Regional U.S. Population Adjustments to Surface Temperatures Since 1973: Still Little Warming

April 6th, 2012 by Roy W. Spencer, Ph. D.

UPDATE: I’ve added 6 regional U.S temperature plots, for the Northwest, North Central, Northeast, Southwest, South Central, and Southeast.

Thanks for the comments and suggestions on yesterday’s post introducing a new U.S. population-density adjusted temperature (PDAT) dataset. As a result of the comments, I have stratified the U.S. into 6 subregions, and performed station temperature trend vs. population regressions, rather than just lumping all 280 stations together. This should help reduce the effect of any fortuitous (but real) regional warming which just happens to be located where more people live. It should better isolate the true urban heat island (UHI) effect on temperature trends.

The results are shown in the following plot, with the regional regression coefficients listed being the scaling factor between station temperature trend (deg. C/decade) and population density (persons per sq. km) to the 0.2 power (click for high res. version):

As can be seen, 4 of the 6 regions have quite strong dependence of the trends on population density. Only the Southwest U.S. has cooling with increasing population density, which is probably the result of people planting vegetation in what is mostly an arid region to begin with.

The impact of making regional — rather than whole U.S. — population adjustments on the U.S. average temperature variations results in only a slight increase in the resulting temperature trend I posted yesterday, which is still well below that computed from the CRUTem3 dataset (click for high res. version):

Of course, the regional trends would change substantially, since now I am actually warming the Soutwest U.S. temperatures with time, based upon station population density. But the Southeast U.S. trends will be cooled even more than before, because of the strong relationship between temperature trend and population density found there (see 1st plot, above).

UPDATE: Here are the population-adjusted temperature variations for the 3 northern U.S. sectors, with just the trailing 12-month averages plotted to reduce the messiness:

…and here are the 3 southern U.S. sectors:

The bottom line is that there is still clear evidence of an urban heat island effect on temperature trends in the U.S. surface station network. Now, I should point out that most of these are not co-op stations, but National Weather Service and FAA stations. How these results might compare to the GHCN network of stations used by NOAA for climate monitoring over the U.SA., I have no idea at this point.

Also, I need to clear up a misconception…the adjustments I perform do not remove the trends in the data. They remove only the component of the trend which is due to population density, using the regression coefficient alone (not the regression constant). There are no adjustments in January 1973 (the beginning of my data record), and then the adjustments increase linearly with time.


21 Responses to “Regional U.S. Population Adjustments to Surface Temperatures Since 1973: Still Little Warming”

Toggle Trackbacks

  1. Eric Hein says:

    Not to trivialize your effort here Dr. Spencer but this is a bit of a “duh” moment. We already know that land use changes cause changes in regional climate. Increases in population cause land use changes so…increases in population must cause changes in regional climate.
    Thanks for taking the time to quantify the affect. Maybe this will be the straw that breaks the CAGW’s back.

  2. Doug Cotton says:

     
    Urban crawl would have less effect on climate measurements if the issue of weighting land v. ocean temperatures were addressed correctly.

    It is logical that ocean (sea surface) temperatures ought to be weighted in accord with the relative thermal energy content compared with land surfaces.

    Rather than about 70:30 it should be more like 89:6 I understand.

    Urban crawl is not just something to do with black roads etc on the ground. Vertical walls of buildings still can remain warmer than the surface in the early hours of the evening. Whatever their temperature, they block a “full view” of the atmosphere from the ground and do send some radiation towards the ground, thus slowing surface cooling. Even human and animal bodies will do likewise.
     

     

  3. iya says:

    One cause of urban heating is, of course, waste heat.
    It’s not too difficult to quantify, assuming all primary energy ends up as heat, which is almost true, as the only exception seems to be light going directly to space.

    A reasonably populated region like New York state produces about 1 W/mē:
    http://www.wolframalpha.com/input/?i=4134640+BBtu+%2F+yr+%2F+NY+state+area

    Another example, I tried was Germany, which gave 1.3 W/mē.

    It’s still negligible on a global scale (~30 mW/mē), but most weather stations are in cities, where it can reach 10 W/mē and more.

  4. Dr. Spencer has a habit of telling us things we have known about for years and years.

    I on the other hand, am devoting my time to things we don’t know, such as my PHASE IN THEORY ,and how that can be applied to show how abrupt climatic change can come about as a result of the phasing in of the elements that control the climate.

    This theory is the best explanation, for this important aspect of climatic change.

    The stage is set for global cooling this decade. The only question that remains is how much cooling. That will depned on large part on future solar activity/volcanic activity, not to mention the cold PDO, with the AMO soon to follow.

    All these items are pointing toward colder conditions going forward. The data seems to be supporting this view since the rise in temperatures last century, have now come to a halt.

    Time will tell.

    • Bob G. says:

      “Dr. Spencer has a habit of telling us things we have known about for years and years.”

      We may know something is likely to be true. However, there is a big difference between “knowing” and providing convincing scientific evidence something is true. Dr. Spencer has found a clever new way to provide convincing scientific evidence that the UHI is larger and more important than is represented in CRUTem3.

      Perhaps the largest problem with AGW alarmists is that they were very sure that CO2 causes a large amount of warming and that there was strong positive feedback associated with it. They went from that “certainty” to looking at what would happen in the future if given that their theory was true. They totally skipped the part where they needed to provide convincing scientific evidence that they were right about positive feedback. And that problem remains today.

  5. phi says:

    iya,

    It should be added that the deficit of evaporation is very nearly equivalent for the mid-latitudes temperate climates. If we relate the numbers to fully drained surfaces (roofs and pavements) and we review energy taking into account solar gain, we find that the heat production of an urbanized area is roughly twice of a natural equivalent surface (200 W/m2 against 100 W/m2). Obviously, these values ??must be reduced according to the coefficient of urbanization (around 0.9 for a town center but less than 0.2 for the periphery). These are just orders of magnitude but they are eloquent.

    • iya says:

      Are you saying urbanized areas are warmer because there is less evaporation?
      But evaporation just transfers energy from the surface to the atmosphere, so the air should actually be cooler?

  6. phi says:

    iya,
    A significant part of solar energy is not converted to heat on the ground but in latent energy of evaporation. In an urban environment, much of the rainwater is dischaged into drain and not evaporated in situ. Hence more energy converted to heat and more warming in urban areas.

  7. Jon says:

    What I recall is that over the COTUS, the tob adjustments contribute a majority of the warming trend.

    I don’t dispute that a tob adjustment is called for–but it has always seems fishy to me that the tob adjustments would have a have a net trend.

    Your results use hourly data so there is no need for tob adjustment. Meanwhile, the tob adjusted coop stations don’t seem to have a clear urban/rural distinction–didn’t Mosh and Zeke estimate something with an upper limit of 0.05C/dec?

    In sum of these facts, I am wondering whether the tob adjustment tables are wrong. Perhaps for reasons as silly as DST but perhaps more concerningly whether there was some computational corner cutting by Karl in the 80s. For one thing the physics of the adjustment depend on the shape of the diurnal temperature profile, this shape is plainly geography/land use dependent not just latitude dependent as the original work assumed.

    Apologies for my ignorance but are their modern studies of tob bias?

  8. BillC says:

    Jon,

    I believe this is an important issue. It would be fantastic to create a crowd sourced analysis where individuals take responsibility for a station or two.

    To my mind, I feel like the hourly vs tob is an issue with BEST.

  9. Kasuha says:

    “Also, I need to clear up a misconception…the adjustments I perform do not remove the trends in the data. They remove only the component of the trend which is due to population density, using the regression coefficient alone (not the regression constant). There are no adjustments in January 1973 (the beginning of my data record), and then the adjustments increase linearly with time.”

    Your description matches the way how I understand the basis of your adjustment and I still have my doubts. Please let me explain how I understand your adjustment method. If I use the data from your previous article where the regression function was 0.0587 x + 0.0128, then assuming the population coefficient is 1 (i.e. 1 person per square kilometer, the lowest population for which any data are available) you are removing trend of 0.0587 °C/decade from it (i.e. adding values of linear function which has value 0 at the beginning, and value -0.0587 at 10 year mark), leaving average trend of 0.0128 °C/decade for stations with population 1. For stations with population coefficient 2 (i.e. 32 people per square kilometer) you are removing trend of 0.1178 °C/decade, leaving average trend of 0.0128 °C/decade for this group as well. Yes I know you are not doing it with whole numbers, rather with fractional numbers corresponding to the stations’s population density but that does not change much on the result. If you do this for each population group and then calculate average of all stations, what you get is the average trend you adjusted everything to, i.e. 0.0128 °C/decade. But there is no population group of stations with average trend 0.0128 °C/decade, that’s just an extrapolation of the regression to theoretical stations with population zero, which, if they existed, would likely have average trend similar to stations with population 1 rather than to this extrapolation.
    There are two things in total which I don’t like on this. First is that the extrapolation is used instead of adjusting it to lowest population station trends, and second is that linear regression is used. You write “we find that the warming trend with time increases rapidly with population at low population densities, then levels off at high population densities” yet you still go on with linear regression.
    Maybe I am wrong and you are right. Maybe I just misunderstand the way you performed the adjustment and you did it differently and if that’s the case please accept my apologies. But if the adjustment was done the way how I understand it then it’s in my opinion wrong.

  10. P. Solar says:

    Dr Spencer,

    it would be very useful if you could post the basic data along with posts like this.

    I have a few ideas I’d like to check out that may pull some more information out of this but I don’t have time to dig the original data and try to imitate the processing you have described verbally to get something similar.

    Would you have any objections to posting a text file or xls with the data used for the graphs you presented here.

    some journals now ask for data to be supplied with submitted papers (though sadly they rarely actually enforce it).

    It would seem like a good policy all round and would likely produce the occasional insight by some sharp minded reader.

    Thanks.

  11. P. Solar says:

    Sorry, I should say I would like to see the pre-adjusted data from the stations you selected.

    best regards.

  12. rob says:

    Hi Roy

    Seems like others have with similar conclusions to your own have managed to get their work published:

    http://www.eike-klima-energie.eu/uploads/media/How_natural.pdf

  13. Ted says:

    Dr. Spencer,

    Your own UAH temperature record shows a clear warming trend. Is it affected by the urban heat island effect as well?

  14. Dr. Strangelove says:

    Dr. Spencer, you said

    “the adjustments I perform do not remove the trends in the data. They remove only the component of the trend which is due to population density”

    Why would want to do that? AGW and UHI are due to human population. Isn’t it also true that the higher the population in an area, the higher the CO2 emission, the higher the CO2 concentration in the air? Of course it will eventually diffuse but air pollution is also correlated to population density. Try that analysis.

  15. Cristoph S says:

    I used to read all the time that we shouldn’t trust the land-based readings and should only trust the satellite data.

    I predicted in a post last year that if the satellite data continues to show warming, it will have to be attacked as wrong. Dr. Spencer is proposing now that we should believe “adjusted” temperature readings instead of (his “own”) satellite data. Interesting twist.

    If the 13 month average goes above the max in 2010 again soon, I would not be surprised if Dr. Spencer stops displaying the updates on this site. Of course, the satellite data could also be “adjusted” too!

    • Kasuha says:

      I think you missed the point here – the whole post is not about satellite data at all. Both GISS data (supposedly adjusted to UHI) and unadjusted ISH data are land-based measurements. On the first picture you can even find locations of individual ISH stations.
      The point of the article is that GISS seems to pronounce UHI effects rather than to suppress them.

      You’re probably mistaking UHI (urban heat island) for UAH (University of Alabama, also synonym for processed satellite measurements).

      Satellite data, neither from UAH nor from RSS, don’t have the problem with switching order of warmest years with every release. GISS does.

      • Cristoph S says:

        I assure you I am not confused about satellite vs. land or UHI vs. UAH.

        My point is that we are seeing a shift in style here from
        saying the data doesn’t show warming to saying the data is wrong. If the data sets (either satellite or ground) show the next decade is warmer than the last (again), then the data will have to be called into question.

        The fact that Dr. Spencer suggested the UAH data might be wrong is pretty stunning to me.

        UPDATE #2: Why the Discrepancy with UAH LT Temperatures?

  16. D J Cotton says:

     

    BREAKING NEWS

     

    NASA SCIENTISTS DEMAND END TO GW FRAUD

     

    Continuing man-made global warming fraud triggers a mass NASA rebellion. Rebels demand U.S. government pulls plug on the climate catastrophe cult. Dozens of top experts including astronauts and engineers trigger meltdown in American space agency.

    http://johnosullivan.wordpress.com/2012/04/11/nasa-in-mass-revolt-over-global-warming-fraud/