5
    57537 N/A
    57606 977
    56287 629
    56751 1992
    57253 100
    57504 357
    56193 894
    57127 509
    53863 745
    53915 1348
    51379 794
    53593 910
    52983 1875
    53336 1290
    50745 148
    50953 143
    53161 N/A
    54342 43
    50978 234
    54237 182
    50774 232
    54511 55
    54527 5
    54662 97
    54823 169
    54857 77
    57083 111
    57073 155
    57461 134
    57494 23
    57679 46
    58027 42
    58238 7
    55606 N/A
    58321 36
    58367 7
    58847 85
    59287 42
    59431 126
    58457 43
    58659 7
    59316 3
    57516 260
    57816 1223
    56294 508
    56778 1892
    57447 458
    56386 440
    56196 522
    57036 398
    53772 779
    52889 1518
    51463 947
    53487 1069
    52866 2296
    53463 1065

  13. Steve McIntyre
    Posted Apr 12, 2007 at 9:00 PM | Permalink

    A CA reader sent me the following link on unidentified site 52983 locating it at Yu Zhong.

    http://weather.gladstonefamily.net/site/52983

    It says:

    The site coordinates appear to be of low precision, and the station may be anywhere within the marked area.

    I guess that was part of Jones’ careful site selection strategy.

  14. Steve McIntyre
    Posted Apr 12, 2007 at 9:12 PM | Permalink

    I wonder if Jones et al site 53161 43 54N 125 20E is really 54161 Changchun 43.90N 125.22E. The Team always sets little booby traps to see if we’re on our toes.

    831 205 54161 0 CHANGCHUN 43.90 125.22 238 283 U 1500 FL xx no -9 x -9 COOL FIELD/WOODS

  15. John Norris
    Posted Apr 12, 2007 at 9:52 PM | Permalink

    re #4

    What are the units for the temperature series? The range goes from -1 to about +225.

    Plotting all the urban locations, the second set of 42, the 40 – 150 numbers (habitable?) look like they might trend up from 1954 to 1983. The others greater than 150 or less than 40 look flat, as one would expect as they might be less habitable.

    It is pushing midnight here, so I may totally misunderstand the temp series, or am just seeing things.

  16. Steve McIntyre
    Posted Apr 12, 2007 at 9:57 PM | Permalink

    Unidentified Jones city: 55606 lat – 28 36 long – 115 58

    appears to be 58606 NANCHANG 28.60 115.92

    I looked at GHCN stations within 1 degree of the lat-long’s from http://www1.ncdc.noaa.gov/pub/data/ghcn/v2/v2.temperature.inv

  17. Steve McIntyre
    Posted Apr 12, 2007 at 11:21 PM | Permalink

    #13. Unidentified Jones site 52983 lat 35.58333 long 103.1833 is probably not Yuzhong.

    In the China archive, there is also:
    130 43 Linxia 52984 35.58 103.18 1917.0 1943 205 14539 16 Bullseye

    which in GHCN v2 is:
    904 205 56080 2 LINXIA 35.58 103.18 1917 2008 U 100 MV xx no -9 x -9

  18. Bob Koss
    Posted Apr 13, 2007 at 1:34 AM | Permalink

    John N,

    Divide those temps by 10 to get degrees celsius.

  19. Bob Koss
    Posted Apr 13, 2007 at 1:55 AM | Permalink

    I might as well post up the elevations for the three new sites and add to the list. They might be an aid for some in visualizing the locations.

    52984 1920m
    54161 238m
    58606 50m

  20. Steve McIntyre
    Posted Apr 13, 2007 at 9:12 AM | Permalink

    For China data, don’t forget
    #ndp039 http://cdiac.ornl.gov/epubs/ndp/ndp039/ndp039.html
    #http://cdiac.ornl.gov/ndps/tr055.html
    #http://cdiac.ornl.gov/ftp/ndp039/

    and interim review of data sources #http://www.climateaudit.org/?p=1193

  21. Michael Jankowski
    Posted Apr 13, 2007 at 9:28 AM | Permalink

    I’m not sure how easy it would be to try and adjust for biases. Take a look at the satellite image for Chengdu, for example:
    google shot. How does one begin to account for the apparent industrial clouds (or is that “natural”)?

    Maybe Chengdu looks a lot differently than it did 30-50 yrs ago, as do many of the cities in the study, but it looks like it could be a mess.

  22. Posted Apr 13, 2007 at 12:10 PM | Permalink

    Further to my comment #11, there is the following.

    Lavely W.R. (1987),
    “Chinese demographic data: a guide to major sources”,
    Australian Journal of Chinese Affairs, 18: 167–178.
    doi:10.2307/2158589

    (The author is currently a professor of sociology at the University of Washington.) Here is a quote: “Only six years ago, the total population of China was not known to within 100 million”.

  23. Steve Sadlov
    Posted Apr 13, 2007 at 12:21 PM | Permalink

    RE: #21 – Typical Chinese city. Industrial smoke and soot, plus all the odd small coal fires for heating, small forges, etc. Some cities have it so bad, you get tricked into thinking there is high overcast on a perfectly sunny day without a cloud in the sky.

  24. John Norris
    Posted Apr 13, 2007 at 9:52 PM | Permalink

    I played around in excel with the 84 sites referenced in #4. 34 of the 84 sites average above 15 degrees c. Those show very flat trends – basically no warming over the 30 years. The other 50 sites averaging below 15 degrees c generally show a slow warming trend.

    If you wanted to claim a lack of GW you could show the warm sites. If you wanted to claim GW you could show you the cold sites. If you wanted to show a lack of UHI, you could give a summary answer of all sites, as the colder sites show some UHI between the rural and urban, which everyone has to believe that there is at least a little of. I suspect the flat liner sites above 15 degrees c, rural and urban, really dumb down the total UHI.

    I have scrunched up excel charts of the 42 rural and 42 urban sites where you can see the warming and lack of warming. We will see if they make it. I know they are ugly. My apologies if they mess up the thread. Please feel free to destroy.

  25. Steve McIntyre
    Posted Apr 13, 2007 at 11:01 PM | Permalink

    I’ve posted up Jones information in ASCII tab-separated format (readable to both Excel and R)

    http://data.climateaudit.org/data/jones90/jones90.info.dat

    This includes the name identification; the corrected wmo number (3 corrections); corrected latitudes (2 corrections) referring to TR055; GHCN v2 identifications; GHCN v2 altitudes, population code, population; xist.org populations (thanks to Doug Hoyt). The difference between GHCN v2 and xist populations is remarkable to say the least. HEre is the collatoin, but it’s easier to read the ASCII file

    id wmo original ghcn data name lat long alt ipop pop xist.pop
    1 50854 50854 50854000 205 Anda 46.38 125.32 150 U 110 473
    2 50949 50949 50949000 205 Qian Gorlos 45.12 124.83 138 R -9 567
    3 54135 54135 54135000 205 Tongliao 43.6 122.27 180 U 80 793
    4 54497 54497 54497000 205 Dandong 40.05 124.33 14 U 450 780
    5 50963 50963 50963000 205 Tonghe 45.97 128.73 110 S 20 203
    6 54324 54324 54324000 205 Chaoyang 41.55 120.45 176 U 60 475
    7 50756 50756 50756000 205 Hailun 47.43 126.97 240 U 75 733
    8 54616 54616 54616000 205 Cangzhou 38.33 116.83 11 U 100 443
    9 54725 54725 54725000 205 Huimin 37.5 117.53 12 S 20 604
    10 54476 54476 54471001 205 Gaixian Xiongyue 40.17 122.15 20 R -9 NA
    11 54936 54936 54945001 205 Juxian 35.58 118.83 107 S 30 1034
    12 54852 54852 54863001 205 Laiyang 36.93 120.7 31 S 30 897
    13 54916 54916 54916000 205 Yanzhou 35.57 116.85 53 U 100 598
    14 57193 57193 57193000 205 Xihua 33.78 114.52 53 R -9 NA
    15 57378 57378 57378000 205 Zhongxiang 31.17 112.57 66 S 45 1022
    16 57584 57584 57584000 205 Yueyang 29.38 113.08 52 R -9 912
    17 57662 57662 57662000 205 Changde 29.05 111.68 35 R -9 1346
    18 58144 58144 58144000 60 QingJiang 33.6 119.03 19 U 110 NA
    19 58150 58150 58150000 205 Sheyang 33.77 120.25 7 R -9 1053
    20 57799 57799 57799000 205 Jian 27.12 114.97 78 U 100 409
    21 58314 58314 58314000 205 Huoshan 31.4 116.32 68 R -9 333
    22 58251 58251 58251000 205 Dongtai 32.85 120.32 5 S 40 1164
    23 58754 58754 58754000 205 Fuding 27.33 120.2 38 S 25 521
    24 59293 59293 59293000 205 Heyuan 23.73 114.68 41 S 25 2782
    25 59417 59417 59417000 205 Longzhou 22.37 106.75 129 R -9 NA
    26 58477 58477 58477000 205 Dinghai 30.03 122.12 37 S 37 369
    27 58646 58646 58646000 205 Li Shui 28.45 119.92 62 R -9 348
    28 59501 59501 59501000 205 Haifen Shanwei 22.78 115.37 5 R -9 344
    29 57537 57537 57633001 205 Pengshui 29.3 108.17 311 S 27 540
    30 57606 57606 57602001 205 Tongzi 28.13 105.83 972 R -9 575
    31 56287 56287 56287000 205 Yaan 29.98 103 629 R -9 338
    32 56751 56751 56751000 205 Dali 25.7 100.18 1992 R -9 186
    33 57253 57253 57259001 205 Yunxian 32.85 110.82 202 S 30 584
    34 57504 57504 57504000 205 Neijiang 29.58 105.05 357 U 240 1385
    35 56193 56193 56193000 205 Pingwu 32.42 104.52 877 R -9 188
    36 57127 57127 57127000 60 HanZhong 33.07 107.2 509 U 120 3480
    37 53863 53863 53863000 205 Jiexiu 37.05 111.93 750 R -9 373
    38 53915 53915 53915000 205 Pingliang 35.55 106.67 1348 U 85 453
    39 51379 51379 51379000 205 Qitai 44.02 89.57 794 U 70 205
    40 53593 53593 53593000 205 Weixian 39.83 114.57 910 R -9 778
    41 52984 52983 56080002 205 Linxia 35.58 103.18 1917 U 100 168
    42 53336 53336 53336000 205 Urad Zhongqi 41.57 108.52 1290 R -9 511
    43 50745 50745 50745000 60 QiQiHaEr 47.38 123.92 148 U 1500 1540
    44 50953 50953 50953000 60 HaErBin 45.68 126.62 143 U 2750 3695
    45 54161 53161 54161000 60 ChangChun 43.9 125.33 238 U 1500 3225
    46 54342 54342 54342000 60 ShenYang 41.77 123.43 43 U 3750 5303
    47 50978 50978 50978000 205 Jixi 45.28 130.95 234 U 350 910
    48 54237 54237 54236001 205 Fuxin 42.03 121.65 144 U 350 785
    49 50774 50774 50774000 205 Yichun 47.72 128.9 232 U 200 814
    50 54511 54511 54511000 60 BeiJing 39.93 116.28 55 U 8500 12843
    51 54527 54527 54527000 60 TianJin 39.1 117.17 5 U 7210 8146
    52 54662 54662 54662000 60 DaLian 38.9 121.63 97 U 1480 3245
    53 54823 54823 54823000 60 JiNan 36.68 116.98 58 U 1500 3506
    54 54857 54857 54857000 60 QingDao 36.07 120.33 77 U 1900 2721
    55 57083 57083 57083000 60 ZhengZhou 34.72 113.65 111 U 1500 6656
    56 57073 57073 57071001 205 Luoyang 34.67 112.42 136 U 750 1202
    57 57461 57461 57461000 60 YiChang 30.7 111.3 134 U 150 1338
    58 57494 57494 57494000 60 WuHan 30.62 114.13 23 U 4250 8312
    59 57679 57679 57679000 60 ChangSha 28.2 113.08 46 U 850 2122
    60 58027 58027 58027000 60 XuZhou 34.28 117.15 42 U 1500 1679
    61 58238 58238 58238000 60 NanJing 32.05 118.78 12 U 2000 3405
    62 58606 55606 58606000 60 NanChang 28.6 115.97 50 U 900 1844
    63 58321 58321 58321000 205 Hefei 31.87 117.23 36 U 400 1659
    64 58367 58367 58362000 60 ShangHai 31.17 121.43 7 U 10980 13065
    65 58847 58847 58847000 60 FuZhou 26.08 119.28 85 U 900 2124
    66 59287 59287 59287000 60 GuangZhou 23.13 113.32 8 U 2300 9496
    67 59431 59431 59431000 60 NanNing 22.82 108.35 73 U 375 6463
    68 58457 58457 58457000 60 HangZhou 30.23 120.17 43 U 1100 4502
    69 58659 58659 58659000 60 WenZhou 28.02 120.67 7 U 250 1915
    70 59316 59316 59316000 60 ShanTou 23.4 116.68 3 U 400 4944
    71 57516 57516 57516000 60 ChongQing 29.58 106.47 351 U 3500 10266
    72 57816 57816 57816000 60 GuiYang 26.58 106.72 1074 U 1500 2374
    73 56294 56294 56294000 60 ChengDu 30.67 104.02 508 U 2000 4645
    74 56778 56778 56778000 60 KunMing 25.02 102.68 1892 U 1300 5781
    75 57447 57447 57447000 205 Enshi 30.28 109.47 458 S 40 755
    76 56386 56386 56385001 205 Leshan 29.57 103.75 424 U 250 1135
    77 56196 56196 56196000 205 Mianyang 31.47 104.68 472 U 52 1129
    78 57036 57036 57036000 60 XiAn 34.3 108.93 398 U 1900 5345
    79 53772 53772 53772000 60 TaiYuan 37.78 112.55 779 U 2725 2558
    80 52889 52889 52889000 60 LanZhou 36.05 103.88 1518 U 1500 2087
    81 51463 51463 51463000 60 WuLuMuQi 43.78 87.62 919 U 500 1753
    82 53487 53487 53487000 205 Datong 40.1 113.33 1069 U 300 1526
    83 52866 52866 52866000 60 XiNing 36.62 101.77 2262 U 250 854
    84 53463 53463 53463001 60 HuHeHaoTe 40.8 111.63 1063 U 700 1407

    The temperature versions at TR055 appear to match the versions now at CRU.

  26. Willis Eschenbach
    Posted Apr 14, 2007 at 1:47 AM | Permalink

    Steve, thanks for the data. You say:

    The difference between GHCN v2 and xist populations is remarkable to say the least.

    Not only are the differences remarkable, the differences are greatly skewed between the rural and urban stations in the two datasets. Here are the figures for the xist population/GHCN population:

    xist/GHCN Rural = 1796% (95% CI = 894% to 2697%)
    xist/GHCN Urban = 420% (95% CI = 275% to 566%)

    Another way to look at it is the difference between the average “rural” and “urban” population centres. In the GHCN data, the rural centers average 84,000 people, and the cities average 1,812,000. Thus the cities average 22 times the population of the rural centres.

    In the xist data, on the other hand, the rural centers average 838,000 people, ten times the GHCN population, and hardly “rural”. The cities average 3,787,000, only twice the GHCN figures. In the xist data, the cities only average 5 times the population of the “rural” centres.

    Once again, we have climate scientists using very dubious data to draw statistically unsustainable conclusions. As I pointed out in a post on another thread, the overwhelming majority (68 of 84) of the stations show no significant trend, and out of the 42 paired stations, only three pairs are statistically different. Add that to the dubious population data … I say there is not one meaningful conclusion that can be drawn from this dataset.

    w.

  27. jcspe
    Posted Apr 14, 2007 at 5:54 AM | Permalink

    While I appreciate the need to quantify UHI effects, they may actually be a secondary question. Even if you are able to make heads or tails of the Jones data, I would wager that one would still need substantially more information to reach any solid conclusions.

    http://pafc.arh.noaa.gov/climvar/climate-paper.html, reads in part:

    An additional consideration is the movement and re-location of a particular climate station within a city or town. Let’s use Anchorage as an example. The official weather station for the city was established in 1915 and was located near the lower end of Ship Creek. Between 1923-1943 the station was re-located several times in what is now the downtown area. From 1943-1953 the station was located at Merrill Field, but from 1953 to the present it has been sited at Ted Stevens International Airport. These station re-locations are important because as everyone who lives in Anchorage knows, winter temperatures (and precipitation) vary considerably from west-to-east and north-to-south across the city. These spatial variations in the data have to be taken into consideration during climate analysis, otherwise they might indicate a change in climate which in reality is due to the re-location of the equipment.

    Anchorage’s population is around 300,000. It sits on a little glacial outwash that is about 10 miles north-south and about 8 miles east-west. The lower end of Ship Creek is in northwest Anchorage at approximately 10-20 FT above MSL, where temperatures are greatly affected by Cook Inlet. Both the downtown area and the international airport are on the west side at about 100 FT MSL. The seawater in Cook Inlet also affects the temperatures of these areas, but less so than the Ship Creek by noticable amounts. Merrill Field is approximately 2 miles east of downtown and farther from the water. There are many days when temperatures at Merrill Field are about 5 degrees F different than the international airport. The east side of the Anchorage Bowl (Muldoon area) is about 6 miles east of downtown at about 300 FT MSL. Just east of Muldoon are the Chugach Mountains. I would guess the tree line is only approximately 8 miles east of dowtown at about 1500-1800 FT MSL or so. See http://www.muni.org/iceimages/Planning/vicinity.gif for a vicinity map.

    During the warmest and coldest days each year, the temperature differences in the few miles between MSL and 300FT MSL can be 15-20 degrees F. Maybe this particular town is unique, or maybe there are a lot of towns where temperature can vary this much in just a few miles. However, moving any weather station opens the whole issue to a far greater uncertainty than any estimate of the UHI effect I have seen.

    I would want to see proof of very good recording keeping in the Chinese data regarding the actual location of each station for the whole of recorded period before I would believe any conclusion at all about UHI calculations.

  28. Posted Apr 14, 2007 at 7:41 AM | Permalink

    The following shows whether on not a station relocated, during 1954–1983. Only the 35 stations for which histories are available are listed. An “X” indicates at least one relocation; an “=” indicates none; a “?” indicates inconsistent history. (A * indicates an incorrect station number.)


    50745 X
    50953 X
    51463 X
    52866 X
    52889 X
    53463 =
    53772 X
    54161* =
    54342 =
    54511 X
    54527 (first year only)
    54662 X
    54823 =
    54857 ?
    56294 ?
    56778 =
    57036 ?
    57083* (first year only)
    57127 X
    57461 ?
    57494 X
    57516 =
    57679 X
    57816 =
    58027 X
    58144 X
    58238 =
    58367 X
    58457 X
    58606* X
    58659 =
    58847 =
    59287 X
    59316 ?
    59431 X

    As an example of inconsistent history, consider station #56294. This shows its address in the 1970s being “as above”, referring its address in the 1960s. The longitude of the address, however, changed from 104°04′ to 104°01′ (about 4 km).

  29. KevinUK
    Posted Apr 14, 2007 at 10:13 AM | Permalink

    #3

    OK I’ve continued to read the rest of this thread and the others related to it. This far it’s clear that there are lots or problems with this dataset that is used as the primaty proof of negligible UHI effects in the mean global surface temperature currently claimed warming trend. What other alternative studies exist which have looked into the UHI effect where there are large difference between rural and urban populations (Europe, US/Canada, USSR, India)?

    KevinUK

  30. Posted Apr 14, 2007 at 11:16 AM | Permalink

    Further to my comment #28, I now see that there are inconsistent histories for stations 54342, 58238, and 58659. Thus, each of those should be marked with “?” (rather than “=”). Such inconsistent histories further emphasize the poor quality of the data.

    The 84 stations can be classified as follows:
    49 with no histories;
    08 with inconsistent histories;
    18 with substantial relocations;
    02 with single-year relocations;
    07 without relocations.
     
    Separately, note the following from p.64 of the CAS-ORNL report: “station (No. 54852) shows identical July and August temperatures over the period 1973–1975″—the indication being that the data (from a station among the 49 without histories) is unreliable.

  31. Douglas Hoyt
    Posted Apr 14, 2007 at 12:36 PM | Permalink

    Ren et al looked at urban heating in China and found that 65% to 80% of the warming was due to UHI for 1960 to 2000. The paper is discussed at http://climatesci.colorado.edu/2007/03/16/two-papers-on-the-urban-heat-island-effect-on-temperatures/

    Population growth in Chinese cities from 1950 to 2000 is tabulated at http://www.xist.org/charts/cy_aggmillion2.aspx If you use the populations there and the following formula to calculate the trend in temperatures caused by growth in urban development:

    UHI trend = 1.5 * log (pop2000/pop1950)/ 5

    where 5 is for 5 decades, pop2000 is the population in 2000 and pop1950 is the population in 1950. You get a predicted UHI warming of 0.190 C/decade for the 139 listed cities.

    The formula above is pretty standard and comes from Torok’s study of UHI in Australia.

    HadCRUT3 gives a warming for eastern China of about 0.238 C/decade (from CO2science website) so it looks to me like about 80% UHI warming may be about right as Ren et al conclude.

  32. Posted Apr 15, 2007 at 4:15 AM | Permalink

    Further to my comments in another thread that discuss the moves of individual stations….

    Station #57679 moved 24–25 km; map shown below (from http://www.informath.org/WMO57679.png)

  33. Posted Apr 15, 2007 at 4:16 AM | Permalink

    Station #57127 moved 10–11 km; map below (from http://www.informath.org/WMO57127.png)

  34. Posted Apr 15, 2007 at 4:27 AM | Permalink

    Station #57494 moved 20–21 km; map below (from http://www.informath.org/WMO57494.png)

    The brown-shaded area shows the city proper; the metropolitan area is much larger. There were actually three locations in all; the third is not shown (it followed a move that occurred in mid 1982).

    (Note that Wuhan is one of the two cities analyzed by G.Y. Ren et al. [GRL, 2007]. The authors begin their analysis in 1961, the year after the relocation shown in the map.)

  35. Willis Eschenbach
    Posted Apr 15, 2007 at 5:25 AM | Permalink

    KevinUK, you ask a good question:

    OK I’ve continued to read the rest of this thread and the others related to it. This far it’s clear that there are lots or problems with this dataset that is used as the primaty proof of negligible UHI effects in the mean global surface temperature currently claimed warming trend. What other alternative studies exist which have looked into the UHI effect where there are large difference between rural and urban populations (Europe, US/Canada, USSR, India)?

    The abstract of a recent study in Climate Research [CR 33:159-169(2007)] says:

    Recent California climate variability: spatial and temporal patterns in temperature trends

    Steve LaDochy1,*, Richard Medina1,3, William Patzert2
    1Department of Geography & Urban Analysis, California State University, 5151 State University Drive, Los Angeles, California 90032, USA
    2Jet Propulsion Laboratories, NASA, 4800 Oak Grove Drive, Pasadena, California 91109, USA
    3Present address: Department of Geography, University of Utah, 260 South Central Campus Drive, Salt Lake City, Utah 84112-9155, USA
    *Email: sladoch@calstatela.edu

    ABSTRACT: With mounting evidence that global warming is taking place, the cause of this warming has come under vigorous scrutiny. Recent studies have lead to a debate over what contributes the most to regional temperature changes. We investigated air temperature patterns in California from 1950 to 2000. Statistical analyses were used to test the significance of temperature trends in California subregions in an attempt to clarify the spatial and temporal patterns of the occurrence and intensities of warming.

    Most regions showed a stronger increase in minimum temperatures than with mean and maximum temperatures. Areas of intensive urbanization showed the largest positive trends, while rural, non-agricultural regions showed the least warming. Strong correlations between temperatures and Pacific sea surface temperatures (SSTs) particularly Pacific Decadal Oscillation (PDO) values, also account for temperature variability throughout the state. The analysis of 331 state weather stations associated a number of factors with temperature trends, including urbanization, population, Pacific oceanic conditions and elevation. Using climatic division mean temperature trends, the state had an average warming of 0.99°C (1.79°F) over the 1950’€”2000 period, or 0.20°C (0.36°F) decade’€”1. Southern California had the highest rates of warming, while the NE Interior Basins division experienced cooling.

    Large urban sites showed rates over twice those for the state, for the mean maximum temperatures, and over 5 times the state’s mean rate for the minimum temperatures. In comparison, irrigated cropland sites warmed about 0.13°C decade’€”1 annually, but near 0.40°C for summer and fall minima. Offshore Pacific SSTs warmed 0.09°C decade’€”1 for the study period.

    Note that over the last 50 years, the increase in maximum temperatures was twice as large as the mean rate for large urban sites, and the increase in minimum temperatures was five times as large. This means that the mean temperature increase in urban areas was 3.5 times as large as the mean rate statewide. The mean rate statewide was 0.2°C/decade, so that means the urban sites averaged 0.7°C/decade … I’d call that significant UHI warming myself.

    Unfortunately, the full report is extremely expensive (€80), so I think I’ll have to pass on it … there is also an interesting analysis of trends in California temperatures by county at Warwick Hughes’s site.

    As a rough rule of thumb, Studies by Torok et al. (2001) and Oke (1973) indicate that the UHI effect is about 1.5°C * log(population/population0). The data on Warwick Hughes page, however, gives a slightly higher value of 1.8°C.

    w.

  36. Dave Dardinger
    Posted Apr 15, 2007 at 8:15 AM | Permalink

    re: #35 Willis,

    And for those wondering, the 1950 world population was 2.556 billion and the 2000 world population was 6.081 billion (according to the first site I turned to) which gives an *expected* UHI according to that formula of .565 deg C. Given that the Jones figure used is .05 deg C for an entire century, and the total temperature rise for the century was about .7 deg C, It’d seem to me that the UHI question needs some real examination.

    BTW, grabbing the 1900 & 2000 figures from Wikipedia (1.650 and 6.071 billion) I get .85 deg C expected UHI for the entire century, which is actually, I believe, more than the actual observed warming. So, just how strong is the evidence for the UHI formula? You can’t try to justify it simply by looking at population in a few cities since the population signal and the putative AGW signal will be highly correlated in any case. You need to find locations which have experienced both very high and very low population increases and which are isolated from confounding problems like other nearby urban areas or say areas of major land use change.

  37. Willis Eschenbach
    Posted Apr 15, 2007 at 4:00 PM | Permalink

    Dave, an interesting post. I don’t know the answer to your questions. The Torok paper is available here. He gives a value of 1.42 log population. I’ve never been able to locate the original Oke paper, but there is an analysis of it (by the usual suspects) here. Dr. Oke has written extensively on the question, a list of his publications is here.

    However, of course, such a rule of thumb is just that. UHI varies by latitude, with colder regions having a greater winter UHI for a given population than warmer regions. For an extreme example of UHI in a small town, check out Barrow, Alaska.

    w.

  38. beng
    Posted Apr 16, 2007 at 9:30 AM | Permalink

    RE #37

    The Barrow UHI study is worth reading & understanding. Keeping in mind it’s an arctic site, it’s still remarkable that it has such an easily detectable UHIE w/a mere 4600 population in a windy, exposed coastal location. Another is that they conclude most of the UHI comes from direct energy conversion. I thought that aspect was negligible at least in my region — that most UHI seems to result from solar heating on dry, heat-retaining urban surfaces like concrete/asphalt/roofing.

    Yeah, right, we (Jones, IPPC, media) “know” that global UHI is only .05C for the last century.

  39. pochas
    Posted May 23, 2007 at 12:02 PM | Permalink

    This from the Idsos:

    http://www.co2science.org/scripts/CO2ScienceB2C/articles/V10/N21/C2.jsp

  40. Steve Sadlov
    Posted May 23, 2007 at 12:32 PM | Permalink

    RE: #38 – one need only look at the explosion in high tension power lines and stand alone motors since 1900 to see just how much of UHI is due to energy conversion. All those electrons and BTUs need to do some work and they have and continue to do so.

  41. bender
    Posted May 23, 2007 at 12:39 PM | Permalink

    And the original abstract:

    Temporal change in urbanization-induced warming at two national basic meteorological stations of China and its contribution to the overall warming are analyzed. Annual and seasonal mean surface air temperature for time periods of 1961-2000 and 1981-2000 at the two stations of Beijing and Wuhan Cities and their nearby rural stations all significantly increase. Annual and seasonal urbanization induced warming for the two periods at Beijing and Wuhan stations is also generally significant, with the annual urban warming accounting for about 65-80% of the overall warming in 1961-2000 and about 40-61% of the overall warming in 1981-2000. This result along with the previous researches indicates a need to pay more attention to the urbanization-induced bias probably existing in the current surface air temperature records of the national basic stations.

    twq, where r u?

  42. Steve McIntyre
    Posted May 23, 2007 at 12:51 PM | Permalink

    One of the things that has come out of my first inspection of the GISS stations is the tremendous increase in the proportion of stations in the post-1990 period that come from urban airports. This is because the majority of the updated data in CRU and GISS other than the U.S. is from the automated airport weather system. I suspect that this may underpin the fanatical obstruction from CRU to even identifying their stations.

    So it’s not just a UHI effect but an urban airport effect. I’ll do a post on this some time. In Toronto, the landscape around the airport has been urbanizing very rapidly. It was on the outskirts of the city when I was a boy and now it’s urban. This must be happening all over the world.

  43. bender
    Posted May 23, 2007 at 1:04 PM | Permalink

    Ren et al. (2007) Fig. 3 has urban temperatures cooler than rural, right up to 1990. That’s more than half the reason for the urban vs. rural trend difference. Curious. That’s not normal. Is this a case of agri-rural vs forested/pre-airport, with agricultural warmer than forested pre 1990??

    Then, a curious statement (bolded) in the conclusion (context kept, for accuracy):

    The annual urban warming at the city stations can account for about 65-80% of the overall warming in 1961-2000, and about 40-61% of the overall warming in 1981-2000. The quality control and the in-homogeneity examination and adjustment for the data of the stations used for the analysis have been made.

    I’d like to see the data and code for this whole paper.

  44. bender
    Posted May 23, 2007 at 1:06 PM | Permalink

    Just noticed for the first time the comment in the opening post:

    Note that the rural station is warmer than the urban station here.

    Duly noted!

  45. bernie
    Posted May 23, 2007 at 2:38 PM | Permalink

    It seems to me that since we are looking at temperature anomalies, a more productive way to classify stations is “high growth vs no growth” or “high development” versus “low development”. As noted above you can find UHI in towns of 4600 people. Based on what I saw in the census data for Cambridge Bay, Canada you could easily find UHI effects in towns of 1000 that have grown in the number of buildings by 50% in 20 years.
    I assume that in China there is also the possibility of UHI tied to dramatic improvements in SOL with the parallel increase in energy consumption.

    I was also interested to see that the just released paper on CO2 emissions uses a model that has population growth as a primary driver of energy consumption – the Kaya identity. The model suggests on its face that UHI should be calculated by correlating temperature changes with net changes not density per se: The rural urban distinction is bogus both in terms of actual station classifications and as a means of evaluating the UHI with respect to estimates of changes in global temperatures. This is yet another reason a reluctance to provide the locations of stations. There is almost a guaranteed positive correlation of local temperature and local population growth that would have to be partialled out before any crude aggregation! I recall seeing an equation that estimates this effect, at least on a global basis.

3 Trackbacks

  1. [...] climateaudit, noting that questions about these claims were raised here (for example, here here and here . Since then, we’ve also looked at adjustments in the USHCN, GHCN and GISS networks, observing [...]

  2. [...] 2007, following receipt of the data, I did a number of posts at CA on the Chinese network e.g. here here here here here, analysis that we now know that Jones was monitoring. One of the few mentions [...]

  3. [...] Immediately on receipt of this information, I wrote some interesting posts on Chinese stations here here here . Doug Keenan followed up on this information as [...]

Follow

Get every new post delivered to your Inbox.

Join 3,246 other followers

%d bloggers like this: