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.
But if you look at the USHCN data from Missouri and Kansas over the last 114 years, plotting against current population, you get a log relationship to population, which would confirm, in a different way, the relationships discussed in the post.
there are problems here.
how about a temp reading in central park New york city? pop density is high!
sprawl might have a higher temp because of malls, parking lots etc.
Anthony found one at an air port that move from one end (UHI city) to the other end (rual ) and changed 5 degrees ? that is just a mile or so.
mods feel free to add the link to WUWT on that moving station.
Tim L.
Has anyone taken a look at this yet? Is it new? Reactions?
http://www1.ncdc.noaa.gov/pub/data/ushcn/v2/monthly/menne-etal2010.pdf
REPLY: Yes. See http://wattsupwiththat.com/2010/01/27/rumours-of-my-death-have-been-greatly-exaggerated/
Pamela Gray (18:25:49) : “By the way, I am still home sick with a roaring sinus infection. So please pardon my drugged up typing and lack of attention to grammar.”
Pam, I’m sitting here LMAO. I’ve been reading your posts and thinking, golly, this sure doesn’t sound like Pam. I wonder if she’s drunk…
Go to bed! And don’t reread your posts tomorrow.
Dave F (20:12:58) : “…If you put the measuring device in an area with no population, but in the middle of a heat island, the effect is still there. Likewise, even in densely populated areas, it is possible to have a well sited thermometer, is it not?”
Possibly, Dave, but it’s considered statistically unlikely.
There is one underlying assumption in previous studies that needs to be
questioned.
I’ll take GISS as an example. Rural was determined by a population bin
of 5K for the nearest city. IN the US Nighlights was used and nightlights was supposed to be a proxy for population density.
The assumption is this. A site that is rural today ( say pop 4K) has always been
rural all the way back to 1880. But if a population growth from 100 ( say in 1880) to 4K today imposes warming then the assumption of rural today = rural yesterday is clearly not adequate.
A very nice piece of work. It would be useful to see some “applications” – estimates how much UHI particular surface weather stations have suffered from.
I don’t think this analysis covers all UHI increases related to population growth.
many cities grow in overall population by increasing their size but not by changing the population per square kilometer that much.
this increase in size should also increase the UHI, but I don’t see this effect represented here.
Is Dr Spencers study another proof of negative feedback?
The initial UHI from pristine land make a difference but if “adding more UHI” is the result not linear. That is negative feedback.
Ivan (19:46:46) :
If this work stands to scrutiny, its nice to have a peer reviewed paper on Urban Heat Effect, Ivan.
If the Long paper was Peer Reviewed and stands to scrutiny too, and published in a paper, then it would be two papers, which together indicates what the warming has been in the US.
Then you have Dr. Pinker’s paper that says something about how much Surface Solar irradiance has varied from 1983 to 2001.
Its the revenge by proper Science, you know.
O just had a funny thought.
Dr. Mann likes to do reconstructions of past temperature based on tree rings.
Since we dont have population density information for all historical sites, I wonder if we could get him to reconstruct the historical population of various sites where we dont have information from those places that we do have information.
“sphaerica (15:12:17) :
Correlation is not causation.
Can anyone name a mechanism that would cause less densely populated areas to warm more?”
Yeah…high-density areas are already “warmed” by UHI. Low-density areas are not. It’s harder to add more UHI to an area that already has higher levels of UHI.
Roy Spencer (13:20:43) :
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).
The rapid UHI change in the low population factor may be due to something less high-tech, like land use changes e.g. vegetation clearance reducing cooling transpiration and exposing the earth to drying and greater heat gain during the day and slower heat release at night.
Ivan – “That’s all fine and good. There is only one problem – what is the purpose of the entire exercise?”
This is science. Look at problems/issues from different points of reference.
Heck, there are still people to this day trying to find holes in E=MC2.
jorgekafkazar (22:02:43) :
Possibly, Dave, but it’s considered statistically unlikely.
Sounds like an assumption to me, that the station would be sited badly just because it is in an area with denser population. Perhaps Dr. Spencer and Anthony can mesh their work together here to show that population density correlates with improper siting?
HIDE THE CODE HIDE THE CODE
Heck, there are still people to this day trying to find holes in E=MC2.
HIDE THE CODE HIDE THE CODE
I found the shape of the first two cumulative graphs interesting because of the big change in slope around 2,000 persons per km2. This level of population is far below the minimum urban threshold of roughly 5,000 per km2. (See graphs above that use 10,307 and 11,031 data points. )
Still, 2000 per km2 is a fairly high density for rural areas. Uninhabited areas have populations below 5 persons per km2. In 19th-century Ohio, a typical figure for rural population density was about 15 persons per km2, before the great migration of the US population to the cities. At a density of around 2,000 per km2, we are looking * at least * two kinds of populations: outer suburban (countries like the USA) and rural (areas like Java, Indonesia). The density could be explained by paddy rice cultivation in the tropics or suburban development anywhere in the world.
Once we get to Dr Spencer’s last graph with only 724 data points, we may be looking at a different phenomenon. The leftmost data point may indicate a temperature-related causative factor that makes the location undesirable for settlement. So what might be interpreted as a biased temperature reading caused by human settlement might better be interpreted with causation going in the opposite direction. The same unknown factor that makes the place unsuitable for settlement causes a real temperature difference observed. To illustrate: maybe the data point location is a volcano or a salt pan.
Dr Spenser’s study could be extended to relate temperatures obtained from satellite measurement to population of one-km2 cells with over 10,000 population. This would pick up cities like Chicago, London, Hamburg but probably miss many small cites. What would be of interest would be how high the dome of the urban heat island effect extends into the troposphere.
Dr Spencer’s model has the great merit that it is productive of a whole line of enquiry. I trust that he is as excited about the potential as many of the readers of this blog.
As other people here have indicated, my guess is that low population density areas are very sensitive to microclimate issues (tarmac v. grass, nearby trees and buildings). Vegetation is cooled by transpiration but paved roads are basically night storage radiators. It’s now known that plants actively control their temperate because of the reaction rates of photosynthesis. Temperatures inside leaves can be tens of degrees cooler than the surrounding air.
I would like to see a similar comparison done using energy use per km2 to see how the results compare with population densities. I don’t think there is a single right answer to UHI adjustments but Roy’s approach looks promising. Climategate means that GISS and CRU have lost their correlated duopoly on estimating temperature change and their approach will have to compete with others now.
Mosher — Since we dont have population density information for all historical sites, I wonder if we could get him to reconstruct the historical population of various sites where we dont have information from those places that we do have information.
Searching for Atlantis? Why rely on archaeological data and boring fieldwork? Using new climatology-developed methods that obviate the need for such trifles, you too can locate Atlantis in the comfort of your living room or office.
AlanG (23:53:08) :
Well this much is clear the controversy brought a lot of interesting new people into the party.. and some cool approaches
Thanks for an interesting post, Roy, and for providing a better way of correcting for UHI effects which looks ‘do-able’.
Apologies if already mentioned, but could the effect be different in countries like India and China which are undergoing rapid industrialisation? Big changes to infrastructure and energy use with low population growth could bias the result.
Yes GDP. Several posters have mentioned this as a potential correlate. A simple function such as Population Density x area GDP would create this weighting: High POP/High GDP > High POP/Low GDP and Low POP/High GDP > Low POP/Low GDP.
One problem I can see is that there are changes to an area that do not depend on population. A small village with 5 houses surrounded by dense forrest in 1920 could change to a small village with 5 houses surrounded by corn fields in 2010. This would have a profound influence on the measured temperature.
Wren (19:23:31) : You wrote:
Why do you think Dr. Spencer isn’t implying that warming over time could be greatest in the rural areas?
Isn’t that what he meant when he said –
“Significantly, this means that monitoring long-term warming at more rural stations could have greater spurious warming than monitoring in the cities.”
All cities were rural at one time or another. The UHI effect is greatest when rural stations transition to urban stations due to growth and development. Once that has taken place, additional growth and development adds an ever diminishing UHI effect. Spencer’s statement simply acknowledges that rural stations have a greater potential for additional, spurious warming from the UHI effect than cities do. That potential, however, is never realized if there is no development at the rural site. The potential in the cities has already been realized and must be calculated into the temperature record to discover any true climate change.
Since we are looking for temperature trends over the last 100 years, the potential for spurious warming in the rural areas is not relevant. It is just potential. It may never actually be realized. The actual warming due to UHI is relevant and exists in the towns and cities, not in the persistent rural areas.
Just a rough comparison here of estimates, so nothing solid, but globally, we have this:
http://data.giss.nasa.gov/cgi-bin/gistemp/do_nmap.py?year_last=2010&month_last=1&sat=4&sst=3&type=anoms&mean_gen=01&year1=2010&year2=2010&base1=1951&base2=1980&radius=1200&pol=reg
vs.
http://www.populationaction.org/Publications/Reports/Mapping_the_Future_of_World_Population/Medium_size_bitmap_image.jpg
I don’t have the software to overlay them and see how well they correlate, but Canada is an issue right off the bat.