I believe this is a truly important piece of work. I hope Dr. Spencer will submit it to a journal. I’m grateful to Dr. Spencer for his email suggesting I post it here. Consider this early peer review. Beat it up, find any errors, and point out flaws, so that he can make it better. – Anthony
by Roy W. Spencer, Ph. D.
UPDATED (12:30 p.m. CST, March 3): Appended new discussion & plots showing importance of how low-population density stations are handled.
ABSTRACT
Global hourly surface temperature observations and 1 km resolution population density data for the year 2000 are used together to quantify the average urban heat island (UHI) effect. While the rate of warming with population increase is the greatest at the lowest population densities, some warming continues with population increases even for densely populated cities. Statistics like those presented here could be used to correct the surface temperature record for spurious warming caused by the UHI effect, providing better estimates of temperature trends.
METHOD
Using NOAA’s International Surface Hourly (ISH) weather data from around the world during 2000, I computed daily, monthly, and then 1-year average temperatures for each weather station. For a station to be used, a daily average temperature computation required the 4 synoptic temperature observations at 00, 06, 12, and 18 UTC; a monthly average required at least 20 good days per month; and a yearly average required all 12 months.
For each of those weather station locations I also stored the average population density from the 1 km gridded global population density data archived at the Socioeconomic Data and Applications Center (SEDAC).
All station pairs within 150 km of each other had their 1-year average difference in temperature related to their difference in population. Averaging of these station pairs’ results was done in 10 population bins each for Station1 and Station2, with bin boundaries at 0, 20, 50, 100, 200, 400, 800, 1600, 3200, 6400, and 50000 persons per sq. km.
Because some stations are located next to large water bodies, I used an old USAF 1/6 deg lat/lon percent water coverage dataset to ensure that there was no more than a 20% difference in the percent water coverage between the two stations in each match-up. (I believe this water coverage dataset is no longer publicly available).
Elevation effects were estimated by regressing station pair temperature differences against station elevation differences, which yielded a cooling rate of 5.4 deg. C per km increase in station elevation. Then, all station temperatures were adjusted to sea level (0 km elevation) with this relationship.
After all screening, a total of 10,307 unique station pairs were accepted for analysis from 2000.
RESULTS & DISCUSSION
The following graph shows the average rate of warming with population density increase (vertical axis), as a function of the average populations of the station pairs. Each data point represents a population bin average for the intersection of a higher population station with its lower-population station mate.
Using the data in the above graph, we can now compute average cumulative warming from a population density of zero, the results of which are shown in the next graph. [Note that this step would be unnecessary if every populated station location had a zero-population station nearby. In that case, it would be much easier to compute the average warming associated with a population density increase.]
This graph shows that the most rapid rate of warming with population increase is at the lowest population densities. The non-linear relationship is not a new discovery, as it has been noted by previous researchers who found an approximate logarithmic dependence of warming on population.
Significantly, this means that monitoring long-term warming at more rural stations could have greater spurious warming than monitoring in the cities. For instance, a population increase from 0 to 20 people per sq. km gives a warming of +0.22 deg C, but for a densely populated location having 1,000 people per sq. km, it takes an additional 1,500 people (to 2,500 people per sq. km) to get the same 0.22 deg. C warming. (Of course, if one can find stations whose environment has not changed at all, that would be the preferred situation.)
Since this analysis used only 1 year of data, other years could be examined to see how robust the above relationship is. Also, since there are gridded population data for 1990, 2000, and 2010 (estimated), one could examine whether there is any indication of the temperature-population relationship changing over time.
This is the type of information which I can envision being used to adjust station temperatures throughout the historical record, even as stations come, go, and move. As mentioned above, the elevation adjustment for individual stations can be done fairly easily, and the population adjustments could then be done without having to inter-calibrate stations.
Such adjustments help to maximize the number of stations used in temperature trend analysis, rather than simply throwing the data out. Note that the philosophy here is not to provide the best adjustments for each station individually, but to do adjustments for spurious effects which, when averaged over all stations, will remove the effect when averaged over all stations. This ensures simplicity and reproducibility of the analysis.
UPDATE:
The above results are quite sensitive to how the stations with very low population densities are handled. I’ve recomputed the above results by adding a single data point representing 724 more station pairs where BOTH stations are within the lowest population density category: 0 to 20 people per sq. km. This increases the signal of warming at low population densities, from the previously mentioned +0.22 deg C warming from zero to 20 people per sq. km, to +0.77 deg. C of warming.
This is over a factor of 3 more warming from 0 to 20 persons per sq. km with the additional data. This is important because most weather observation sites have relatively low population densities: in my dataset, I find that one-half of all stations have population densities below 100 persons per sq. km. The following plot zooms in on the lower left corner of the previous plot so you can better see the warming at the lowest population densities.
Clearly, any UHI adjustments to past thermometer data will depend upon how the UHI effect is quantified at these very low population densities.
Also, since I didn’t mention it earlier, I should clarify that population density is just an accessible index that is presumed to be related to how much the environment around the thermometer site has been modified over time, by replacing vegetation with manmade structures. Population density is not expected to always be a good index of this modification — for instance, population densities at large airports can be expected to be low, but the surrounding runway surfaces and airplane traffic can be expected to cause considerable spurious warming, much more than would be expected for their population density.
wow, lots of questions. I’ll respond to as many as possible early in the morning…most are simply the result of the brevity of my post.
Keep in mind that these are the average UHI effects across ~10,000 stations. Obviously, the UHI effects at specific stations could be much larger.
Also, as the post states, this effect is not due to the sensible heat generated by people’s bodies….it’s due to the sensible heat generated by things that people build, whether an active source (e.g. AC exhaust), or a passive source (pavement rather than grass).
Dear Dr. Spencer,
as far as I remember, Ferd. Engelbeen did something similar with belgian/netherland/german longterm temps datas and UHI effect,
maybe eight to ten years ago. (sorry, the URL’s don’t work anymore)
Or was it Jan Janssens (solaemon)?
Conclusions were nearly same, steepest UHI effect/curve starts at/with lowest
populated area. This validates your approach.
Is it possible to do a logarithmic plot as well? Especially if there is some evidence that there is a logarithmic relationship.
What I also would like to know is what the next step might be in this piece of research since population density isn’t entirely reflective of amount of human infrastructure intervention on the thermometer record (human land use effect).
I just have to say I like this idea. Hope it works out as planned.
This looks to me to be correlated to Dr. Longs article here;
http://wattsupwiththat.com/2010/02/26/a-new-paper-comparing-ncdc-rural-and-urban-us-surface-temperature-data/#more-16726
Why not join forces? Spencer & Long et.al.
Give NCDC something to think about.
Having earlier said that inverted, the graph looks like temp change sensitivity, another random “thought” (work with me here) occurs. Assuming more CO2 in an urban setting (cars, etc.) isn’t this also demonstrating how GHG’s work in a microclimate i.e. less upward swing per higher CO2 concentration? That could explain the curve.
As such this one graph looks to corroborate basic GHG theory, a lot of the IPCC stuff, i.e. everything important.
@Jan Frykestig (11:39:22) :
—> “” Spencer: “this means that monitoring long-term warming at more rural stations could have greater spurious warming than monitoring in the cities.”
One would then reasonably expect to see much higher trend at rural stations than at urban ones, ok? However, the USA data show exactly the opposite – just 0.1 deg C of warming in 20th century, but 0.6-0.7 degrees C in the cities. How that can be, if the rural stations should have a larger warming bias? “”
No. Spencer is talking 2nd derivative and you’re talking first derivative. Spencer is saying that you may see greater change in temperature anomaly at sites where the population is increasing from near zero to 1000 per square km. You are talking about greater temperature anomalies in the same areas. Do you see the difference? Spencer is saying depending on site, you might expect a rural station to show much greater change in anomaly per with respect to population changes, and you are asking why we don’t see greater temperature anomaly with respect to population. That is a mismatched question because Spencer is simply talking about a different slope than you are asking about.
What is interesting to me is how the lines in the first two warming bias graphs bend down at about 500 pop per sq km, but after that, do not level off. I would have expected each additional person to have a marginal impact on warming bias, but that does not seem to be the case.
Good idea. Never thought of it that way before. This makes me wonder has anyone ever done an affluence and UHI effect analysis? As a region get more affluent more paved roads instead of dirt, larger homes, more suburbia type energy usage.
I can see packing more and more people into a small area (ie. Mexico City) as affecting UHI but I can also see massive suburban sprawl (ie. Calgary) having similar effect on UHI for quite different reasons.
“Land Use/Cover Relationship for Thermal Urban Environment Studies Using Optical and Thermal Satellite Data”
From:
http://www.fig.net/pub/vietnam/papers/ts01g/ts01g_mallick_etal_3751.pdf
“satellite datasets are used to analyze the spatial structure of the thermal urban environment and the ‘hot’ surfaces within the urban settings were identified that is related to the urban surface characteristics and land use/land cover.”
“All station pairs within 150 km of each other had their 1-year average difference in temperature related to their difference in population. Averaging of these station pairs’ results was done in 10 population bins each for Station1 and Station2, with bin boundaries at 0, 20, 50, 100, 200, 400, 800, 1600, 3200, 6400, and 50000 persons per sq. km.”
Roy: This section really needs cleaning up. You calculated a running average for any pair of stations within 150 km of one another using four time points per day and omitted any time points where the two stations didn’t both have a reading. That’s the 1-year pair-wise average of temperature difference. Then, you assign each of the stations to a population bin. What, exactly, are you averaging the second time? Are you averaging the average? I don’t see how you are averaging them in bins or why the bins are necessary.
What I don’t understand about this blog is that people are congratulating Roy on proving something, when I don’t think his article can even be approached due to lack of definitions & clearly explained method. What is up with that?
I have a thermometer on my bicycle. Temperatures usually drop 3-4 degrees in the afternoon, pedaling from a wide paved road to a park two blocks away.
One road is enough to generate a significant UHI effect.
Dr. Spencer, your post contains some very good analysis! Bravo! Respectably, you almost carried that analysis its final end, yet not quite.
The above graphs, of Station Warm Bias vs. Population Density, need to be carried to its y-intercept. At that y-intercept, you should find the temperature anomaly increase that earth has experienced from some unspecified cause(s) over that period and should converge over longer periods of time to near the SST anomaly over the same period. That is the final answer as to why published temperature anomaly values are all hugely exaggerated.
The causes of this small increase could take many forms. AGW proponents are going to try to blame it totally on CO2. Those like myself are going to place a large component on the secular increase in TSI seen over the last three or so cycles and has been termed in some places and papers as The Modern Solar Grand Maximum, but this has yet to be proven. A certain amount definitely needs to be placed on global-wide nuclear electricity generation as it is an unnatural source of heat whose tiny fraction of total input is still continually increasing. A certain amount can be placed in the factional increase in various fossil fuels over the last few decades, the heat given off by combustion, not CO2 secondary effects. But all in all, we are talking about a magnitude smaller delta temperature anomaly over past decades, not the 1.7ºC or so seen when UHI sites are improperly included.
It seems our beloved water is the major player in the reason CO2 is not causing the so called “greenhouse gas effect” as many have tried to conjunct. The effects of water evaporation and condensation and water vapor itself dwarf any CO2 component in a planet scale climate system. This Earth does seem to have some mechanism(s) that consistently keep Earth’s temperature in balance with the overall input to the Earth climate system. I could list the many candidate hypotheses here but I would leave some out, which wouldn’t be fair, so I will restrain.
Please carry your paper to its very end at the y-intercept over longer periods. In your last post I commented a simple hypothetical system that would continually measure this anomaly after all daily solar fluxes were removed. That will force science to finally turn to answer the real base questions that have needed answered for so long.
Where are the error bars?
1. In meteorological dispersion modeling population density is just as acceptable as land use category in rural/urban determinations by the EPA. However, land use categories may convey a more direct correlation with the climate forcing in question.
2. Anticipating attacks on such a study, I would settle for a smaller sample size which has a higher degree of credibility. In this light, I would reduce the distance between stations by at least half, and set a criteria for elevation difference.
3. You may want to section the correlations into nightime vs. daytime temperatures in addition to what you have already presented.
Dr Spencer.
Firstly, thank you for undertaking this task. I hope all sides of the debate will contribute in a positive manner to ensure that your work ultimately achieves benchmark staus within climate science.
We are faced with a situation where total numbers of stations are declining, but I suspect that the proportion of airport stations is actually increasing (perhaps you could let us know out of interest the proportions in your sample set). I think it is important to attempt to account for them separately so that you can determine whether their effect is skewing your overall findings. On this basis I am in agreement with David (12:57:12).
G.L. Alston (12:34:41) :
I’ll have to agree with G.L. Alston (12:34:41) : … this doesn’t seem to measure UHI as much as it does sensitivity to temperature change per population density. Which in and of itself could be of great use when doing site adjustments.
Dr. Spencer,
How difficult woud it be to do the same analysis over time, as the population density of rural sites doesn’t increase uniformly? Areas that developed faster than others should show a warming bias compared to non-developing neighbors, then as those neighbors catch up their temperature should likewise catch up.
It should provide a confirmation of your delta T/pop density relation in time as well as across areas, further enhancing the robustness of the result.
P.S. Is saying “robust” kind of like saying “shrubbery”?
Sangaman & Robert of Ottawa:
You guys beat me to it. If there are 10 people living in a km^, chances are good they are farmers. Bring in a couple hundred more, and chances are, they’re into farming also. So we get substantial changes in land use which do not occur with corresponding population changes in urban areas. Point is, and consistent with Spencer’s work: it’s not CO2, its other human behaviors.
Thanks Dr Spencer, an interesting read. Confirms work by others.
Robert of Ottawa: UHI appears to work more at night, when buildings, roads etc cool more slowly than grass and trees. As well, the structures act like hairs or feathers to keep the ground warm on a cold morning. So it affects minimum temperatures more than max, which we’ve seen here in Australia.
I’m pleased to see recognition that even very small settlements will have UHI. Any change to natural environment will change temperature a little- just clearing a few trees and ploughing the ground, for example. Adding buildings and concrete makes it worse, and you don’t need many people to do that.
So presumably the population assiciated with each of thse stations has also varied with time. Is this data known, and can it be factored into the historic temperature records; since if population is affecting today’s data, it presumably also affected historic data.
Great work!!
I suggest that the data be sorted by average temperature in a month. I believe that would show a strong effect. When the average temperature is between 60 and 70F, most residences will not be heating or cooling and I would expect the UHI to be less by a factor of 2. It would also create 12 times more data and give that much more resolution and confidence in the results.
First of all, good work.
I have a question. Why adjust temperatures to sea level? The relationship is often not linear. I live in Reno NV and ski at lake tahoe resorts. There are often inversion layers here which leaves Reno colder than the higher ski resorts. Why add another adjustment, which can add error, if temps are normalized locally?
scienceofdoom said: Even if we adjust for population maybe someone then does another study on rural temperature stations and finds that there is a vegetation growth effect due to temperature so we have to then correct the temperature in stations by a factor relating to temperature! and proximity of nearby vegetation..
Or even, a temperature effect due to vegetation.
New York City Temperature and Vegetation
If I read this right you are saying that the greatest rate of increase in temperature (due to UHI) will be at sites where there is a development at a site with a existing low population density. In that case, if we are looking at changes of temperature over time, urban weather stations will be the most stable (i.e. have the least changes due to urban development) over time. Is this the correct conclusion.