JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 116, D14113, 12 PP., 2011
Observed surface warming induced by urbanization in east China
- The rapid urbanization has significant impacts on temperature over east China
- A new method was developed to dynamically classify urban and rural stations
- Comparison of the trends of UHI effects by using OMR and UMR approaches
Xuchao Yang, Shanghai Typhoon Institute of China Meteorological Administration, Shanghai, China Institute of Meteorological Sciences, Zhejiang Meteorological Bureau, Hangzhou, China Yiling Hou, Shanghai Climate Center, Shanghai, China, Baode Chen, Shanghai Typhoon Institute of China Meteorological Administration, Shanghai, China
Monthly mean surface air temperature data from 463 meteorological stations, including those from the 1981–2007 ordinary and national basic reference surface stations in east China and from the National Centers for Environmental Prediction and National Center for Atmospheric Research (NCEP/NCAR) Reanalysis, are used to investigate the effect of rapid urbanization on temperature change.
These stations are dynamically classified into six categories, namely, metropolis, large city, medium-sized city, small city, suburban, and rural, using satellite-measured nighttime light imagery and population census data. Both observation minus reanalysis (OMR) and urban minus rural (UMR) methods are utilized to detect surface air temperature change induced by urbanization. With objective and dynamic station classification, the observed and reanalyzed temperature changes over rural areas show good agreement, indicating that the reanalysis can effectively capture regional rural temperature trends. The trends of urban heat island (UHI) effects, determined using OMR and UMR approaches, are generally consistent and indicate that rapid urbanization has a significant influence on surface warming over east China. Overall, UHI effects contribute 24.2% to regional average warming trends. The strongest effect of urbanization on annual mean surface air temperature trends occurs over the metropolis and large city stations, with corresponding contributions of about 44% and 35% to total warming, respectively. The UHI trends are 0.398°C and 0.26°C decade−1. The most substantial UHI effect occurred after the early 2000s, implying a significant effect of rapid urbanization on surface air temperature change during this period.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
@ur momisugly Mark Wilson – jumping from this study to the globe writ large isn’t a valid leap.
“but even then 75.8% of the warming is from other sources”
Even the IPCC admits that changes in solar intensity account for 20 to 30% of the warming.
Then there is the all but confirmed cosmic ray/cloud formation factor.
There there is the documented problems with the ground based temperature network causing a warming bias.
All in all, not much of the signal is left for CO2 to be responsible for.
I think even the hardened warmista accept that UHI exists.
The problem is (just like alleged CO2 based climate sensitivity) how much is it actual effect? The likes of Hadley/UEA and Mr Jones are supposed to have ‘adjusted’ for UHI – but have they?, and is the adjustment enough or realistic, etc, etc. Without raw data and adjusted data, preferably with the adjustment notes – how the flip does anyone here know what has been done?
@ur momisugly Mark Wilson – you are going random math again.
I don’t think you’ll find an mainstream climate scientist claiming that CO2 is responsible for 100% of the warming. What you will find is that it is an important forcer that gets augmented by other GHG.
It is interesting, China has a national vested interest in debunking AGW. With their attempt to modernize their economy and improve their standard of living, they are massively increasing their fossil fuel generating capacity which creates very large amounts of C02. While I think this study is correct, I would be skeptical about the size of the actual numbers.
In the West, the politics encourage exaggeration of the negative effects of CO2 while in China, the politics may encourage the opposite, understating the effect of CO2.
At least it provides some peer reviewed science that isn’t filtered by the pro AGW political correctness.
Kev-in-Uk,
I think that there are interesting leads. Böhm et al. 2001 is a great door. To pass with care.
Bystander says: “I don’t think you’ll find an mainstream climate scientist claiming that CO2 is responsible for 100% of the warming.”
IPPC Net anthropogenic forcing= 1.6 W/m^2, forcing from CO2 1.66W/m^2, total IPCC forcing adding in the sun’s brightness (they say this is the only effect they need to include, everything else is unproven, according to them), 1.72 W/m^2. Therefore, “mainstream” science says the warming is 96.5% from CO2. Pretty dang close to saying that’s all that really matters.
http://i23.photobucket.com/albums/b370/gatemaster99/forcings.png
phi says:
July 28, 2011 at 12:52 pm
I am not sure of your point – I just quickly googled Bohm et al 2001 and it would appear from the abstract that they are suggesting a mean temp trend increase over and above what CRU produce. Presumably, in the Alps, this would not be due to UHI – I will have to look for a direct link to the paper and it’s graphs. Nevertheless, I am still perturbed by the references in the abstract to various statistical analysis and homogenisation……
I am fairly convinced (as a geologist – who deals with ‘spatial’ awareness!) that spatial determination and averaging is a very subjective and locally specific problem – meaning that ‘local’ weather/conditions are reflected by local topography, prevailing winds, station siting, etc, etc. My own simple analysis of the UK station data (off the metoffoce site) shows an amazing difference from west coast compared to east coast stations (this is totally to be expected BTW) – Ergo, a direct ‘averaging’ of the two would not necessarily be representative. I don’t really consider that ‘gridded’ data are valid in any significantly varying topographical area. A couple of stations in the sahara desert are more likely to be similar than a couple of stations either side of the english channel for example – yet the channel only spans 20 miles!
I am not overly familiar with all the statistical methods – and I really don’t have the inclination to learn it in detail at my age! – but on the simple points regarding UHI, a station within an urban sprawl must logically be affected by that urbanisation. IMO – Rural stations should be ‘king’ in the real data world, because adjustments for UHI, if and when applied, are highly likely to be subjective. Taking a very basic and simplistic view of averaging say, 2 rural stations with say 6 urban stations to give a gridded average must introduce a bias. Personally I’d suggest ignoring urban stations altogether (or at least use them only for indicative purposes). I think, though I accept I may be challenged here – that basic and trusted data (i.e known to be unaffected by UHI) should always be the ‘preferred’ data – that fact that some airport or other has reams of data – does not automatically make that data any better (probably worse in some sense!) than some rural station manned by a couple of old codgers with thermometers! In terms of actual individual station data, I would prefer to see each station analysed seperately – if you like, with each plot overlaid onto a base graph – then you would see which stations were grossly affected by UHI as those stations would show markedly different trends – whereas simple UHI ‘adjustment’ and then averaging them altogether is likely to introduce a bias in my opinion (could be either way depending on how good the UHI adjustment is guessed!) and so what value would one put on the resultant ‘combined’ average plot?. I am reasonably sure that if anyone actually plotted all the UK stations on a graph, one would see a range of trends, with most/all the higher trends being in the urban stations – and lower trends in the rural stations – in effect, I wouldn’t be surprised to see two distinct ‘bands’ emerging, and logically THAT, not direct comparison between an urban station and its nearest ‘rural’ neighbour – would provide a realistic indication as to the likely UHI effect? Has anyone ever done this?
Verity Jones says: “[Unfortunately] claims of the contamination of ‘global average temperature’ by UHI are readily countered by those who point out that plotting all the rural stations worldwide separately … gives the same shape of graph and the same trend as the complete set of urban and rural stations.”
In Australia, Canberra Airport is officially classified as Rural.
http://members.westnet.com.au/jonas1/GoogleEarth_70014_CanberraAirport.jpg
as are many other obviously non-rural sites.
Kev -in-Uk asks if anyone has ever done a direct comparison between urban and rural sites. From memory one was done by a schoolboy and his dad in the USA – posted on WUWT – but I don’t have the link. I did a more general analysis for Australia http://wattsupwiththat.com/2011/02/21/an-analysis-of-australian-rural-vs-non-rural-stations-temperature-trends/ which indicated that “Over those periods in which there were increasing temperatures, the rural stations appear to have warmed at about 60-70% of the warming rate of the non-rural stations.“.
Kev-in-Uk says:
July 28, 2011 at 2:13 pm
“…I am fairly convinced (as a geologist – who deals with ‘spatial’ awareness!) that spatial determination and averaging is a very subjective and locally specific problem – meaning that ‘local’ weather/conditions are reflected by local topography, prevailing winds, station siting, etc, etc..”
/////////////////////////////////////////////////////////////////////////////////////////////////////////////
Absolutely. I live in Spain in the foothills overlooking the Med. I am probably at about 150 – 200m above sea level and as the crow flies less than 1km from the sea. Where I live there is a horseshoe mountain range. Within 15 minute drive, down towards the coast, or along the valley. or up into the mountains, or just one or other side of the mountain range (ie., either East or West of where I live) one bay can be rather windy and cool, the next bay almost dead calm and warm, I witness a substantial number of different temperatures with a huge range in temperature. The whole vacinity has its own micro climates such that the idea that my region has the temperature as indicated by the station at the nearest airport some 60km away, is rediculous. Of course, this may be more important if one were considering ascertaining in absolute terms the global average temperature. To get a proper handle on the global average temperature, one would have to increase coverage probably a billion or more fold.
This point highlights the futility of the ‘global’ approach. There is no such thing as ‘global’ warming. Some areas may be warming, others staying broadly as they are, and some places even cooling. Some places may be experiencing more rain, otheres broadly the same amount of rain as usual, and others less rain. Sea level rise is a factor for some countries not for others. The approach of averaging everything just clouds/obscures what is going on.
Global warming/climate change is a local issue not a global issue. Each country (or at any rate relatively small region) needs to analyse their own data and evaluate how global warming/climate change will affect them.
The only reason that this is being presented as a global problem is a political one to garner control on an international basis. The only possible consequence which could have widespread effect is sea level rise (although that will not affect countries which have no sea coast and even countries that dio have sea coast lines will be affected differently) and we all know that sea level rise is happening only slowly and rather steadily such that there are no foreseeable imminent disaster problems with that.
If looked upon on an individual basis, each country can plan what it needs to do to adapt to such changes as can (with a reasonable degree of certaintyj be predicted. Some countries may need no adaption and for them any warming may merely bring positive results. I for one consider that the UK would benefit greatly if there was to be warming of 2 or 3 degC. It may be that some countries will be seriously affected and may not have the wherewithall to adapt and in which case then consideration could be given to the desirability of ‘rich countries’ giving that country aid. That is the only political issue that should arise out of this perceived drama. .
“The definition of ‘rural’ for GIStemp is a population of >10,000, (also airports are frequently classed as rural) ”
Wrong.
that changed in 2010. Nightlights are now used.
I’ll have to check which Nighlights these guys used. One hopes they checked with the PI and did not use the nighlights that are unsuitable for this purpose.
Kev-in-Uk,
Böhm paper is available at: http://onlinelibrary.wiley.com/doi/10.1002/joc.689/pdf
In my opinion the whole problem of homogenization comes down to a ratchet mechanism, see this comment:
http://rankexploits.com/musings/2011/analyzing-surface-stations-part-1/#comment-76476
There are other ways to approach the total effect of perturbations, I described some on the same thread.
Verity Jones says:
July 28, 2011 at 6:02 am
“Much as I hesitate to say that scientists should be in any way subjective, there is a need to look in detail at individual stations…”
You have correctly identified the need for vetting ALL records to be used in climatic analyses. This can be done very objectively with proper signal analysis tools that are strangely absent in ” climate science.” The paramount requirement is for the maintenance of an unchanged environment at the sation site and a consistent datum level throughout the duration of the record. The fact of the matter is that in the GHCN data base so-called “rural” records are frequently corrupted by land use changes and abrupt datum-level changes. That is what blind, crank-turning computational exercises on an unrepresentative, faulty database miss entirely. As this duration-limited study from a region very poorly represented in the GHCN data base clearly shows, UHI is a very significant factor in records from population centers of considerable size. That significance only increases with increased duration.
Mike Jonas says:
July 28, 2011 at 2:44 pm
I am well aware of the valuable work done by Anthony et al – that wasn’t really my point. I was wondering if anyone had literally simultaneously obtained and plotted on a graph many rural versus urban station data to see if they fall into two distinct ‘bands’…. (yeah, I know it would look a little cluttered – but if there was some way of presenting it – I feel it would demonstrate unequivically the difference in temp trends between urban and rural stations – I’ve see it done for a few individual examples, but not for a whole dataset, e.g giss or hadcrut3)
phi says:
July 28, 2011 at 3:19 pm
thanks – I will try and get a few hours together to read through…(unlikely as too much work – but I will try, honest!) – but I hope you can understand my (probably badly presented) suggestion, that a graphical representation of essentially raw data from various stations superimposed together should easily identify the trend differences between urban and rural stations. (This being infinitely more preferable that a direct averaging of rural/urban stations)
Even if it just shows a concentration of urban stations in a different graphical position to the rural stations, this would give an empirical indication of the UHI value. I just feel this provides a more realistic assessment of the actual levels of UHI in the overall sense, yet I have never seen such a graph…
I am thinking of a way to plot it in such a way as to be indicative – perhaps the simplist way would be to set the opaqueness and/or colour of the plotted station line according to its urban/rural status. e.g. urban sites could be red hued and rural sites more blue hued – whats the betting that the redsites would be grouped together etc – and the blue hued sites would show a lower rising trend than the urban sites?
In my view, those who insist that the UHI effect is negligible (or small) are loony.
For instance, the temperature of Tokyo has risen by more than 3 degC over 100 years:
http://data.giss.nasa.gov/cgi-bin/gistemp/gistemp_station.py?id=210476620003&data_set=1&num_neighbors=1
Undoubtedly this is due to the ever-intensifying UHI effects through energy consumption and motorization in the area surrounding the thermometer, located in a courtyard of our MET office, which sits on the very center of metropolis Tokyo.
richard verney says:
July 28, 2011 at 2:39 am
“I know that we will soon see a number of the ‘usual’ posts suggesting that since we are looking at anomalies and not absolute temperatures, the overall trend is unaffected and that the statistic of large numbers cancels these anomalies out. Personally, I consider such arguments to be bunkum although I do accept that once a temperature measuring station has been fully swamped (saturated) by the effects of urbanisation, thereafter, but only thereafter, the anomalies will become little affected by UHI.”
I applaud your excellent post. You are way too kind to the Warmista. No temperature station has ever been fully swamped by the effects of urbanization. Warmista must learn that as the Gates of Hell are widened so the temperatures are raised proportionately. Urbanization is not a one-shot proposition like clear cutting a forest or damming a river. Urbanization is ongoing. As your truck stop succeeds, it will get busier and larger indefinitely and it will attract other truck stops. (We all know the principle that if you want to open a new book store you should open it next door to the most successful existing bookstore.)
Anomalies are just tricks for avoiding actual empirical observations and better hiding the pea.
I have been putting some of my job skills to work on a personal project. Some of you may find my “Urban Heat Island” web site of interest. If there is a city in the USA (or the world) you would like a UHI web page created for, send me an e-mail. I will download and process the landsat data to create a web page similar to the ones shown at the web site below. This prcess takes about 2 to 3 hours, so I would only be able to do a few of them. I plan on doing one or two per week (at least one city per state) over the course of the next year.
http://www.mascookin.com/Mas_Cookin/Urban_Heat/Urban_Heat.html
midimikeh@aol.com
Steven Mosher,
I wouldn’t worry about whether they have the correct nightlights or not. Just the fact that they found a number of stations that have trends about .25c more than a bunc of others shows that your mantra about it not mattering is unsupported in other than the contaminated, overadjusted surface records.
You’re not kidding! The heat index here in Shanghai today was 128 deg F/53 deg C (98 deg F/31 deg C and 65% RH). Same tomorrow as well – but I’ll be out in rural Jiangsu, visiting some friends a way out from Suzhou (about 70 km NW of Shanghai) where they’re around 98 deg F/36 deg C heat index (88 deg F/31 deg C and 65% RH).
And I’m in the Shanghai suburbs (Minhang district). Downtown Shanghai is probably 4-5 deg F/2 deg C warmer.
Kuhnkat of course everybody was interested when it was hansen using the wrong nightlights.
and everyone was skeptical of using nightlights in china and india.
funny how the skepticism goes away.
.25C? absolutely you will see that. On several occassions I’ve pointed to sites with MORE that .25c of UHI. MORE than that is easily found. That’s NEVER been the question.
the question has NEVER been can you find UHI. Of course you can. You can find great examples of it. I’ve even pointed you guys at sites that map it down to the block.
The question has ALWAYS been this:
If you look at the sites ACTUALLY USED to calculate the global mean.
Can you find a SIGNIFICANT difference between an ALL RURAL
selection and an ALL URBAN selection.
What i’ve said repeatedly is I expect the difference to be between the values of Jones
And mcKittrick: Jones at .05 and mcKittrick at ,3C
If you want an estimate I will give the same one I have given since 2007. .15C which is
magically between those two estimates.
If the difference is ,15C do i expect to find areas where its greater? YOU BET I DO
Do I expect to find areas where its less? YOU Bet.
Was I surprised when I found urban areas that COOL? a little.
Kev-in-Uk,
In my opinion, the distinction between rural and urban stations is not the solution but the problem. Disturbances from urbanization are not limited to this effect which concerns only the major aglommeration. All studies are based on this distinction that leads to an impasse. The treatment of UHI by Giss succeeds to invent for instance the notion of Urban Cooling Island. Odd.
A couple of people have asked about other Rural – Urban comparisons, particularly the “Peter and his Dad” comparison covered here on WUWT in 2009. In that comparison, Peter & Dad extracted temperature data for 28 location ‘pairs’, one Urban and one Rural, spread around the US for years 1900 -2006. http://wattsupwiththat.com/2009/12/09/picking-out-the-uhi-in-global-temperature-records-so-easy-a-6th-grader-can-do-it/
I repeated the data extraction and cleanup as described in that article.
Here is what came from the analysis:
Rural 0.57/century C
Urban 1.07/century C
More results comparing Rural and Urban temperatures at the 28 paired sites, all in DegC:
……………….Rural ……..Urban
Minimum …….5.8………….7.2
Maximum ….22.4…………23.3
Average……..13.1…………14.3
As others have pointed out, the slopes of temperature change per time are probably more important than the averages, since there may be true siting differences in some pairs. Additional poking at the data showed no particular trends in the slopes based on warmer or colder areas in the paired sets.
*There are 6 of 28 (~20%) Rural negative slopes (cooling).
*None of the Urban temperature slopes are negative/cooling.
*Five of the Rural slopes are higher than the corresponding Urban slope (almost 20% of pairs), which also means that those five Urban slopes are lower temperature change than the corresponding Rural slope.
for what it’s worth.
Mosher, for the last damn time, McKitrick’s work was NOT a study of UHI! You can’t say that you expect the urban difference to be between “Jones and McKitrick” Because A) McKitrick has not estimated the effect of UHI, rather all socioeconomic biases in the data. and B) There is ZERO reason for you to expect it to be “inbetween” the results. McKitrick has the ONLY results in climatology that come anywhere close to meeting the standards of statistical significance in particle physics, p values literally smaller than any others in this entire field…and you don’t believe them. You aren’t impressed by them. Bah. What would McKitrick have to do to make his model “impress” you? Every objection people have made, he said, “Okay, well when I take your hypothesis as to why I’m wrong into account, it doesn’t change the result. Now do you believe it?” They inevitably say “No!” Heck they’ll even admit that, if he takes their concern into account (ie before he actually does) that they still won’t believe the result – they admit that their faith is unshakeable. Your faith is probably equally unshakeable. It’s “inbetween” Jones and McKitrick…for no reason other than the fallacy of the golden mean.