The Impact of Urbanization on Land Temperature Trends

by Zeke Hausfather , Steven Mosher, Matthew Menne , Claude Williams , and Nick Stokes

[Note: this is an AGU poster displayed at the annual meeting, available here as a PDF. I’ve converted it to plain text and images for your reading pleasure. I’m providing it without comment except to say that Steven Mosher has done a great deal of work in creating a very useful database that better defines rural and urban stations better than the metadata we have available now. – Anthony]

Introduction
Large-scale reconstructions of surface temperature rely on measurements from a global network of instruments. With the exception of remote automated sensors, the locations of the instruments tend to be correlated with inhabited areas. This means that urban
ares [sic] are probably oversampled in surface temperature records relative to the total land surface that is actually urbanized.

It has long been known that urbanized areas tend to have higher temperatures than surrounding less developed (or rural) areas due to the concentration of high thermal mass impermeable surfaces (Oke 1982). This has led to some concern that changes in
urban heat island (UHI) effects due to rapid urbanization in many parts of the world over the past three decades may have been responsible for a portion of the rapid rise in measured global land surface temperatures. This concern is reinforced by lower
observed trends in some interpretations of satellite measurements of lower tropospheric temperature over land areas during the same period (Klotzbach et al 2009).


An analysis of the impact of urbanization on temperature trends faces multiple confounding factors. For example, an instrument originally installed in a city frequently will have warmer absolute temperatures than one in a nearby rural area (especially at night), but will not necessarily show a higher trend over time unless the environs change in such a way that the UHI signal is altered in the vicinity of the instrument. Similarly, microsite characteristics that may be unrelated to the larger urban environment can have
notable effects on temperature trends and act counter to or in concert with the ambient UHI signal.

Moreover, the definition of urban areas is subject to some uncertainty, both in terms of how urban form is characterized and at what distance from built surfaces urban-related effects persist. Published station metadata often includes outdated indications of whether a station is urban or rural, and instrument geolocation data can be imprecise, out of date, or otherwise incorrect.

There is also uncertainty over how much explicit correction is needed for urban warming in global temperature reconstructions, and how well homogenization techniques recently introduced in GHCN-Monthly version 3 both detect and correct for inhomogenities
arising from changes in urban form.

To address these issues and obtain a more accurate estimation of the impact of urbanization on land temperature trends, we examine different urbanity proxies at multiple spatial resolutions and urbanity selection criteria through both simple spatial
weighting and station pairing techniques. This study limits itself to unadjusted average temperature data, though we will examine homogenized data in the future to see how much of the UHI signal is removed.

Methods


We examine GHCN-Daily version 2.80 temperature data rather than the more commonly used GHCN-Monthly data as it contains significantly more stations, particularly during the past thirty years, and allows for separate examination of maximum and minimum
temperatures. A relatively high spatial density of stations is useful to allow sampling into various urban and rural station subsets while minimizing biases due to loss of spatial coverage. After excluding stations that have fewer than 36 months at any time in the
period of record or at least one complete year of data during the 1979 to 2010 period, we are left with 14,789 stations.

A complete set of metadata is calculated for each station using the station location information provided in station inventories and publically available GIS datasets. These datasets include: Distance From Coast (0.1 deg), Hyde 3.1 historical population data (5
arc minute), 2000AD Grump Population density (30 arc seconds), Grump Urban Extent, Land use classes from the Harmonized Land Use inventory (5 arc minutes), radiance calibrated Nightlights (30 arc seconds), ISA- Global Impervious Surfaces (30 arc
seconds), Modis Landcover classes (15 arc seconds), and distance from the closest airport (30 arc seconds). In addition, area statistics at progressive radii are calculated around each putative site location.

Stations are then divided into two classes based on various thresholds for urbanity and two analytical methods are used to estimate the bias in trend due to urbanity: a spatial method and a paired station approach. The spatial averaging method relies on
solving a set of linear equations for the stations in each class. For each group of stations, urban and rural, a time series of average temperature offsets was created by fitting the model:

where T represents the observed temperature for each station, month and year, L is a local average temperature for each station for each month (incorporating seasonal variation) and G is the desired global (or regional) average, varying by year. This is fitted
with a weighting that is inversely proportional to a measure of station density. With a G calculated for both urban and rural, the trends can be compared.

The pairwise method proceeds with the same classification of stations and the following steps are taken. An urban base pair is selected based on the length of record. To qualify as a base urban pair a station must have 30 complete years of data in the 1979-2010 window.

Ten out of 12 months of data are required to count as a complete year. For every urban base station rural pairs are selected based on distance and data overlap. For every urban base station the rural stations are exhaustively searched and all those rural pairs within 500km are assigned to the base station. Since rural stations may have short records the entire rural ensemble is evaluated for data overlap with the urban base pair. 300 months of overlap are required. If the collection of rural stations has less than 300 months of overlap with its urban pair, it is dropped from the analysis. A weighting function is deÞned in the neighborhood of each urban station, which diminishes with distance and is zero outside a certain radius. An average trend is computed for the rural stations within that radius by fitting the model

where t is time in years, and B is the gradient. This trend is then compared with the OLS trend for the central urban station. The differences in the shapes of the distributions of the trends is a function of the number of stations that form the trend estimation.

Urban trends are trends for individual stations, while rural trends are the result of computing a trend for all the rural pairs taken as a complete ensemble.

Discussion

While urban warming is a real phenomenon, it is overweighted in land temperature reconstructions due to the oversampling of urban areas relative to their global land coverage. Rapid urbanization over the past three decades has likely contributed
to a modest warm bias in unhomogenized global land temperature reconstructions, with urban stations warming about ten percent faster than rural stations in the period from 1979 to 2010. Urban stations are warming faster than rural stations on average across all urbanity proxies, cutoffs, and spatial resolutions examined, though the underlying data is noisy and there are many individual cases of urban cooling. Our estimate for the bias due to UHI in the land record is on the order of 0.03C per decade for urban stations.

This result is consistent with both the expected sign of the effect and regional estimates covering the same time period (Zhou et al 2004) and differs from some recent work suggesting zero or negative UHI bias (Wickham et al, submitted).


Stricter urbanity proxies that result in a smaller set of rural stations show larger urban-rural differences in trend. The upper limit on UHI bias between rural and urban stations is on the order of 0.06 to 0.1C per decade. However, these cases are clearly problematic from the spatial coverage aspect, as the number of rural stations becomes vanishingly small when the most stringent filters are applied. Adopting cutoffs that define rural less strictly leads to more reasonable spatial coverage and an estimate of UHI bias in the record that converges on 0.02C to 0.04C per decade across the proxies. The station pair approach avoids this issue by limiting the analysis to areas with both rural and urban stations available, but has limited global coverage and excludes large areas in India and coastal China where rapid urbanization has been occurring in recent decades.

It is likely that homogenization will further reduce the observed UHI-related bias, as many urbanity biases are detectable through break-point analysis via comparison to surrounding rural stations. We are currently in the process of using the Pairwise Homogenization Algorithm (Menne and Williams 2009) on GHCN-Daily data to examine the effects in more detail. However, it remains to be seen to what degree UHI bias can be removed via homogenization in areas like coastal China and India where there are few rural stations and where station densities are not particularly high in the current version of GHCN-Daily. In any case, the acquisition of additional station data outside of urban areas in these parts of the world would likely be benefitial.

Acquiring more accurate station location data will allow us to use higher-resolution remote sensing tools to identify urban characteristics below the 5 km threshold, and better test effects of site-specifc vs. meso-scale characteristics on urban warming biases. In addition, validated site locations allows for more refinement in the definition of rural stations as a function of distance from urban cores of various sizes.

References

Center for International Earth Science Information Network (CIESIN), Columbia University; International Food Policy Research Institute (IFPRI); The World Bank; and Centro Internacional de Agricultura Tropical (CIAT). 2004. Global Rural-Urban
Mapping Project, Version 1 (GRUMPv1): Population Density Grid. Palisades, NY: Socioeconomic Data and Applications Center (SEDAC), Columbia University. Available at http://sedac.ciesin.columbia.edu/gpw.[Aug 14, 2011].
Elvidge, C.D., B.T. Tuttle, P.C. Sutton, K.E. Baugh, A.T. Howard, C. Milesi, B. Bhaduri, and R. Nemani, 2007, “Global distribution and density of constructed impervious surfaces”, Sensors, 7, 1962-1979
Fischer, G., F. Nachtergaele, S. Prieler, H.T. van Velthuizen, L. Verelst, D. Wiberg, 2008. Global Agro-ecological Zones Assessment for Agriculture (GAEZ 2008). IIASA, Laxenburg, Austria and FAO, Rome, Italy.
Klein Goldewijk, K. , A. Beusen, and P. Janssen (2010). Long term dynamic modeling of global population and built-up area in a spatially explicit way, HYDE 3 .1. The Holocene20(4):565-573.
Klotzbach, P., R. Pielke Sr., R. Pielke Jr., J. Christy, and R. T. McNider, 2009. An alternative explanation for differential temperature trends at the surface and in the lower troposphere. J. Geophys. Res.
Menne, M.J., I. Durre, R.S. Vose, B.E. Gleason, and T.G. Houston, 2011: An overview of the Global Historical Climatology Network Daily Database. Journal of Atmospheric and Oceanic Technology, submitted.
Menne, M.J., and C.N. Williams, Jr., 2009. Homogenization of temperature series via pairwise comparisons. J. Climate, 22, 1700-1717.
Schneider, A., M. A. Friedl and D. Potere (2009) A new map of global urban extent from MODIS data. Environmental Research Letters, volume 4, article 044003.
Schneider, A., M. A. Friedl and D. Potere (2010) Monitoring urban areas globally using MODIS 500m data: New methods and datasets based on urban ecoregions. Remote Sensing of Environment, vol. 114, p. 1733-1746.
T. R. Oke (1982). “The energetic basis of the urban heat island”. Quarterly Journal of the Royal Meteorological Society 108: 1–24.
Wickham, C., J. Curry, D Groom, R. Jacobsen, R. Muller, S. Perlmutter, R. Rohde, A. Rosenfeld, and J. Wurtele, 2011. Inßuence of Urban Heating on the Global Temperature Land Average Using Rural Sites IdentiÞed from MODIS ClassiÞcations.
Submitted.
Zhou, L., R. Dickinson, Y. Tian, J. Fang, Q. Li, R. Kaufmann, C. Tucker, and R. Myneni, 2004. Evidence for a signiÞcant urbanization effect on climate in China. Proceedings of the National Academy of Sciences.
Ziskin, D., K. Baugh, F. Chi Hsu, T. Ghosh, and C. Elvidge, 2010, “Methods Used For the 2006 Radiance Lights”, Proceedings of the 30th Asia-PaciÞc Advanced Network Meeting, 131-142

134 thoughts on “The Impact of Urbanization on Land Temperature Trends

  1. Another microsite issue is “snuggling up”, for lack of a better term. As data collection went electronic, direct cabling substituted for observers’ feet, and recording stations ended up close in to recording buildings, virtually guaranteeing that their measurements lost almost all benefits of natural surroundings and isolation. This applies even to many rural stations.

  2. As I read it, this paper finds average UHI in USA to be 0.06 to 0.1C per decade with the number of genuine rural stations being “vanishingly small”. As the spatial reality is that the rural percentage is dominant, the sample is clearly unrepresentative.

    However, as it is not “reasonable” to have such biased spatial coverage, the definition of “urban” has been eased off so that the category covers only those stations which converge on 0.2 to 0.4C.

    Is this a case of changing the data to fit the hypothesis?

  3. The upper limit on UHI bias between rural and urban stations is on the order of 0.06 to 0.1C per decade. However, these cases are clearly problematic from the spatial coverage aspect, as the number of rural stations becomes vanishingly small when the most stringent filters are applied.

    This is not a justification for using less stringent rurality criteria. That the right answer is difficult to achieve does not make the wrong answer more correct.

  4. While it is beyond the scope of your research, I would also be interested in the impact of multidecadal changes in atmospheric weather patterns on UHI over long periods of time. For example, if we were to see a fairly reliable seasonal storm track make a persistent move in location, we might see a significant change in regional temperature trends. This would be particularly so if the track moved from impacting a more urban location to a more rural or vice versa. If a city, for example, were to get fairly regular seasonal rainfall for a period of, say 30 years, and then the dominate storm track were to move, that urban area might experience a lot more solar heating. A rural area that has gone from mostly sunny conditions to more stormy conditions might not see as much of a change in temperature so the impact of the movement of the storm track might act to cause an amplified response.

    If the move of the weather pattern moved from one rural area to another or one urban to another, it might not have as great an impact. What prompted this thought was a paper I read a while back where evidence was found of past movement of the ITCZ in Africa where it has apparently changed position by some 300 miles for a considerable period of time. Basically, it moves from time to time. I was then thinking of how the PDO might impact storm tracks in the US, particularly along the West Coast. Storms making landfall around the Portland/Beaverton/Vancouver area reliably over a long period of time might have considerable impact on the “normal” temperatures of those cities. Should the storm track move a considerable distance for a fairly significant amount of time leaving those cities with more sunshine, the temperature trends for the Pacific Northwest might be improperly biased when what we are actually seeing in the trend is solar heating of the concrete and not really much of a real change in climate. Whereas a movement in the opposite direction, from landfall in a relatively rural area to a track over a major population center would show a rather pronounced negative trend in the data that is exaggerated by the attenuation of UHI due to increased cloud cover.

    Considering how sensitive policy makers are these days to tiny changes measured in tenths of a degree and how these tiny changes over short periods are often extrapolated over long periods for planning purposes, such changes could cause over-reactions and adoption of needless and expensive policy measures to “combat” something that is simply a reflection of a natural cyclical storm track change and not a general change in global climate.

    I’ll be interested to see what impact your homogenization techniques have on the difference in urban-rural trends. Frankly, for purposes of climate monitoring, I don’t think we should be using urban stations at all. I believe we should be using rural or parkland for locations for such monitoring. Most states have many acres of state/national forest/park/preserve land available for such monitoring locations.

  5. Well, the first thing I want to say is that what a biased map you show up at the very beginning! How many rural sites are partially (or even fully) obscured by urban sites—particularly in the US? This is not really better than the way Mann et al. enjoy putting a big fat line on top of their reconstructions to hide any recent insensitivity. I’m very sensitive to people who try to lie (or mislead) with graphs.

  6. I wondered about that, too, Alexander, because I know there is some pretty rural area in Kansas and Nebraska and the Dakotas that show up as “urban” in that map. As in most of those states are rural except for a few smallish cities. Heck, some places out there you can go 40 miles between tiny little towns.

  7. Thanks for posting this poster. It deserves a careful and thoughtful read. I have used some of these same techniques in my work on ore-bodies. The pair wise comparison can be very useful but as pointed out can sometimes be problematic and needs to be applied with care. In some locations it may not work. It seems to me, instead of always needing to make adjustments for urban growth and other changes, we should quickly and diligently establish new rural locations that will remain unchanged or unaffected for a long period of time. If enough true rural stations are available we would not need to use “adjusted data” which will always present difficulties.

  8. Considering that virtually NONE of the existing global historical temperature data is of proper sample size nor of a proper random sampling regime to obtain a reasonable confidence level or interval, it barely matters that UHI sampling bias is present. A sample size of 1 from a effectively infinite population size for each of 30 days does not entitle the scientist to 30 DF for minimum nor maximum temperature. The corruption of the science of sampling is truly awesome all the way down to the purported precision down to hundredth of a degree..

    Neither have I ever seen an honest accounting of the actual extent / range of all the various errors in the data set.

    GIGO – the data is trash no matter how you stack it.

  9. One thing to consider is that for a lot of these sites, there will also be wind data. You could also extract wind data from the weather maps at the time on a large scale.

    So now you could test a site that is to the east of an urban area. You also need a site to the west of an urban area. Depending on the wind direction you either get no difference, or a bias one way or the other.

  10. Interesting work.

    You say;

    “Ten out of 12 months of data are required to count as a complete year.”

    Do those ‘missing months’ vary-i.e one record misses out January and February, whilst another misses out June and July?

    How is the missing data accounted for?
    tonyb

  11. Thanks for posting this, v interesting work to read. Congratulations.
    The pdf link shows the whole poster.

    The two maps on the bottom left of the poster, when viewed at 150% show that
    1. Map of Urban & Rural Stations – One station in the Antarctica ice & one to the NE in the sea. None in the Arctic
    2. Map of Urban % Rural Pairs – None in the Antarctic and none in the Arctic

    Studies that use twins would provide a useful baseline methodology!

  12. Why do I, in these studies, not see more of formulations like
    – CHANGE through time in the DEGREE of urbanization/ruralness (probably most important to some rural areas),
    – CHANGE through time in the way we “furnish” our urban areas (more or less parks and gardens and more or less of offices and residential usage),
    – CHANGE through time in the way we heat our urban areas (during the last 40 years all major urban areas in Sweden has changed to area wide heating systems – more efficient and thus less leakage to the general environment )?

    Gösta Oscarsson
    Stockholm

    -

  13. BioBob, I think for purposes of “climate change” it really isn’t all that important how accurate it data are so much as you capture the proper trending of it. And when you are doing things like creating a global aggregate average and simply paying attention to the trend and not so much the actual value, you can probably get by with a small sample. The notion isn’t really to accurately measure the “average” temperature of the planet as it is to get an accurate “signal” or the variation of the data.

    Even a single station per US state could do it provided you had a location with no other influences to its trend except changes in the climate. That is what make land use changes so tricky. If you have 100 stations in a given area and over half the record suddenly a dam was built and massive irrigation started, you get a change in trend. I believe that FEWER stations would actually make it easier to spot trends but at the same time you have to be very careful to select stations with minimal surrounding changes over the years. A station in a county with about the same population it had 100 years ago with a surrounding terrain that has remained basically unchanged for 100 years would probably give a good indication of the local climate trends.

    My personal belief is that for the most part, the station records have too much locally induced noise to be of much value. That is the price that you pay for using stations near people. A station in the middle of Redden State Forest, DE or Tumbling Run Game Preserve, PA or Dixie National Forest, UT probably won’t show much human induced variation.

    Places away from airports, away from towns, away from heavy ag use are probably going to give you are more accurate climate signal. More stations just means more chances for noise and makes it harder to see any signal. Heck, a collection of stations on Islands in the various oceans might even be good enough but in general, I would want a network of isolated stations about 250 miles from each other.

  14. BioBob @ 12.22

    Touché

    And with that level of knowledge perhaps you would consider assisting Donna L and plot the distribution of lead and secondary authors, the sample they were likely to have been obtained from and then the final sample offered of authors in the IPCC summaries. Three maps?
    And then construct a sample of media outlets for disseminating that work. These days collection & dissemination is the field of market research and economics. And visuals always have great impact.

    As McKintrick stated so well in his recent report
    What is Wrong with the IPCC: Proposals for Radical Reform
    3.1 Appearance and Reality

    “Simple card tricks often work by adding in steps that seem to make the outcome impossible, but which in reality have no effect. For example, if the card that needs to be revealed is known to be at the top of the deck, it is a simple matter to shuffle the deck repeatedly without changing the position of the top card. While it looks like the entire deck has been randomised, in reality the shuffle was neutralised with respect to the one card that matters.
    The analogy applies to the IPCC process”

  15. In some parts of the westernised world there has been a pervasive shift in vegetative cover. Trees have been making a comeback in many rural areas that were previously completely cleared. This affects the local and regional albedo and local wind patterns. Relatively recent (last 30 years) changes in the nature of the land surface and the extent of irrigation need to be controlled for. Has this been carried out in the work that the post above summarises?

  16. Even if the months with extremes of heat and cold were ditched the data would still only give a false positive result. Aristotle would be dumbfounded at the continual error of climatology in arguing from the particular to the universal. A a good enough guess in science ain’t science it’s pseudoscience.

  17. Reconstructions of temperature rely on a global network of stations?

    The problem is they are NOT global and probably not fall within the definition of ‘network’.

    The alarmist reconstructions rely on a global network of scientists who are not bothered about altering, deleting or otherwise changing their data to fall in with the AGW ideas.

  18. I really need to do some proper science about UHI as i live in the perfect city for it.

    I remember as a kid doing some simple geography field trip work in a village and we could dectect a 1-2 differnce in temperature between the village and the sourrounding countryside.

    Which is almost exactly the global warming signal for the past 100 years.

    From the city centre where i live Sheffield UK, a city with 500,000 people and the 4-5 biggest in the UK i can be within a national park in 20 minutes drive and a few miles.

    Almost directly along a straight road, out into the countryside. The temperature in winter can drop at least 2-3 degrees from work in the city centre to where my folks live in the countryside.

    This is actually greater thean the global warming human induced signal.

  19. What is e(smy), it has not been described above.
    I find it interesting that they would look at Rural stations up to 500km away to compare to an urban station. Everyone knows that “climate” can change in as little as 10km. I know they “grade” them by distance, but it still seems a strange thing to do.
    What is very misleading is that there are very few Rural stations compared to Urban ones and yet spatially there is much, much more rural land area than urban. Where in their algorithms is this allowed for. Shouldn’t there be a major biasing for rural stations to match the rural area difference?
    What is the difference in Trend between all RAW Rural Vs all RAW Urban without any kind “adjusment”?

  20. The most interesting UHI study that I have read is about Barrow, AK (Hinkel, et al. 2003 and Hinkel, et al. 2007). They used enough sensors to map the temperature field around Barrow. This then allowed them to establish a scientific basis for which urban sites were most representative of the urban UHI and which rural sites were most representative of the “normal” temperature. On this basis, mean Winter UHI was about 2°C. Barrow grew from a population of about 300 in 1900 to about 4,500 in 2000. Hinkel attributed substantially all of the UHI to use of natural gas extracted from local fields. On the basis that these fields were not in production in 1900 (my guess) and that the population then was much smaller, one could estimate that the UHI in 1900 was probably close to 0°C. Thus, the mean Winter UHI of 2°C would have developed over about 100 years or ten decades. I would submit then that the Winter UHI trend bias for Barrow would be 2°C divided by 10 or about 0.2°C per decade. Since fossil fuel use in Barrow is primarily for Winter heating, I would then divide by 4 seasons to obtain an estimate of 0.05°C per decade.

    Given that Barrow would likely be classified as a “rural” station in the lower USA based on population alone, and that this study assumes that the mean UHI for “rural” stations is zero, I think one could question whether this study significantly underestimates the “rural” mean UHI trend bias. The argument could be made that using the Barrow study results as an estimate of the “rural” trend bias in the lower 48 overestimates it, because there is a larger difference between inside and outside temperatures in the Winter in Barrow than would be the case on average for the lower 48. However, this study implicitly assumes that the “rural” station UHI trend bias is zero, when it may well be significantly different from zero. ANY non-zero “rural” UHI trend bias would be in addition to the difference bias that this study estimates.

    Unlike Hinkel, this study does not seem to develop a scientific basis for representativeness of the urban or rural stations as estimators of the “urban” mean UHI or the “rural” mean UHI. Such representativenes is simply assumed. Also, the study appears to be heavy on statistical examination of the station data, while not checking for the presence of other factors that might cause differences between “urban” and “rural” station pairs. Hinkel refers to Daly 2006:

    The Barrow Urban Heat Island Study (BUHIS) network has unusually high instrument density, and was designed specifically to avoid many of the problems inherent in networks producing spatial climate data sets.

  21. Re AC Osborn +1 for how climate can change over very short distances. 20 miles from me, farmers grow wheat and oilseeds, I’d struggle to get a decent crop of barley even with its shorter growing season. (Same altitude, 100M above sea level)
    Also, in the same way as you get ‘light pollution’ whereby one can see the (usually orange) glow of city streetlights many tens of miles away from the actual city, there must surely be an equivalent for the down-welling infra-red. Can rural thermometers ‘see’ an infra red glow from urban areas that may be dozens of miles away and over the horizon – are all thermometers thus affected?
    It might just explain this entire global warming scam…..

  22. A very interesting and intriguing chart. IMO it looks pretty reasonable. A few years ago prompted by a post on CA, I looked at the population data on all South American GISS locations. Much of the GISS population data is seriously at odds with local census data in countries with rapidly growing and urbanizing populations. The only current rural site in South America that I could find appeared to be an island off the coast of Brazil and even that was growing rapidly.
    This raises the point that if you can identify sites that have shifted from rural to urban in the last 50 years you should get a pretty clear UHI signal. Is another coloured dot possible? After all it is the change in population density that is relevant not simply whether a site is rural or urban.

  23. Are there really no stations in Brazil? It is a small map, and I have bad eyes, but I am not seeing a dot, either blue or red, in Brazil.

  24. Right off the bat, I’m suspicious. The classic UHIE study of Barrow, AK, showed substantial effects even in such a remote “rural” location, due to buildings/siting. How many other “rural” sites are similarly affected? Probably most.

    FAIL.

  25. Land temps come and go (literally every night). It is the temperature directly over the oceans that are of interest to me. If they demonstrate warming and cooling commensurate with oceanic conditions, you not only have a very good temperature record that clearly cannot be affected by urbanization, you also have cause.

    Land temps are useful for weather prediction, clothing selection, and airplanes.

  26. crosspatch says:
    December 5, 2011 at 11:16 pm

    While it is beyond the scope of your research, I would also be interested in the impact of multidecadal changes in atmospheric weather patterns on UHI over long periods of time. For example, if we were to see a fairly reliable seasonal storm track make a persistent move in location, we might see a significant change in regional temperature trends.
    _____________________________

    I have one “pair” that is interesting and it is not even a rural/urban pair but a urban/airport pair:

    The city is on the North Carolina/Virgina border and right on the ocean.
    Norfolk City

    Norfolk International Airport

    Here is the raw 1856 to current Atlantic Multidecadal Oscillation Amazing how the temperature follow the Atlantic ocean oscillation as long as the weather station is not sitting at an airport isn’t it?

    Other Coastal Cities:
    North – Elisabeth City

    South – Wilmington NC

  27. BioBob says:
    December 6, 2011 at 12:22 am

    ….The corruption of the science of sampling is truly awesome all the way down to the purported precision down to hundredth of a degree..

    Neither have I ever seen an honest accounting of the actual extent / range of all the various errors in the data set.

    GIGO – the data is trash no matter how you stack it.
    ____________________________________________
    AMEN!

  28. ****
    Phil says:
    December 6, 2011 at 3:45 am

    ****

    Thanks, Phil, I hadn’t read your post before posting mine. You provide alot more detail on the Barrow study.

    Since fossil fuel use in Barrow is primarily for Winter heating, I would then divide by 4 seasons to obtain an estimate of 0.05°C per decade.

    The actual heating season is longer than 3 months in Barrow, isn’t it? :)

  29. Very interesting. A recent paper showed that in semidesert to desert environments (Southwest USA for example) the urban environments were actually cooler because of the vegetation people plant. The effects will then be affected by the distribution of rural sites, since many more of them are in undeveloped parts of the world, which tend to be drier.

  30. John Marshall says:
    December 6, 2011 at 2:08 am

    ….The alarmist reconstructions rely on a global network of scientists who are not bothered about altering, deleting or otherwise changing their data to fall in with the AGW ideas.
    __________________________
    OH, you mean like this? http://i31.tinypic.com/2149sg0.gif

  31. Gail Combs says:
    December 6, 2011 at 7:02 am
    Here is the raw 1856 to current Atlantic Multidecadal OscillationAmazing how the temperature follow the Atlantic ocean oscillation as long as the weather station is not sitting at an airport isn’t it?

    Let me guess, a growing airport and the daily onshore wind blows straight across a long section of hot tarmac right to the weather station? Wouldn’t be suprised if this is a systemic problem for all airport stations near an ocean.

  32. All this discussion has missed a basic point.

    The concern raised in the AGW hypothesis is that heat is being trapped in the Earth system leading to potentially catastrophic changes in climate. Therefore, to assess the truth of this statement the metrics used should quantify heat not atmospheric temperatures. In the atmosphere temperature is not directly related to heat content.

    To quantify the amount of heat being retained in the Earth system it is far more accurate to measure ocean heat content; or, measure incoming radiation and outgoing radiation from satellites outside the system. The metrics that correctly quantify heat content show that there does not appear to be an increase in heat being retained within the Earth system.

    The correct metrics failure to show the desired AGW result is not a reason to try to use incorrect metrics. Similarly just because historically the incorrect metrics have been recorded is not a reason to use them – that is searching for keys under the lamppost.

  33. “Summer land surface temperature of cities in the Northeast were an average of 7 °C to 9 °C (13°F to 16 °F) warmer than surrounding rural areas over a three year period, the new research shows. The complex phenomenon that drives up temperatures is called the urban heat island effect.”

    http://www.nasa.gov/topics/earth/features/heat-island-sprawl.html

    They looked at 42 cities.

    Zeke, NASA says 7C to 9C in the summer. You say .03C per decade. Does your data show 7C – 9C in the summer? If not, why not? What is wrong with your methodology? How do you plan to correct it?

  34. The obvious conclusion is that rural stations are becoming more rare and have cleaner data. They should receive more attention and preservation. And some more should be added.

  35. Pete in Cumbria UK says:
    December 6, 2011 at 3:50 am
    ….. Can rural thermometers ‘see’ an infra red glow from urban areas….
    _______________________
    NO.

    The most likely problem is micro siting problems. That is moving from a glass min max thermometer sitting out in a field some where to a digital thermometer with a cable. Cable length means the actual “Site” will be moved from the middle of the field to near the side of the building and all of a sudden your “rural” site is no longer actually “rural”

    That is what Anthony’s surface station project was all about http://www.surfacestations.org/

    list of Several discussions on the subject: http://wattsupwiththat.com/category/weather_stations/

  36. They could save even more money on cabling costs if they were to simply site the weather stations inside the buildings and be done with it. This would pretty much stabilize temperatures at 70F/20C around the globe. We would have a very brief 5C jump in average temperatures from the current 15C, probably due to elevated CO2 inside the buildings, then temperatures would stabilize globally.

  37. crosspatch says:
    December 6, 2011 at 1:12 am
    BioBob, I think for purposes of “climate change” it really isn’t all that important how accurate it data are so much as you capture the proper trending of it.

    I agree, however, this statement from the above is key “…but will not necessarily show a higher trend over time unless the environs change in such a way that the UHI signal is altered in the vicinity of the instrument.

    I am very skeptical that one could find any station under “UHI/LHI” influence hasn’t had the effect altered multiple times over the time frame that it was under that influence.

  38. .Someone please do the fact-check on these recollections I have.

    As I recall, in the mid-1980’s Weathermen were just itching to utilize the next generation of weather radar, but it was going to cost a lot, and at that time, (unlike today,) Congress was digging in its heels when it came to spending like drunken sailors. NOAA was told it couldn’t get its new radar unless it made significant cuts in other areas.

    It was for this reason the excellent system of gathering temperatures got mangled. A lot of faithful employees were replaced by gizmos that needed to be connected to the station by a cord. The cord could not cross a highway, for it was too expensive to cut trenches through the tar. Therefore a lot of thermometers were moved closer to the weather stations, both in Urban and Rural locations.

    What is interesting to contemplate is the possibility of ulterior motives being involved. After all, the mid 1980’s was when they left the windows open in Congress to make it uncomfortably hot when Hansen made his speech about Global Warming. If sneaky stuff like that could be done, why not mangle the excellent system of gathering temperatures, to make it easier for Hansen to “adjust” the temperature records?

    In the old days I would have deemed this sort of suspicion sheer paranoia on my part, and dismissed it from my thinking. However so many crazy suspicions have turned out to be true I don’t think anything would surprise me any more. And I do know that members of congress change their votes for reasons that have nothing to do with what the vote seems to be about.

    The NEXRAD system was passed by Congress in 1988, I think, and ground based temperatures started to get messed up at the same time, “to save money.” Who would have dreamed Hansen would adjust as he adjusted, and the cost-cutting measure would turn out to cost trillions?

  39. Ian W says:
    December 6, 2011 at 7:27 am

    …..To quantify the amount of heat being retained in the Earth system it is far more accurate to measure ocean heat content; or, measure incoming radiation and outgoing radiation from satellites outside the system. The metrics that correctly quantify heat content show that there does not appear to be an increase in heat being retained within the Earth system.

    The correct metrics failure to show the desired AGW result is not a reason to try to use incorrect metrics. Similarly just because historically the incorrect metrics have been recorded is not a reason to use them – that is searching for keys under the lamppost.
    ________________________________
    That has nothing to do with the “Science”

    The reason IPCC exists and climate scientists are funded is to produce “Data” that scares the crap out of Joe Sixpack so he will cough up more of his hard earned cash and fork it over to the bankers/financiers/”green” corporations and of course the politicians.

    Rising temperatures do a dandy job of this. There is no cause for alarm does not.

  40. Imho, local weatherstations were built and run to document local events, local developments.
    Very useful.

    To have thousands such stations, manipulating their data in order to compensate local developments, and add all this together to show a global mean temperature is the “work hard” approach. Not very smart. Way too many variables, error sources and local dependencies.

    Wouldn’t the smarter way be to have a selection of a few truly rural stations only, like National Parks, with a long standing record, unchanged siting and instrumentation?
    Not more than a handful per continent. Less work, more reliable results, no manipulation needed.

    These few stations would most likely give an unbiased impression about the general land surface temperature trend. Complemented with satellite data for sea surface and all is done.

  41. My question is this. Why do we include urban sites in any temperature reconstruction? When you only look at rural temperatures there is no positive trend in temperatures. The continental USA has shown a drop in temperatures over the past 100 years when you exclude urban sites.

    Oh wait, I just answered my own question. Without a scary trend there would be no grant money.

  42. So, using the lowest estimate of UHI of 0.03C/decade over the last 150 years there is a 0.45 degree C signal. Considering the estimate of 0.8 degrees warming during that time period in the US, more than half of the overall warming would be related to UHI. Now, the question to ask is how is this signal being compensated for in the US temperature data? From what I’ve seen it has either been ignored or exaggerated.

  43. Caleb says: December 6, 2011 at 7:49 am

    .Someone please do the fact-check on these recollections I have.

    I agree with you. It is difficult to ignore the simultaneous rise of temperatures c. 1980, with the change in equipment and therefore siting.

  44. JohnWho

    “I am very skeptical that one could find any station under “UHI/LHI” influence hasn’t had the effect altered multiple times over the time frame that it was under that influence.”

    If you are looking for a “climate change” signal you just want a place that hasn’t changed and there are many of them in practically every region. Little towns off the beaten path with small populations can still be found as can areas where land has been set aside as park or preserve. The problem with UHI impacts on temperatures, though, is that they give you a sunlight bias in your temperatures.

    So lets say you have a city in a place that experiences much different weather in La Nina conditions than it does in El Nino conditions. Say your city tends to experience drought in La Nina years (Amarillo?). If the PDO changes from positive to negative and we have more La Nina years than El Nino years, your temperature readings may now be biased by more sunny days per year heating up that concrete in town. So in this case you get an artificial biasing of urban temperatures because of changes in cloud cover. While it always had UHI , the degree of UHI can change depending on cloud cover. UHI would have less of an impact on a cloudy day than on a sunny day. In Phoenix with a missed “monsoon”, UHI would have more impact than in a year with a strong summer “monsoon” and lots of afternoon clouds.

    In other words, UHI is basically a sunlight bias. Changes in weather patterns that cause changes in the amount of sunshine will change the amount of UHI bias. I don’t think UHI can be precisely quantified because the amount of UHI impact on the temperatures in a location will vary depending on the amount of sunshine a place receives. I will be willing to bet there is a rough correlation between the amount of UHI offset in a location and precipitation amounts (unless it all falls at night!).

    I believe that UHI is so tricky and so fickle that it can’t be accurately removed in any reliable way. It would be best to completely eliminate it from measurements being used to track climate trends.

  45. Gail Combs says:
    December 6, 2011 at 7:50 am
    Ian W says:
    December 6, 2011 at 7:27 am

    …..To quantify the amount of heat being retained in the Earth system it is far more accurate to measure ocean heat content; or, measure incoming radiation and outgoing radiation from satellites outside the system. The metrics that correctly quantify heat content show that there does not appear to be an increase in heat being retained within the Earth system.

    The correct metrics failure to show the desired AGW result is not a reason to try to use incorrect metrics. Similarly just because historically the incorrect metrics have been recorded is not a reason to use them – that is searching for keys under the lamppost.
    ________________________________
    That has nothing to do with the “Science”

    The reason IPCC exists and climate scientists are funded is to produce “Data” that scares the crap out of Joe Sixpack so he will cough up more of his hard earned cash and fork it over to the bankers/financiers/”green” corporations and of course the politicians.

    Rising temperatures do a dandy job of this. There is no cause for alarm does not.

    I fully agree – but why are the ‘correctly skeptical scientists’ (most posters above) being drawn into complex arguments on how to quantify the incorrect metric? The approach should be – “The IPCC, HADCRUT, GISS etc., are measuring the wrong variable.” and not enter into long debates on how to more accurately measure the wrong variable.

  46. How can UHI be high enough to kill people but not so high as to register in the temperature record? That is a question left unanswered at a recent NJ conference on climate change.

    “Month-long spells of 100-degree weather will kill more people in the New York/North Jersey metropolitan area, with the greatest danger in urban “heat islands” like Newark and Jersey City.” I believe-though not certain- this is attributed to Dr. Kim Knowlton a Columbia University clinical professor of environmental health sciences.

    http://www.newjerseynewsroom.com/science-updates/global-warming-experts-paint-a-bleak-picture-of-new-jerseys-future

  47. What about issues from conversion from manual recording to automated? Sometimes people just fudge the logs when they forget to record on time, or forget to check at all. This usually happens when it’s very cold out, or when the responsible person otherwise doesn’t feel like trudging outside to make a reading.

    Is this a problem? Is it quantifiable? We know it was an issue on the old Soviet Union when remote sites would fudge their cold-weather data in order to get a larger allocation of fuel.

  48. Changes in land use within the “rural” designation may also be significant. If forests and wild grasslands are put under the plow to feed the growing world population (and fuel for cars), then the “rural” designation is confounded, and a true mixed urbanization/cultivation of wild lands bias on averaged global temps may be present.

  49. but why are the ‘correctly skeptical scientists’ (most posters above) being drawn into complex arguments on how to quantify the incorrect metric?

    I fully agree with that. There is really no need to spend the time on the issue at all. You can look at the temperature rise from the 1970’s to around 2000, then look back in the record and see a nearly identical pattern of warming from around 1910 to the late 1930’s and say “oh, it’s doing the same thing it did 60 years ago” another step up in the recovery from the LIA. We are probably due for one more such step starting in 2040 or so and I have no indication that it is unnatural in any way at all. We are probably going to go about 30 years or so with relatively flat temperatures.

    In this link that was provided in another thread here recently:

    http://www.woodfortrees.org/plot/hadcrut3gl/from:1998/plot/hadsst2gl/from:1998/plot/hadcrut3gl/from:1998/trend/plot/hadsst2gl/from:1998/trend/plot/hadcrut3gl/from:1980/to:1998/plot/hadcrut3gl/from:1980/to:1998/trend/plot/hadcrut3gl/from:1934/to:1980/plot/hadcrut3gl/from:1934/to:1980/trend/plot/hadcrut3gl/from:1905/to:1934/plot/hadcrut3gl/from:1905/to:1934/trend/plot/hadcrut3gl/from:1880/to:1905/plot/hadcrut3gl/from:1880/to:1905/trend/plot/hadcrut3gl/from/to:1880/plot/hadcrut3gl/from/to:1880/trend

    We see a rise in temperatures from 1850 to 1880. About a 30 year rising trend in temperatures. We weren’t emitting a lot of CO2 in 1850 to 1880. Then we have another rise from 1905 to 1934, another 30 year uptrend in temperatures. We were not emitting a lot of CO2. This second rise is steeper and longer than the earlier rise. Then we show another uptrend in the graph starting in 1980 but I would change the timings of some of the trends to show where they really switch and not place them on even 5yr boundaries:

    http://www.woodfortrees.org/plot/hadcrut3gl/from:2004/plot/hadcrut3gl/from:2004/trend/plot/hadcrut3gl/from:1975/to:2004/plot/hadcrut3gl/from:1975/to:2004/trend/plot/hadcrut3gl/from:1942/to:1975/plot/hadcrut3gl/from:1942/to:1975/trend/plot/hadcrut3gl/from:1911/to:1942/plot/hadcrut3gl/from:1911/to:1942/trend/plot/hadcrut3gl/from:1879/to:1911/plot/hadcrut3gl/from:1879/to:1911/trend/plot/hadcrut3gl/from/to:1879/plot/hadcrut3gl/from/to:1879/trend

    The “1945 problem” is clearly seen here and that is due to a change in the mix of sea surface temperatures being reported during WWII with US ships doing it one way and UK ships a different. After 1945 the number of US reports greatly declined and so you see a sudden drop in the graph that is probably not indicative of actual temperatures. In HadCRUT3, locations near the coast a GREATLY biased by any SST measurements nearby. So a change in SST measurements can also change what HadCRUT3 reports for grids that are even 90% land.

    What that second graph I posted shows is that we catch the tail end of a long term rising trend ending in 1879-1880. Then temperatures decline to about 1910 or so. Then temperatures start another sustained rising trend until about 1940-ish. Then we probably had a decline/flat period until about 1975 or so when temperatures continued their climb in a fashion very similar to the 1910-1940 increase and since 2004 temperatures have begun to decline but 6 years of data isn’t enough to say with any certainty though if the pattern of the past holds going forward, we should expect about 30-ish years of flat to cooling temperatures.

    The trend from 1975-ish to 2000-ish is nearly identical in duration and amplitude as that from 1910-ish to 1940-ish. What this graph captures is a three-step recovery from LIA temperatures. There is nothing in the more recent rise that looks “alarming” or scary or human created or needs to be “mitigated” in any way and doing so is simply throwing good money after bad. What we have is are NATURAL CYCLICAL periods of warming as we recover out of the Little Ice Age temperature regime. If we are to completely recover to what temperatures were before the LIA, I would expect to see another step up in warming starting in around 2030 or so.

    There is no reason to waste a lot of time and resources on this issue. We have seen warming like this in the past, in the very RECENT past, in fact (early 20th century). There is nothing in the temperature record that shows anything unusual is going on. No abnormally high rate of change or amplitude of change has happened. In short, nothing to see here, move along!

  50. To add to crosspatch’s charts, this chart has a CO2 overlay: more evidence that natural variability has a much larger effect than CO2. In fact, the effect of CO2 is so small it is down in the noise, and can’t be separated from natural variability.

    Looking at the long term trend, there is nothing unusual happening. Nothing.

  51. Also note that so far the slope of decline since 2004 is very similar (looks nearly identical) to the slope of temperature decline between 1880-ish and 1910-ish. Yet another indication that there is really nothing new going on and that we are likely seeing natural variation.

  52. crosspatch says:
    December 6, 2011 at 9:10 am
    I believe that UHI is so tricky and so fickle that it can’t be accurately removed in any reliable way. It would be best to completely eliminate it from measurements being used to track climate trends.

    I agree completely with that.

    Something wrong with a system that puts a calibrated station in a distant area and then someone sits many miles away and determines a methodology to adjust and manipulate the station’s data in order to get a “corrected” reading.

  53. crosspatch says:
    December 5, 2011 at 11:46 pm
    “I wondered about that, too, Alexander, because I know there is some pretty rural area in Kansas and Nebraska and the Dakotas that show up as “urban” in that map. As in most of those states are rural except for a few smallish cities. Heck, some places out there you can go 40 miles between tiny little towns.”

    Are you kidding? Kansas has road signs that say “No gasoline for the next 80 miles.”

  54. > crosspatch says: December 6, 2011 at 1:12 am
    > it really isn’t all that important how accurate it data are so much as you capture the proper
    > trending of it. And when you are doing things like creating a global aggregate average and
    > simply paying attention to the trend and not so much the actual value, you can probably get by > with a small sample.
    ————————————————

    How would you know what is adequate/accurate with N = 1 ? Sorry Crosspatch but you need remedial sampling theory 101 and stats 101, as do most climate scientists by in large. The point is that aggregating values is one of the ABSURD statistical processes involved here.

    Not only is the sample size (1) totally inadequate, but stats are routinely employed which REQUIRE that the population the sample is drawn from be NORMALLY distributed and that the sample drawn from that population is statistically valid. The number of untested assumptions about the temperature population the single sample is drawn from, the assumptions about the methods used, and the assumptions about the statistics employed to describe that sample are enormous and in most cases in error. You do not need to be a climatologist to conclude that, thanks very much, since that field of investigation more resembles divination than science.

    This is most assuredly a case of trying to make a silk purse from a sows ear simply because some people have an axe to grind [lol]. And damn the actual science/stats, or the data processing, as we have seen.

    It is WAY PAST TIME for scientists to do the science CORRECTLY with PROPER random sampling and production of PROPER descriptive statistics with a minimum of absurd assumptions. So, most of all, let’s see the use of proper sampling and statistics with reasonably accurate standard deviations and identification of all sources of error.

    Proper sampling and use of stats won’t happen simply because the variance and errors are monstrously large and expose the fact that we can not measure any global temperature change because any change is lost in the “noise” in human timeframes and we are pissing into the wind in the attempt.

    My take ? Let’s get back to investigating problems and processes that we actually CAN find the answers to rather than straining at describing thousand year behavior of gnat fart’s in chaotic global heat engines that we don’t even fully understand and let some reality penetrate thru this unproductive brain fug.

  55. Smokey: your first link is USHCN which I would like to point out to folks is a CONUS land station record only. HadCRUT3 includes sea surface temperatures. Things such as La Nina can have a greater impact on HadCRUT3 than on USHCN. Just wanted to point that out to folks. USHCN would also, by the way, be unaffected by changes in the way sea surface temperatures were measured by ships and the replacement of ship data by satellite and later buoy/float data. Also note that the CET temperatures don’t pick up the early 20th century warming, it seems to happen much later there (1950s?) which might be due to a major North Atlantic influence on temperatures there.

  56. Are you kidding? Kansas has road signs that say “No gasoline for the next 80 miles.”

    So does California actually! You might have to go through a town to get to the next one with the gas station. Little hint, though, many of those farms in Kansas (and elsewhere) have their own gas pumps on the farm. If you are really in a jam out that way and come to a decent sized farming operation, you can sometimes persuade them to sell you a spot of fuel just to get you to town.

  57. To answer a question raised above, there may be a hundred deaths from heat in New York city, but it’s cold that consistently kills: deaths per month.

    Next, who decides which stations are removed? The average temperature compared to the number of stations also raises serious concerns. Who decides if the stations eliminated are urban or rural?

    And as Gail Combs points out, the AMO appears to have a huge influence on temperatures compared to the unmeasurably small impact of CO2.

    Naturally the warmist crowd is going to say the number of stations is irrelevant. But there seems to be a concerted effort by government agencies to eliminate most of the temperature recording stations. Why? And again, who decides which particular stations are removed?

    With so much money at stake, added to the documented fact that GISS, USHCN and other agencies manipulate the data to show either lower past temperatures [creating an alarming-looking rise] or artificially “adjust” current temperatures [always up, never down], the only reasonable conclusion is that satellite temperatures are the only reliable data.

  58. crosspatch,

    I was showing the non-correlation between CO2 and temperature. I don’t have a global temp/CO2 chart handy, but global temps mirror U.S. temps pretty closely, and CO2 still seems to be a non-playa.

    And before the hand-waving over tiny fractions of a degree gets to gale force, sometimes it’s best to just sit back and look at the situation in human terms, by changes in degrees. At times during the Holocene temperatures have changed by as much as 15°C in a very short time. The current global temperature has been unusually mild and pleasant for a long time now. A degree or two warmer would be even better. I guess that makes me a genuine warmist.

  59. Here is another interesting piece of data: The Blue Hill weather observatory has the longest continuous temperature record in the US:

    You will see the trend in temperature rise from 1870 to about 1950 is rather steady and then FLATTENS from that time to the present. At that observatory, the rate of overall temperature rise since the 19th century has declined in the late 20th century.

    And about this comment: “Sorry Crosspatch but you need remedial sampling theory 101 and stats 101″

    I don’t think so because I am not trying really to get an accurate measurement of temperature. I am trying to get a sense if it is rising or falling and at what rate. For example, if I have a series of balls coming off an assembly line, I don’t need an extremely precise measurement of the weight to know if they are generally trending heavier or lighter as time goes by. As long as I am measuring with the same instrument at the same point in the production line, the calibration of my instrument doesn’t matter as long as it isn’t drifting. I have a perfect sample. I have 100% of the temperature measurements (or very close to 100%) that were taken at a given station. I only want to see if those measurements are trending generally higher or lower.

    Lets say I have 50 machines that make these little balls and each machine makes them just a little bit differently. I suspect the material the balls are being made of is more dense than it used to be but I’m not sure because there is natural variation of the density from batch to batch of raw material. Over that period of time, some machines have been in and out of service, some are completely gone, other new ones have been added and yet others have had their measuring instrument changed. Even worse, some of these stations have instruments that change their measurement in response to ambient temperature at different rates than the others do but I do have records of the density of each ball that came out from each of the 50 machines.

    If I try to get too “perfect” I am only going to add additional uncertainty. The more samples I add from more different places, the more uncertainty I add to my result. I can identify the instruments that have the least amount of change according to ambient temperature so that reduces the amount of measurement uncertainty (I will use that as an analog to UHI). Then of those, I select the ones that have the longest running continuous data set. I don’t need to measure ALL of them, I simply need to measure a few in each different one of my production plants (say I have 100 production plants around the world all using the same raw material which I suspect might be changing in density). If the density of the material going in (an analog of general global climate) is increasing, I should see a general increase in density of balls at all plants. So if each plant has 50 machines, I can probably get away with trending the output of three of these stable instruments at each of my 50 plants. I do not care what the total average density of all of the balls I produce is so I don’t need an accurate sample because that is not what I am trying to quantify. I am only looking only for a trend.

    So I look at my production for the past 20 years from three machines at each plant that have been in continuous operation with instruments whose ambient conditions haven’t changed and whose measurements themselves don’t change much with any change in ambient conditions. I notice that at 49 of the plants the density has trended up and at one plant it has trended down because, as it turns out, the plant operator was using a little less material per ball so he could skim some off to make his own balls for sale on the local black market without me noticing any reduction in production.

    The problem with UHI is that it can cause different trends in the same place with the same thermometers. I can have 30 years of cool PDO which gives me one weather pattern at a location and so influences the measurement in some way and then have PDO switch to its warm phase and get a completely different trend. UHI is, for the most part, a sunshine proxy. Combined with land use changes, it can be both a sunshine AND a humidity proxy combined.

    For example, the temperature profile in Phoenix in summer given exactly the same weather conditions was probably much different in 1900 than it was in 2000. This is because I now have a bunch of concrete and asphalt absorbing a lot of solar energy. Then at night everyone’s sprinklers turn on and cranks up the humidity. So the conditions at both midnight and dawn and now completely different. Now if I go into a prolonged drought and they tell people they aren’t allowed to water their lawns, suddenly UHI takes on a completely different profile! Same if I get a persistent “monsoon” with an unusually cloudy summer. So UHI can not really be quantified because its impacts can be completely different this year than it was last year.

    Measuring climate with urban data is stupid because too many things change. It is a noisy environment with too many things influencing measurements that are not related to climate.

    I believe I could get a better handle on the general climate of the earth’s surface with regularly spaced rural stations than I can with a bazillion mixed urban/rural stations. All those additional stations do is add noise.

  60. Esh., I changed from 100 to 50 plants in that mental cartoon, sorry about that. make that 99 trending denser and 1 plant operating skimming from me )

  61. I would be interested in seeing whether the results depend on the exact classification rule for “rural”.

    Also, will the data used and the code be made public?

  62. I was showing the non-correlation between CO2 and temperature. I don’t have a global temp/CO2 chart handy, but global temps mirror U.S. temps pretty closely, and CO2 still seems to be a non-playa.

    Yes, that graphic points out exactly the point I was trying to make … nothing new is going on. We don’t see any unusual increase in temperatures let alone a change we can attribute to humans. IT JUST ISN’T HAPPENING.

    Also, on the word playa: Living in the Western US that word has a different meaning. It is pronounced PLAH – yah and is Spanish for beach or shoreline and is used to describe dry lake beds. So for example we would describe Groom Lake as a playa. There are a lot of playas in the Great Basin region.

  63. To whom is this article addressed? Statisticians? Possibly. Mathematicians interested in your use of linear equations? Possibly. Persons interested in the facts on the ground. No. Nada. Nein. Nyet. If this is the kind of study that you are going to do, I recommend that you submit it to Real Climate.

    When will you realize that you will have achieved nothing until you address the facts on the ground? You are encouraged to discuss your statistics and other mathematical techniques but only for the purpose of explicating to the non-mathematician how they do justice to the facts on the ground. What is at issue is whether your statistical work has any foundation in empirical matters. We are very worried about this because we know that in other equally important matters you are completely oblivious to the facts on the ground.

    For example, in the matter of tree ring proxies for temperature, you insist that historical records of tree rings can be used for proxy measurement of temperature yet you know very well that those historical records float free of the facts on the ground because no one has ever done the empirical work necessary to explain what environmental factors cause the changes in tree ring width.

    Finally, everyone knows that if you use only the US stations that have no interruptions in their records then there is no warming trend at all. Could you please explain the facts on the ground that caused these stations to behave in this way. Consider this an introductory assignment that will get you prepared for the larger assignment.

  64. If only somebody would start a project to survey all the sites and quantify the issues with their siting, such as being built on top of a parking garage, or under and air conditioning vent!

  65. Everyone knows (and it has been stated above) urban areas have warmer temperatures than rural areas, and are greatly over represented in virtually every country. However, the powers to be insist that if the population of the urban area has not greatly increased, and the general nature of the urban area has not greatly changed, the temperature trend (i.e. rate of increase of temperatures) in urban areas should be the same as in rural areas (and thus representative of the country in general). This is absolutely untrue because it totally ignores the change in lifestyle in virtually all first world cities. Consider air-conditioning (in the summer in northern cities and year round in southern cities). Anyone who has stood outside near an air-conditioning vent can attest to the rush of hot air being introduced to the local environment as warm air inside the building is being expunged, and replaced by cool air. Fifty years ago, only wealthy people in northern cities had air-conditioning in their homes. (I can remember people who would rush to movie theatres on a hot summer’s evening – not because they cared about the movie being shown, but because the theatre was advertised as being air-conditioned!) In Europe, as recently as twenty years ago, air-conditioning was an outrageous extravagance that even the rich didn’t have. Today, while not as common as in North America, air-conditioning in Europe is not uncommon. The huge migration from the north to the south in the United States over the last forty years or so was made possible by the fact that in urban areas of states like Florida and Texas (where the summers are quite uncomfortably hot), virtually every indoor place you go to has air-conditioning, and the air-conditioning is set to temperatures that don’t just make it bearable, but make it outright cool! Lifestyle in the wintertime brings another unaccounted factor. Fifty years ago, many homes in the northern cities did not have central heating. In southern cities and in Europe, almost no homes did. On cool nights, you wore flannel pajamas and put extra blankets over you. On very cold days, you might try to heat up the one room you were in, while leaving the rest of the house with relatively little heat. Today, many people forget about winter pajamas and extra blankets, and just turn the thermostat up to keep the whole house nice and warm. This of course, also keeps the environment close to the house relatively nice and warm too. Thus, just the increased use of heating and airconditioning have caused a quite measurable increase in urban temperature trends that is ignored by most (if not all) climate scientists.

  66. I am still a little puzzled about how Spain is more urban than the UK….but no doubt there is method in the seeminjg madness

  67. Nice work guys! I hope it is just a start. The magnitide of difference sounds reasonable, but for individual stations where does it sit along the timescale? Does the difference kick in with the advent of aircon from the 1950s? How does it manifest itself in relation to multidecadal cycles such as the AMO? Does it mean a reduced cooling from the 1950s and/or then an increased warming when the cycles turn? Lots more interesting stuff to tease out.

  68. Jeff S. says:
    December 6, 2011 at 1:53 pm

    Yes, what Jeff said.

    More specifically, what are the physical conditions that constitute an urban area? What causes that increase temperature measurements operate in urban areas? What are the effects of those causes? What is the process of migration as those causes move from urban areas to rural areas; that is, how does urbanization occur.

    If I visited CERN and posed similar questions about particle collisions, they would have highly detailed answers to each. Why cannot climate scientists pony up and deal with the facts on the ground?

  69. crosspatch says:
    December 6, 2011 at 11:33 am

    Thanks for the info. But I grew up on a working farm. I know about the resourcefulness of farmers.

  70. crosspatch, I was [apparently unsuccessfully] trying to be funny, or at least flippant, with “playa.” In ghetto talk it’s ‘player’. Dang, I should have used a smiley. ☹

  71. Zeke, Steve, and all, first, my congratulations a very nice piece of work. In particular I appreciated the sensitivity analysis regarding the urban/rural cutoff. It was also good to see some comparisons of various ways to estimate the “urbanity” of a given station.

    You seem to suggest that a likely figure is on the order of a third of a degree per century. If so, this is on the order of half of the warming of the 20th century. Accordingly, your analysis is in line with that of Ross McKitrick, who from memory found the same thing (that about half of the 20th century warming was due to industrialization / urbanization).

    One unrealized aspect of the increase in temperature in the second half of the 20th century is that it was also the time of a doubling of the global population. The resulting urbanization and surface modifications of the areas around the stations, purely from nothing but the sheer increase in populations, is reflected in your figures.

    All the best,

    w.

  72. BioBob says:
    December 6, 2011 at 11:28 am

    …..Not only is the sample size (1) totally inadequate, but stats are routinely employed which REQUIRE that the population the sample is drawn from be NORMALLY distributed and that the sample drawn from that population is statistically valid. The number of untested assumptions about the temperature population the single sample is drawn from, the assumptions about the methods used, and the assumptions about the statistics employed to describe that sample are enormous and in most cases in error. You do not need to be a climatologist to conclude that, thanks very much, since that field of investigation more resembles divination than science…….

    _____________________________________________________

    Agreed. It never really passed the “sniff test” Data good to 0.1C for the average temperature of the entire world for a century??? give me a break.

    You might like to take a look at this temperature study where at least an attempt has been made to do real science.

    …Jonathan Lowe, an Australian statistician, has performed extensive analysis of weather data recorded at fixed times by Australia’s Bureau of Meteorology (BoM). This analysis is available at his blog, A Gust of Hot Air. The data comes from 21 weather stations manned by professional meteorologists….. http://bishophill.squarespace.com/blog/2011/11/4/australian-temperatures.html

    A Gust of Hot Air: http://gustofhotair.blogspot.com/2009/04/analysis-of-australian-temperature-part_16.html

  73. crosspatch says:
    December 6, 2011 at 12:25 pm

    ….I believe I could get a better handle on the general climate of the earth’s surface with regularly spaced rural stations than I can with a bazillion mixed urban/rural stations. All those additional stations do is add noise.
    ____________________________________

    Frank Lancer has done just that. But instead of sausage making à la CRU he is looking at what story the data actually tells us.

    http://joannenova.com.au/2011/10/messages-from-the-global-raw-rural-data-warnings-gotchas-and-tree-ring-divergence-explained/

  74. The fundamental problem with determinations of UHI from station records is the dearth of baseline “rural” records in much of the world that extend over the ENTIRE 20th century. It is over time-frames of a century or longer that most urban footprints have changed. And it is particularly in the 20th century that experienced the introduction of heat-producing technonolgy . Every urban station record thus has a fairly unique time-history of deviating away from the temperatures in the surrounding countryside. Analysis of definitional metadata criteria for urbanization scarcely addresses the real problem.

  75. The ‘weekend effect’ is strong evidence for an Urban Cool Island effect.

    You can read more about it at the link below, but in summary weekdays are warmer than weekend days. This is almost certainly due to higher aerosol levels causing increased clouds/less solar insolation on weekdays from factories operating and more vehicles on the road.

    The weekend effect is relatively large, as much as 0.5C.

    In the developed world aerosol levels have been declining for decades, and this will have resulted in a decreasing weekend effect and IMO is the main source of an increasing UHI trend. Although aerosols are transported by wind and will affect nearby rural locations. Which makes me think the weighted for proximity methodology used in the study above will artificially depress any UHI warming trend and a weighted for distance methodology would give a more accurate picture.

    http://www.appinsys.com/globalwarming/DTR.htm

  76. Willis,

    Our analysis showed that urban stations are on average warming about 10% faster than rural stations. This would result in a bias smaller than 10% in practice, as global land record are composed of both urban and rural stations. Given that the land warming over the last 30 years was around 0.27 C per decade, an urban-rural difference of 0.04 or so would not amount to a third of modern warming per se (closer to a tenth or so, though more work needs to be done on controlling for spatial coverage and testing how much UHI is removed by pairwise homogenization). Extending this analysis back to the full century is rather difficult, as in most cases we only have a current snapshot of urbanity.

    By the way, you can find a slightly better version of the station location map that randomizes the visual ordering of urban and rural stations (rather than always showing urban stations on top of rural stations) here: http://i81.photobucket.com/albums/j237/hausfath/URPairs5.png . Unfortunately it was completed shortly after we had to submit the poster.

  77. Bruce,

    I fail to see how a relatively modest difference in trends is inconsistent with a large absolute difference unless the instrument started out in a pristine location and was subsequently urbanized, a case that is the exception rather than the rule at least over the past 30 year timeframe. We’ve been over this before, however, and may have to agree to disagree.

    We cover this in the poster when we remark that: “an instrument originally installed in a city frequently will have warmer absolute temperatures than one in a nearby rural area (especially at night), but will not necessarily show a higher trend over time unless the environs change in such a way that the UHI signal is altered in the vicinity of the instrument.”

  78. “Why do I, in these studies, not see more of formulations like
    – CHANGE through time in the DEGREE of urbanization/ruralness (probably most important to some rural areas),
    – CHANGE through time in the way we “furnish” our urban areas (more or less parks and gardens and more or less of offices and residential usage),
    – CHANGE through time in the way we heat our urban areas (during the last 40 years all major urban areas in Sweden has changed to area wide heating systems – more efficient and thus less leakage to the general environment )?

    Gösta Oscarsson
    Stockholm

    ############################################

    Funny you should ask as I am working on that problem this very minute.

    First you need to understand that the way I select rural stations leads to stations that
    have undergone little change. Simply, I select rural stations that have no BUILT pixels
    within 11km ( And did sensitivities on this paramater from 5km to 20km). I further restrict
    rural stations by eliminating all near airports, even rural airports and heliports. I further restrict
    these stations by eliminating any that have significant nightlights and impervious surface.
    This eliminates sites that have small infrastructure that may be missed by Modis classifications.
    So there is no change in economic activity at these sites because there is no economic activity
    to speak of. What you see when you look at this all rural dataset is a trend from 1979-2010
    that is approximately 0.1C less than the trend you see from urban stations.
    Looking at these rural stations further I can say that they also do not see much change
    in the percentage of cropland, and grassland. The small population that lives in the vincinity
    ( 5 people per sq km and in most cases ZERO people) does not grow over the period
    we look at. Some areas see negative growth and some areas see small positive growth
    ( say from 5 to 6 people per sq km) With no changing population the motivation for other changes
    is diminished.

    Next, if we look at the rural stations and try to “explain” the trends via regression you will
    find that the trend is not explained by population, not explained by land use or land use changes.
    The only two significant regressors are Latitude ( polar amplification) and the distance from Coast.

    So, I basically selected my rural stations in such a way that Minimized the changes that everybody worries about.

  79. Willis

    “You seem to suggest that a likely figure is on the order of a third of a degree per century. If so, this is on the order of half of the warming of the 20th century. Accordingly, your analysis is in line with that of Ross McKitrick, who from memory found the same thing (that about half of the 20th century warming was due to industrialization / urbanization).”

    No. Both we and Ross looked at the 1979-2010 period ( I think he looked at 2000)
    Ross’s work suggests that .1C per decade may be associated with changes in economic
    activity. We find that there is a UHI bias of approx .03C per decade IN THE PEROID
    of 1979-2010. You cannot extrapolate that to the century.

  80. “Nice work guys! I hope it is just a start. The magnitide of difference sounds reasonable, but for individual stations where does it sit along the timescale? Does the difference kick in with the advent of aircon from the 1950s? How does it manifest itself in relation to multidecadal cycles such as the AMO? Does it mean a reduced cooling from the 1950s and/or then an increased warming when the cycles turn? Lots more interesting stuff to tease out.”

    The bias effect is slightly lower if you move the start date back to 1950.

    As time permits I will finish the work that “explains” what factors drive the warming at
    Urban stations. It is a combination of Latitude, Distance from Coast, The AREA in sq
    km of land that is BUILT and a couple other land use factors.

  81. Theo

    ‘Yes, what Jeff said.

    More specifically, what are the physical conditions that constitute an urban area? What causes that increase temperature measurements operate in urban areas? What are the effects of those causes? What is the process of migration as those causes move from urban areas to rural areas; that is, how does urbanization occur.”

    ########################

    In this study We defined Urban by using a concatention of features to identify a station as Urban

    1. A rural station can have no significant human built area within 11km. or 380 sqkm.
    2. a rural station cannot be at an airport
    3. a rural station must not be lit
    4. a rural station must have virtually no impervious surfaces.

    Urban by my definition is anything that violates any of those rules,
    So, in 380sq km around the site if there is 1 sq km of human built surface. Its urban
    Lights over 30: urban: airport? urban.

    Increases in temperature in urban areas is directly related to the sq area of the urban landform.
    That is, changes to the surface properties ( cities have lower albedo) changes to the material
    properties ( heat capacity) and changes t o the vertical profile ( building height) are the key
    drivers for UHI.

    In fact, I can show how the Trend at Urban sites is a direct function of the amount of built area.

  82. “kwinterkorn says:
    December 6, 2011 at 10:39 am
    Changes in land use within the “rural” designation may also be significant. If forests and wild grasslands are put under the plow to feed the growing world population (and fuel for cars), then the “rural” designation is confounded, and a true mixed urbanization/cultivation of wild lands bias on averaged global temps may be present.”

    Sorry, I got that covered as well. The rural sites are selected by looking at the 380sq km
    around the site. We determine that there is no “built” areas within that zone. We can also look
    at the land use in those same areas ( you will see some of that in the presentation) So, I can
    tell if there has been changes to amount of cropland and amount of grassland in these rural
    areas. In general there are slight changes in these variables and those slight changes do not
    explain the trends seen. Conversely, the trends seen at the urban stations ARE a function
    of the amount of urban area.

  83. @ beng says on December 6, 2011 at 7:11 am

    The actual heating season is longer than 3 months in Barrow, isn’t it? :)

    Technically, you are right, but I tried, perhaps inartfully, to insert a caveat that Barrow’s UHI bias may not be comparable to the UHI bias in “rural” stations in the lower 48. My intent was to raise the question of whether the “rural” stations were, in fact, controls for measuring UHI bias in “urban areas” (i.e. that the trend bias of the rural stations did not depart significantly from zero).

    sky (December 6, 2011 at 7:56 pm ) probably phrased the issue better than I did.

  84. “Alexander says:
    December 5, 2011 at 11:32 pm
    Well, the first thing I want to say is that what a biased map you show up at the very beginning! How many rural sites are partially (or even fully) obscured by urban sites—particularly in the US? This is not really better than the way Mann et al. enjoy putting a big fat line on top of their reconstructions to hide any recent insensitivity. I’m very sensitive to people who try to lie (or mislead) with graphs.”

    ##############

    sorry about that, in the end, Nick came up with a method of plotting that alternated the plotting of rural and urban so you could see more clearly, but it didnt make the deadline. You can blame me for that graphic.

  85. @ steven mosher says on December 6, 2011 at 11:32 pm

    …We find that there is a UHI bias of approx .03C per decade IN THE PEROID of 1979-2010. You cannot extrapolate that to the century.

    I would like to clarify/withdraw my comments as I was thinking of UHI on a century scale, so what I said is not directly applicable to your commendable effort.

  86. Nick

    “One thing to consider is that for a lot of these sites, there will also be wind data. You could also extract wind data from the weather maps at the time on a large scale.

    So now you could test a site that is to the east of an urban area. You also need a site to the west of an urban area. Depending on the wind direction you either get no difference, or a bias one way or the other.”

    ##################

    the wind data would be more interesting to explain why there is no UHI in the city. A 7m/sec wind will eliminate most UHI. The more interesting effect would be changes to precipitation downwind.

  87. “As I read it, this paper finds average UHI in USA to be 0.06 to 0.1C per decade with the number of genuine rural stations being “vanishingly small”. ”

    Ah no, its global not the USA. The pairs study end up with pairs in USA and Aus.

    ‘However, as it is not “reasonable” to have such biased spatial coverage, the definition of “urban” has been eased off so that the category covers only those stations which converge on 0.2 to 0.4C.

    Is this a case of changing the data to fit the hypothesis?”

    #####################

    there is no “changing” of the data.

    The baseline case is a definition of rural that defines Rural by the following rules
    No “built” pixels within 11km of the site. 11km was selected to account for possible
    errors in station location. Station locations, in some cases, are reported to accuracy
    of 1/10th of a degree. That is 11.112 km at the equator. This radius is used
    to evaluate all sites. If there are no built pixels within that radius it is designated as
    Rural, otherwise it is Urban. In addition there is additional screening on the rural
    stations: They cannot be at airports. Since a Modis pixel is 500m on a side
    there is a small chance that it will not see a runway. or the Runway can be grass.
    So I removed the small number of rural airports that slipped by Modis. Then,
    there are some unique cases where nightlights are used to further perfect the selection
    of rural. After this we have over 5000 rural sites.

    Now the selection of 11km may look arbitrary ( BEST used .1 degree ) So we varied that
    parameter and looked at zones from 5 km to 20km. Also, the decision to allow
    NO built pixels may be too stringent, so we varaied that paramater as well to create a
    a 3D surface.

    By stretching the radius out to 20km you can start to see bias that goes above .06C
    However, the problem with this is that the spatial coverage goes below 50% of the land surface
    This is problematic because Latitude, more than ANYTHING ELSE, drives the trend.
    That is, polar amplification is real. As you select higher latitude stations you are selecting
    stations with higher trends

    Want to know the stations with the HIGHEST TRENDS?

    Tundra.

    So, if you change the radius to 20km you simultaneously drive the selection of stations
    poleward. the higher the latitude, the higher the trend. polar amplification.

    Anyway, all the sensitivity studies are there. If you are trying the measure the bias “in the record” then you need to select a rual sample that has good spatial coverage. Less than 50%
    of the land is not good spatial coverage.

  88. SirCharge says:
    December 6, 2011 at 8:04 am
    So, using the lowest estimate of UHI of 0.03C/decade over the last 150 years there is a 0.45 degree C signal. Considering the estimate of 0.8 degrees warming during that time period in the US, more than half of the overall warming would be related to UHI. Now, the question to ask is how is this signal being compensated for in the US temperature data? From what I’ve seen it has either been ignored or exaggerated.

    ##################

    you cannot extrapolate beyond the time period we examine.

    at some point one could try to take this back in time but other measures of urbanity
    would be required.

  89. A. C. Osborn says:
    December 6, 2011 at 2:16 am
    What is e(smy), it has not been described above.
    I find it interesting that they would look at Rural stations up to 500km away to compare to an urban station. Everyone knows that “climate” can change in as little as 10km. I know they “grade” them by distance, but it still seems a strange thing to do.
    ###############################################

    You should not mistake the climate for the weather. climate is a long term average. In this
    case we are calculating a trend for all rural stations within a 500km radius. Thats a 32 year
    trend. Given that stations are correlated out to 1000km, 500km was a conservative choice.
    However, its a simple matter to change that to 100km or whatever. All that does is reduce
    the number of rural stations you have contributing to the estimate of the rural trend for that zone.

    #####################
    “What is very misleading is that there are very few Rural stations compared to Urban ones and yet spatially there is much, much more rural land area than urban. Where in their algorithms is this allowed for. Shouldn’t there be a major biasing for rural stations to match the rural area difference?”

    Dont confuse the pairs study with the global study. In the global study we use 5000+ rural
    stations. the estimate of the all rural network is about 10% lower than the estimate created
    using all stations.
    ##############
    What is the difference in Trend between all RAW Rural Vs all RAW Urban without any kind “adjusment”?

    .03-.04C per decade over the period of 1979 to 2010.

  90. Septic Matthew says:
    December 6, 2011 at 12:30 pm
    I would be interested in seeing whether the results depend on the exact classification rule for “rural”.

    ################

    see the charts. we varied the definition in both dimensions. the radius from the site and the amount of built land .

  91. JJ says:
    December 5, 2011 at 10:57 pm
    The upper limit on UHI bias between rural and urban stations is on the order of 0.06 to 0.1C per decade. However, these cases are clearly problematic from the spatial coverage aspect, as the number of rural stations becomes vanishingly small when the most stringent filters are applied.

    This is not a justification for using less stringent rurality criteria. That the right answer is difficult to achieve does not make the wrong answer more correct.

    ##########################

    the issue is that as you lose spatial coverage ( <50%) you tend to migrate northward.
    That means your rural stations will come from higher latitudes. Higher latitude have
    higher trends, due to polar amplification, so that you are confounding the UHI bias
    estimate by introducing a latitude bias in your rural selection.

  92. crosspatch

    ‘Places away from airports, away from towns, away from heavy ag use are probably going to give you are more accurate climate signal. More stations just means more chances for noise and makes it harder to see any signal. Heck, a collection of stations on Islands in the various oceans might even be good enough but in general, I would want a network of isolated stations about 250 miles from each other.”

    yes, however you have to be careful with your geographic selection.
    Coast and Island trends are markedly lower than inland trends, even when the coastal
    station is urban. Its dominated by the SST trend. The network would have to represetative
    in Latitude and not be overly heavy in coastal stations.

    At some point I’ll work up what that sampling has to look like based on the percentage of land
    by latitude band and percentage of coast by latitude band. But a smaller more representaive
    sample is way better than thousands of stations 10s of km apart.. at least for climate.

  93. Phil
    “Given that Barrow would likely be classified as a “rural” station in the lower USA based on population alone, and that this study assumes that the mean UHI for “rural” stations is zero, I think one could question whether this study significantly underestimates the “rural” mean UHI trend bias.”

    Well, I love the Barrow study. The nice thing is that I used my rural proxy to identify rural stations.
    Then I checked it against some rather famous cases that I love, like barrow. The barrow study
    used a rural network. So I can actually test whether my method says they are rural.

    One of the rural stations used in the Barrow study has been turned into a CRN station.
    I classify Barrow proper ( the city) as URBAN. I do this on the basis of 500 meter satillite data.
    Population has nothing to do with how I classify sites. I classify the surrounding CRN station at Barrow
    as Rural. that is, the SAME station used in the UHI study of Barrow that was classified by them
    as rural was classified by my method as rural.

    basically, I did some independent test of the rural proxy to make sure that it picked out rural
    stations.

    I have a write up on that but there is way too much work behind this for one paper.

  94. “Interesting work.

    You say;

    “Ten out of 12 months of data are required to count as a complete year.”

    Do those ‘missing months’ vary-i.e one record misses out January and February, whilst another misses out June and July?

    How is the missing data accounted for?
    tonyb

    ###########################################

    The complete year test is only used for selecting the urban base pair.
    That is I look through all urban stations to find those with 30 years in the 1979-2010
    window. This is needed only for the pairs study. As you note 10/12 months counts as
    a full year. There was no test for which months are missing out. Of course, I can simply
    redo the test requiring all 12 months and the number of urban stations selected as base pairs
    will go down. That takes 1 minute. Not sure if the answer would say anything about urban Bias.
    In general here is what happens as you increase the data quality: Negative trends disappear.
    That is maybe 10% of the urban trends you see are actually negative. If you move to
    12 of 12 months those negative trends diminish. Also, as you move away from the coast
    negative trends diminish.

    With some of Carricks help I may be closer to explaining most of the spurious negative
    trends that you see in the station trends. Some odd balls will persist, like dwarfs and giants.

  95. Ron says:
    December 6, 2011 at 6:06 am
    Are there really no stations in Brazil? It is a small map, and I have bad eyes, but I am not seeing a dot, either blue or red, in Brazil.

    ######################

    remember this is GHCN DAILY. so as it stands there are no stations there that meet the data
    quality standard. we could always bring in data from other repositories

  96. Bruce says:
    December 6, 2011 at 7:28 am
    “Summer land surface temperature of cities in the Northeast were an average of 7 °C to 9 °C (13°F to 16 °F) warmer than surrounding rural areas over a three year period, the new research shows. The complex phenomenon that drives up temperatures is called the urban heat island effect.”

    http://www.nasa.gov/topics/earth/features/heat-island-sprawl.html

    They looked at 42 cities.

    Zeke, NASA says 7C to 9C in the summer. You say .03C per decade. Does your data show 7C – 9C in the summer? If not, why not? What is wrong with your methodology? How do you plan to correct it?

    ########################################

    Bruce you are referring to Imhoff’s study and I’m not sure that you quite understand it.
    I’m pretty sure you havent read both of his studies. If you want to claim you have
    then I’ll gladly ask you some test questions. we will save that bit for later. for now…

    1. Imhoff looked at cities where the urban area varied from 10sq km to several hundred
    sq km. So, first off you are comparing apples to oranges.

    2. Imhoff compared the average LST ( not air temp) over the entire urban form to the LST
    of the rural landform 5 to 10km outside the city. Measuring UHI by LST
    will overestimate the difference. put another way, you cant compare SAT with LST

    3. Imhoff did not calculate a Tave in the same way that the temperature record does.
    His Tmin is not a Tmin. Its the LST taken at 130AM. His Tmax is not a Tmax
    it is the LST at 130PM.

    4. The difference in temperature between a big city and the rural zone around it
    has Nothing to do with the trends in the network we worked with.

    5. Our Rural sites are actually more rural than Imhoff’s rural as his is based on distance
    from the edge of the city.

    Like Imhoff we show a increase in trend with the increasing area of the urban form. So our work meshed pretty well with his ( heck I used his data ) The difference is this. to create his methodology he picked areas with huge urban forms. Guess what?

    A. there are not that meany of those
    B. UHI reaches a limit. And that means it doesnt show up in the trend.

  97. steven mosher says:

    “More stations just means more chances for noise and makes it harder to see any signal… however you have to be careful with your geographic selection.”

    My repeatedly asked question has never been answered: who cherry-picks the thousands of stations that were eliminated? Job security is a powerful motivator. Without knowing who dropped all those stations, it is reasonable to conclude that it was done with an eye toward skewing the results.

    Before objecting, name the perps.

  98. Smokey

    “My repeatedly asked question has never been answered: who cherry-picks the thousands of stations that were eliminated? Job security is a powerful motivator. Without knowing who dropped all those stations, it is reasonable to conclude that it was done with an eye toward skewing the results.

    Before objecting, name the perps.

    ############
    Huh?
    I start with 77,000 stations.
    I eliminate about 50,000 of those because they dont collect temperature. Go figure.
    Of the 26000 remaining I eliminate around 12000? why? They have records
    with either
    A. less than 36 months of data over this 384 month period
    B. less than 1 complete (12 month) year of data.

    This is not GHCN data.

  99. steven mosher says:
    the issue is that as you lose spatial coverage ( <50%) you tend to migrate northward.
    That means your rural stations will come from higher latitudes. Higher latitude have
    higher trends, due to polar amplification, so that you are confounding the UHI bias
    estimate by introducing a latitude bias in your rural selection.

    Again, that the right answer is difficult to achieve does not make the wrong answer more correct.

    You are imposing the necessity of a “rural selection” of existing stations upon yourself. If you cannot arrive at the correct answer by selecting from existing station data using stringent rurality criteria, then that is the result that you should report.

    If a sufficient number of rural stations do not exist in the mid latitudes to correctly reconstruct a surface temperature from the weather station proxy, then you don’t just do it anyways. You report that the the current network is not sufficient for that purpose.

  100. mosher, I asked a simple question that you avoided answering.

    Where is the 7C to 9C UHI found by NASA’s satellite?

    Surely you would want to corroborate your “findings” with an alternative method of finding UHI.

    It appears you did not find the UHI values NASA did. Either they are wrong or you are. Since they are trying to measure UHI directly, I’ll suggest you are underestimating UHI by a large amount.

  101. JJ,

    You are correct. The coverage of unpopulated areas is obviously insufficient. The thermometers have been placed where the people are. Move on to the satellites, and ocean heat content.

  102. steven mosher says:
    December 6, 2011 at 11:53 pm

    “kwinterkorn says:
    December 6, 2011 at 10:39 am
    Changes in land use within the “rural” designation may also be significant. If forests and wild grasslands are put under the plow to feed the growing world population (and fuel for cars), then the “rural” designation is confounded, and a true mixed urbanization/cultivation of wild lands bias on averaged global temps may be present.”

    Sorry, I got that covered as well. The rural sites are selected by looking at the 380sq km
    around the site. We determine that there is no “built” areas within that zone. We can also look
    at the land use in those same areas ( you will see some of that in the presentation) So, I can
    tell if there has been changes to amount of cropland and amount of grassland in these rural
    areas. In general there are slight changes in these variables and those slight changes do not
    explain the trends seen. Conversely, the trends seen at the urban stations ARE a function
    of the amount of urban area.

    ****

    So, you have all that land-use change (empirical) data, all the way back to ~1880?

    I’d reckon a guess of back to ~1980 at the earliest…..

  103. steven mosher says:

    “Huh?”

    An appropriate response, since you didn’t read the posts I was referring to, and therefore you made a wrong assumption.

    My comment on December 6, 2011 at 11:41 a.m. above is a good starting point. In particular, click on the last link in that post. Look at what happened between 1990 and 2010. [The other links are also relevant.]

    Who cherry-picked the stations that were eliminated? After reading that post you will understand that I was not being critical of you or your article. I simply want to know who, exactly, decided to eliminate most of the temperature recording sites.

    Did they have any scientifically defensible procedure for which sites, urban or rural, were eliminated? Or is this simply another addition to the mountains of evidence that government agencies are deliberately skewing the temeperature record?

  104. Steve and Zeke,

    You have a fair number of blue dots that are apparently on islands. What kind of trend do they show?

  105. Don.

    I’ll have to look at them precisely. Islands and coastal stations have much lower trends
    primarily because they are modulated by SST. In fact the effect of being near the coast is
    more important than urbanization.

  106. JJ

    “You are imposing the necessity of a “rural selection” of existing stations upon yourself. If you cannot arrive at the correct answer by selecting from existing station data using stringent rurality criteria, then that is the result that you should report.

    If a sufficient number of rural stations do not exist in the mid latitudes to correctly reconstruct a surface temperature from the weather station proxy, then you don’t just do it anyways. You report that the the current network is not sufficient for that purpose.
    ############

    I think you misunderstand. The issue is what happens to your sampling error due to spatial coverage

    So: with 5000 rural stations You have a bias estimate and an error on that estimate
    When you drop to 1000 stations classified as rural, you bias goes up, BUT so does your
    error, It’s a balance between the two. Very hard to adress in the limited space of
    the poster world. At the extreme imagine if I defined rural as having no built pixel within
    1000 km. Would you use that one station versus all the rest to measure the MEAN bias
    in the entire sample? hardly, Imagine if I told you that the most rural station compared to
    the 14000 other stations only had a . 04C bias? Why youd say that it was crazy to pick the most rural to compare to all the others. So, we present the curves.

  107. Bruce

    ‘Where is the 7C to 9C UHI found by NASA’s satellite?

    Surely you would want to corroborate your “findings” with an alternative method of finding UHI.”

    #####################

    first off you have to start with the source of that data and understand that they are looking at an entirely different thing than we are.
    ###########################################################
    “In terms of the UHI, we calculated the average temperature
    differences (LSTurban core – LSTrural) for 323 US cities
    at summer (June, July, August) daytime (1:30 PM). As
    expected, the UHI responses for the two sets of ISA were
    significantly related, with a positive correlation of 0.94
    (Figure 2B). On average, we found that UHI of Landsat
    ISA was not different from that of Nightlight ISA at
    the 0.05 significance level. For both ISAs, the amplitude of
    summer daytime UHI appears to be related to the standing
    biomass of the surrounding biome decreasing from forests
    (7 C), grass–shrubs (4 C), and then to almost no heat island
    or sometimes a heat sink in arid cities (0 C). One of the
    major contributions of such UHI differences in biomes is
    the role of vegetation during the process of precipitation,
    evaporation, and photosynthesis (Imhoff et al., 2010).

    ##################
    Several things to note that you fail to realize since you didnt read the paper.

    1. The UHI they measured was UHI in LST. That is Land Surface Temperature. As I have
    tried to explain many times LST is different than SAT. LST is the temperature of the surface
    This is both HIGHER THAN SAT ( surface air temperature ) and more variable than SAT.
    If you read the paper would would have seen that early in in the description and you would
    have read that the two measures are not comparable. SAT is taken at 2m above the surface
    That is what the temperature record is SURFACE AIR TEMPERATURE

    2. The way they measure UHI is different from the way we measure. They take two readings
    one at 130AM and one at 130PM. We take Tmin ( which happens towards dawn) and Tmax.
    So the two figures are not comparable.

    3. They restrict their measure of temperatures to cloudless days. So they measure a peak
    UHI. As we all know wind, rain, clouds all cause lower UHI. For them to take a measure
    ment they need to select days that are 99% cloud free. Peak UHI is not the measure that we are after. A peak UHI that only occurs on a few days cant be found in a monthly average
    where all days of the month are averaged. Now did the peak occur? sure. Was it 7C
    in SAT? no. it was 7C in the LST which is higher than SAT. Here is the tiny slice of
    data that they look at:
    “We use MODIS Aqua Collection 5, 8-day composite LST
    with high quality control (Wan and Dozier, 1996) and 16-
    day composite NDVI (Huete et al., 1994; 1997) at 1 km 6
    1 km resolution covering the time period 2003–2005. LSTs
    from the MYD11A2 product are retrieved from clear-sky
    (99% confidence) observations at 1:30 PM (daytime) and
    1:30 AM (nighttime) local time using a generalized splitwindow
    algorithm (Wan and Dozier, 1996). The comparisons
    between MODIS LSTs and in situ measurements across
    a wide set of test sites indicate the accuracy is better than 1 K
    with a root mean square (RMS) (of differences) less than
    0.5 K in most cases (Wan, 2008; Wang et al., 2008). LST
    observations are used to characterize the horizontal temperature
    gradient across the urban area, and NDVI is used to
    describe the vegetation density temporal variation for each
    urban zone.”

    4. The Peak UHI ( 7C ) is only found in a special class of urban environment
    You need a HUGE urban area embedded in a forest biome. In
    our data there are not many of these types of urban areas. I think there might
    be a handful of urban stations that have 400 sq km of built area. you need
    in excess of 700 sq km of urban area to generate that kind of UHI.

    There is a linear relationship between the % of area that is covered
    and the UHI. So, the problem is that you quote the worst case
    number for the worst case type of city, for a period of 8 days.
    We look at the trend over 30 years> Trends are different
    than absolute temperature differences. Further you wont
    find those kinds of cities in our data. over 5000 of our sites had NO built
    pixels for 11 km around the site.
    For our 9000 Urban stations
    1. 50% of the urban stations were located in areas that had less than
    7 sq km of built area
    2. 75% had less than 30 sq km of built area
    As figure 5 in the paper you referenced shows this is the UHI you can expect to see
    as a function of Urban Area. Im surprised you didnt see this. you did read the
    paper, right?

    A. For cities with > 500 sq km 4.7C ( average of both seasons tested)
    B for cities with 50-500 sq km 3.7C
    C for cities with 10-50 sq km 2.7C

    50% of our urban sites fall below the 10 sq km tested in this study.
    So basically, you dont see 7C in our work because a) we are looking at decadal trends
    NOT 8 days of cloud free weather. b) we measure SAT which is LOWER THAN LST
    c) we measure tave, not the temp at 130PM and AM. d) we dont have cities
    that sized in our data.

    I wish you would read papers rather than google the abstract. If you want a copy of the paper, just do what I did.

  108. Smokey

    “Who cherry-picked the stations that were eliminated? After reading that post you will understand that I was not being critical of you or your article. I simply want to know who, exactly, decided to eliminate most of the temperature recording sites.”

    you are pointing to ENTIRELY DIFFERENT DATASET than the one we use.

    that is GHCN MONTHLY. we use GHCN DAILY.

    The stations that go into MONTHLY are decided by the COUNTRY REPORTING THE DATA.
    If they decide to report a monthly figure to GHCN Monthly, then it goes in. Please understand
    the acronymn

    G = Global
    H = Historical
    C = Climate
    N = Network

    This was a project that was completed some time ago back in the 90s. Historical records were collected and QA’d. Finished. Updates to the series depend upon countries supplying the
    data. They stop supplying monthly data, the station disappears.

    With the data we used DAILY, you have a set of stations were the countries agree to
    supply daily data. last week 500 stations were added to the list. So there is no one person who decided to “drop” stations. The HISTORICAL collection was made in the 90s and then
    certain countries decided to continue to send data for a subset of their data.

    Finally, the station drop out does not matter. we proved that a dozen times.

  109. steven mosher,

    It doesn’t matter to me about the ‘station drop out’. As I asked: “Who cherry-picked the stations that were eliminated? After reading that post you will understand that I was not being critical of you or your article. I simply want to know who, exactly, decided to eliminate most of the temperature recording sites.”

    Got names?

  110. Smokey

    1. The stations were not cherry picked, so your question makes no sense. The stations
    were not eliminated, so your question makes no sense.
    2. I’ve explained the proceedure, no one at NOAA decided what stations countries
    choose to report.
    3. Over the period in question you can bet that various people at various NWS have
    decided to stop reporting monthly figures for stations. They may choose to report
    Daily, for example.
    4. Fewer NWS send there reports into NOAA. If you want to know the name of the people
    at the NWS at 280 reporting countries have decided not to report, then you have 280
    letters to write. Get busy

  111. steven mosher,

    I suspect you’re deliberately misunderstanding what I asked. You stated that the stations that were eliminated were not cherry-picked. You know this, how?

    It appears to be misdirection to say, “no one at NOAA decided what stations countries choose to report.” I’m concerned about how my tax money is spent, so let’s just confine the cherry-picking to U.S. stations. As you can see, between around 1990 and now, thousands of stations in the U.S. have been eliminated. What was the criteria? Was it an honest, documented policy, or was it arbitrary?

    Since you unequivocally stated that “1. The stations were not cherry picked…“, then there must be a documented procedure for determining which stations remained, and which were eliminated. You put yourself forth as a knowledgeable expert, so I’m asking again: who is/was the decision maker regarding which stations would be eliminated? I’m not arguing. It’s a straightforward question.

  112. Say what you will about cherry picking, but there is a clear difference between the data systematically reported by the met services of various countries (as evidenced by monthly summaries shown at “tu tiempo”) and the subset available through GHCN. The latter inexplicably fails to update many reltively trendless small-town records in Turkey, China and many other countries beyond 1990, while religiously providing results from trending urban stations.

  113. mosher:

    “The way they measure UHI is different from the way we measure. They take two readings
    one at 130AM and one at 130PM. We take Tmin ( which happens towards dawn) and Tmax.
    So the two figures are not comparable.”

    That means NASA’s UHI could have been underestimated.

    Have you compared the UHI in NASA’s 42 cities to the same 42 cities in your dataset? Have you found similar patterns of UHI?

    If not, why not? Aren’t you interested in confirming whether you are even close to being right?

  114. Sunshine:

    “That means NASA’s UHI could have been underestimated.

    Have you compared the UHI in NASA’s 42 cities to the same 42 cities in your dataset? Have you found similar patterns of UHI?

    If not, why not? Aren’t you interested in confirming whether you are even close to being right?”

    #####################

    1 wrong, if anything the UHI would be over estimated. It’s pretty simple once you understand how UHI works. But you can read this study and see that in detail in the hourly data
    for a couple cities in texas.

    http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=10&ved=0CHEQFjAJ&url=http%3A%2F%2Fams.confex.com%2Fams%2Fpdfpapers%2F126615.pdf&ei=D3biTpmuLaSZiQLB0eHdBg&usg=AFQjCNFXuXjv6LbLu65fEGqPkexcQw_Leg&sig2=yqwoIeQP8PU_vkIX_yaduA

    2. Imhoff does not list his 42 cities. Over 50 % of our urban sites are smaller than the smallest
    cities that Imhoff looked at. 75% of our urban areas have less than 30 sq km of
    urban surface. Which would fall in his smallest category of city. In short,

    3″ Have you found similar patterns of UHI?.” The two studies are asking entirely different
    questions. In Imhoff’s study he selects cities of certain sizes. Then he measures the LST
    at 130am and 130 PM. LST is higher than SAT. the two are not comparable.
    I can explain it for you but I cannot understand it for you.
    read this

    http://land.umn.edu/documents/Urban_heat_island–Impervious__RSE_paper.pdf

    See table 1: LST runs anywhere from 0K warmer to 11K warmer versus Tave
    calculated from Tmin and Tmax.

    Since we have SAT measures we CANNOT compare those to LST. we cannot
    and would not for the following reason. LST is not used to build the land record
    so estimating the UHI in the SAT record by looking at LST would be pointless.
    Also any differences in UHI measured by LST and that measured by SAT would
    be pointless as LST measures one thing and SAT measures an entirely different
    thing. That said, we do find a simllar relationship between the SiZE of the
    city and the size of the bias. That is, there exists a linear reationship
    between the size of LST UHI and the size of the city. There is also a linear relationship
    between the size of the city and size of the SAT bias. BUT, the SAT bias is smaller
    than the LST bias for a variety of reasons. Further we are looking at TRENDS.
    Not 8 special days in the summer.

  115. sky says:
    December 8, 2011 at 3:52 pm
    Say what you will about cherry picking, but there is a clear difference between the data systematically reported by the met services of various countries (as evidenced by monthly summaries shown at “tu tiempo”) and the subset available through GHCN. The latter inexplicably fails to update many reltively trendless small-town records in Turkey, China and many other countries beyond 1990, while religiously providing results from trending urban stations.

    ##################################

    GHCN does not fail to update. The updates are provided TO GHCN and GHCN reports what is reported to it. They CANNOT update data when they dont recieve it!

    Secondly, all ALL RURAL record does not differ substantially from an ALL URBAN record.
    How much does it differ? about .03C per decade over the 1979-2010 period.

    How do we know that UHI is small over this period. Simple. Look at the trend over land for
    RSS. compare that to the land record. The difference? less than .1C per decade
    for the 1979-2010 period. UHI is real. You can find some days where for some cities
    it is huge. However, those days get averaged with other days, those months get averaged
    with other months and when you finally look at the trends over decades, those effects
    are small. on the order of .03C per decade ( +- .03C)

  116. steven mosher says:
    December 9, 2011 at 1:31 pm

    “GHCN does not fail to update. ”

    When clearly available data does not appear in the time series provided by GHCN, only sophists would contend that there is no failure to update.

    “ALL RURAL record does not differ substantially from an ALL URBAN record.”

    Nonsense! When vetted “rural” records are available in sufficient abundance to establish 20th century temperature variations in a region (e.g., USA), they show that the urban records contain a substantial (> 0.5K/century) trend not found in the former. The “ALL- record” comparison is a piece of sophistry, which exploits the fact that century-long “rural” records are UNAVAILABLE over much of the globe. Statements relying upon the urban bias in the available database, rather than in nature, just don’t cut it. And the cherry-picked interval 1979-2010, which coincides roughly with the rising half-cycle of a mutidecadal swing, tells us nothing about SECULAR trend. Your contrived contentions may appeal to a naiive audience, but they don’t begin to address the scientific issues at hand.

  117. Sky

    “Nonsense! When vetted “rural” records are available in sufficient abundance to establish 20th century temperature variations in a region (e.g., USA), they show that the urban records contain a substantial (> 0.5K/century) trend not found in the former. The “ALL- record” comparison is a piece of sophistry, which exploits the fact that century-long “rural” records are UNAVAILABLE over much of the globe. Statements relying upon the urban bias in the available database, rather than in nature, just don’t cut it. And the cherry-picked interval 1979-2010, which coincides roughly with the rising half-cycle of a mutidecadal swing, tells us nothing about SECULAR trend. Your contrived contentions may appeal to a naiive audience, but they don’t begin to address the scientific issues at hand.”

    1. The all rural record for the USA using Ghcn daily disagrees with your unsubstantiated
    claims. Urban records don’t contain the signal you claim. Sorry, but I have to believe the data.

    2. No one I know of has applied the kind of strict Filter I use for rural, so I’m skeptical of
    the truth of what you say.

    3. Your comments about nature are nonsensical.

    4. 1979 to 2010 was selected for several reasons
    A. Urbanization takes off dramatically
    B. Anthony’s study covered that time period
    C. we can compare to satellites and see that we are correct.
    D. we do measure the secular Trend, sorry no cookie for you.

  118. steven mosher says:
    December 9, 2011 at 6:47 pm

    The simplistic assumption that all station records consist of a linearly trending climate “signal” that is spatially coherent over great distances plus a constant station-specific offset plus noise is what keeps you from any truly incisive analysis of measured data. The situation is far more complicated than that and is further obscured by pernicious data quality problems. Vetting of individual records is required to ascertain that it provides a credible time-history of temperature at a FIXED location–the sine qua non for studying climate CHANGE. By no means are ALL records suitable for such studies.

    Cross-spectrum analysis of century-long vetted records shows that truly urban records (where the station is located within the “bubble” of elevated urban temperatues throughout the entire duration) DO contain strong very-low- frequency components that are INCOHERENT with such components in neighboring non-urban records. It is these spatially varaible VLF components that make for highly inconsistent “trends” amongst neighboring stations. Over the course of the past century USA urban records in the aggregate manifest a >0.5 K rise that is NOT there in the vetted non-urban aggregate of raw data.

    To the extent that proxy records are indicative of climate change, the indication from many spectral studies is that multidecadal and quasi-centennial oscillations dominate decadal -scale rates of change. Your idea that a 30-year regressional trend IS a secular one only reveals a lack of comprehension of analytic basics. That’s the way the cookie often crumbles in analysis of climate.
    I don’t have time to elaborate more or to argue fruitlessly against uncomprehending crank-turning approaches to doing “science.”

  119. Smokey: You asked

    “Since you unequivocally stated that “1. The stations were not cherry picked…“, then there must be a documented procedure for determining which stations remained, and which were eliminated. You put yourself forth as a knowledgeable expert, so I’m asking again: who is/was the decision maker regarding which stations would be eliminated? I’m not arguing. It’s a straightforward question.”

    ###

    I think Steven did a good job explaining it directly above the comment you cite, so maybe that’s the reason he chose not to bother hitting his head against the keyboard again. The extent to which he was willing to take time out of his life to clarify his research to what seems to me (first time reader) be a hostile audience is kind of amazing. Maybe at some point spending time with his family takes priority over responding repeatedly to the same question?

    To me, Cherry picking implies that you think someone intentionally removed stations in order to generate a desired result. Doesn’t it seem most likely that changes like this are due to decisions made by hundreds or thousands of people over multiple years? It seems very likely to me that these decisions were “arbitrary”, but that’s extremely different from “cherry picked”.

    ###

    You said: “Did they have any scientifically defensible procedure for which sites, urban or rural, were eliminated? Or is this simply another addition to the mountains of evidence that government agencies are deliberately skewing the temeperature record?”

    Again, presumably these decisions had nothing to do with science. If scientists were the ones making the decisions (and with an unlimited budget) clearly the bias would be towards keeping as many stations operating as possible :)

    If there was some kind of a grand conspiracy, don’t you think any one of the many right and center-right governments currently ruling in multiple countries would have discovered it?

    ###

    Finally, to many of the earliest posters (who I know will never see this buried comment), I don’t think you understood the third plot. The urban and rural stations showed basically the same trend, there was just a small difference in the amplitude of the trend they observed.

Comments are closed.