A couple of days ago there was a guest post from Russ Steele citing a California study “Feeling the Heat” on global warming that just didn’t seem to add up. One of the stations cited as having climate change related warming was Reno, NV. So, I decided to do a field experiment to test this. The results show clearly that UHI exists in Reno.
Here is what Russ wrote a couple fo days ago:
Feeling the Heat was published by Environment California a non-profit group a few weeks ago, claiming 2007 was the tenth warmest year on record and that the mountain west was experiencing above-average temperatures. Full report here: Download feeling_the_heat_ca.pdf One of the examples given for the high western temperatures was Reno Nevada with a average temperature of 55.3 degrees in 2007, four degrees higher than the 30 years average temperatures from 1971 to 2000.
…Up front in the EC report the author dispatches UHI as having any influence on the climate change, citing studies by Easterling, PD Jones and Parker…
Well I decided to test this myself tonight, since I’m driving through Reno on my return home, I arranged an overnight stay. With me is my NIST calibrated data logger, NIST Calibrated temperature probe, a vehicle mounted Gill IR shield, my laptop computer, and my trusty vehicle. See my previous post “Road Kit”
I chose Virginia Street as the transect route, since it remains relatively straight, level, and crosses all of Reno, including the built up southern suburbs and downtown. It is the original “main street” for Reno.
Here is the result of my South to North transect driving Virgina Street overlaid on a Google Earth image oriented to match the timeline of the transect:
Click for larger image
The weather tonight was perfect. Light winds, clear skies.
Here is the data from the Reno airport ASOS, which also happens to be a USHCN climate station:
Time Temp Dew RH Wind Wind Vis WX Sea Level Altimeter Station
Point Dir Speed Pressure Setting Pressure
(PDT) (f) (f) (%) (mph) (miles) (mb) (inches) (inches)
1:55 am 44 25 47 CALM 10.00 CLR 1023.0 30.30 25.788
12:55 am 48 24 39 CALM 10.00 CLR 1023.4 30.31 25.797
11:55 pm 51 23 33 WSW 3 10.00 CLR 1023.7 30.31 25.797
10:55 pm 54 23 30 S 6 10.00 CLR 1024.1 30.32 25.805
For those interested, I have the raw source data from my datalogger in CSV form for the South to North Reno transect here. (PDF)
Note the placement of the airport, which has it’s ASOS weather station used in many climate studies essentially in the north end middle of the airport. The Reno UHI bubble does extend into this area.
Click for a larger image
I also did a reverse transect, driving the same route in reverse immediately. Plus a route near the airport. I’ll have more tomorrow, its 2AM and I’m tired.
UPDATE:
Jeff Id inquired in comments “how is it mounted to the car?” Here is the answer:
The temperature sensor (inside the Gill IR shield) mounted on the vehicle using an improvised window mount.
Also, the time of night that I made the transect (11:15PM to 11:39PM) allowed me to maintain a nearly constant speed during the transect due to the lack of traffic. Plus Virginia street has stoplights set for all green unless there is cross traffic. I was fortunate to have to stop only once during the entire drive, and that was in the downtown area. I kept an eye on the temperature reading during the stop, and no change was recorded.
I’ll have a complete post in the next day, still catching up from my trip.
Weather Underground offers a great way to measure UHI real time in many cities around the world. This view of private weather stations in Denver on a still, clear winter morning around dawn, frequently shows the outlying areas (of equal elevation) 15-30 degrees colder than downtown.
http://www.wunderground.com/wundermap/?zip=80002&magic=3&wmo=99999
Lief Svalgaard:-)
Thanks for that. What I was aiming at was that there were several predictions in 2006 that it was about to start back then, & it was supposed to be fast & furious, & worse than its predecessor Cycle 23! Yet all the evidence seen here & on other sites suggest that it is stuttering along a bit like an old car being churned into life with a flat battery & a grinding starter motor, never quite firing up but only managing an almost got there! I conceded that it may have officially started, but no fireworks as yet with only 14-18 months or so to go!
However, I willing defer to your good self as an expert in this subject. I’m an engineer not a solar physicist, but having witnessed 30 years of adult life of claims, predictions, projections, assessments, of all sorts of things in life, many based/reliant on the latest technology, & usually coming to nothing much in the end. I suppose you could call me sceptical about many things!
@Steve Keohane (09:12:41) :
I hypothesize the following:
What is important about siting bias is that it is not likely to be a fixed distortion of the data (ie, the biased station always reports 5 degree higher no matter what is the actual temperature).
The bias should be a percentage of the ambient (and actual) temperature, so that as temperature goes up, the bias grows. This percentage may also be variable. This behavior would make a modest positive trend from the PDO or regional UHI and magnify the trend, not just the temperature. It could also amplify a cooling trend, showing the surface cooling faster than reality.
So a station that shows a 5 degree bias now may have only shown a 3 degree bias in 1975 when the planet was slightly cooler.
Just another reason I trust the satellite data over the surface data.
ared is missing the boat entirely:
“Suppose”?? FYI, the U.S. has added ~100 million people since the late ’70’s, when the population was around 200 million. Now the U.S. population is over 305 million. Almost all of that increase has occurred in urban areas.
In his statement above, ared presumes that no population increase has occurred in Reno over the past 50 years. Those of us who visit Reno know that its population has exploded during that time. The recorded rise in urban temperatures track that population increase.
It is not clear if ared is deliberately trying to misrepresent the UHI situation, or if he is simply ignorant of the problem of corrupted data. If it is the latter, clicking on the SurfaceStations link on the upper right section of this page would begin an extremely educational process. But if the former is the case, ared had best be prepared to be set straight by the majority of folks here who understand the “adjustment” shenanigans going on at the UN/IPCC, the NOAA, GISS, etc.
ared,
You may feel like quite a few folks are ganging up on you, and I am sorry if that is the impression. Nevertheless, this blog site is full of people who have done a lot of research on UHI, what adjustments are made in GISS for UHI, what the IPCC says about UHI, and how the IPCC has responded to UHI challenges.
Although the IPCC acknowledges the existence of UHI, it dismisses concerns about UHI in two ways. First, it says UHI is small and refers to two studies (questionable studies in my mind) that say UHI is small while ignoring other studies that say that UHI is large. Second, it says that the UHI has already been accounted for in its selected temperature data sets, but IPCC authors ignore convincing studies that show UHI is vastly underestimated and not adequately handled. For example, if UHI is adequately handled, then temperature trends should not depend upon the economic development around the themometers — but temperature trends are extremely highly dependent upon the economic development.
There might be an interesting question of the burden of proof. You seem to suggest that it is the responsibility of skeptics to prove that UHI is biasing the trends. (And many skeptics have whole-heartedly accepted the challenge and believe that they have done just that.) Meanwhile, it seems just as legitimate — or even more legitimate — for AGW pessimists to have the responsibility prove that UHI is not biasing the trends. After all, AGW pessimists propose massive and costly measures which just in their infancy have proven to be laden with immense negative unintended consequences.
(To be sure, some AGW pessimists have argued that it is up to the skeptics to prove that altering the atmosphere’s chemistry will not have dire consequences for the earth. That could be an interesting question for exploration in another posting.)
Many feel that a legitimate approach to examine the UHI effect is to look at satellite data. This data trend shows that temperarture trends are consistent with PDO phases (and major volcanoes) — the satellite data trend is not consistent with a dominating influence of monotonic increases in CO2.
“Come on, guys. No one denies UHI is real.”
Yes, they do…
UHI effects are well known and accounted for.
If the “it’s just UHI” folks were correct, then the global hotspots would by and large be over areas with lots of development. They are not.
Interesting work Anthony.
It would be interesting to plot speed traveled (or time spent at stoplights/stopsigns per 1/4-mile) as an additional parameter on the graphic. If I remember correctly from my past trips to Reno, the origin of your transect is relatively open road with few stopsigns, the latter portion of your transect has stoplights and idling cars at nearly every block (and tall casinos/hotels on the side). I wonder what is the effect of travel-speed and the heat of your (and neighboring) automobiles on the temps?
As a trivial sidenote, to all contributors to this Blogg (I’ve held my tongue on so many previous posts)..”data” are plural, “datum” is singular…”data are”, “datum is”.
Again, interesting project Anthony.
John D.
The important thing about this method is that it shows that you can get real UHI numbers. You can use these numbers to _accurately_ adjust the surface stations for UHI. This method needs to be carefully repeated all over the country/world.
I wonder if it would be useful if Google to added a good temp sensor to their StreetView cars.
One more note, previous commenters have mentioned the possible affect of elevation on the transect itself (~500′ difference?). But perhaps as important is proximity to the base of Mt. Rose, which is very near the transect origin, and quite removed from the terminus. Mt. Rose is ~7,000 feet elevation and nightime downslope flow of cold air may contribute to the cooling of the transect origin just as UHI contributes to heating of the transect terminus? It would be interesting to repeat such a transect elsewhere in the Great Basin with similar topography without a City at it’s end. Just wondering. Complex world out there.
coby: If the “it’s just UHI” folks were correct… Straw man argument, coby. No one here says that “it’s just UHI”. Nice try.
Coby, Steven Mosher, it appears from the EC article that they’re using the red line from your graphs, not the fully-adjusted green line. If so, they’re counting UHI, big time. Did I miss something?
Barrow, Alaska. Population 4500. 70 thermometers. 4 year study.
http://www.geography.uc.edu/~kenhinke/uhi/Hinkel&Nelson_JGR-A_2007.pdf
This study is strictly concerning UHI. Microsite issues are not addressed.
How is it IPCC ignores studies like these, yet rely on creative mathematical iterations to wipe out UHI and poor siting effects?
Coby, your claim that UHI is “accounted for” is absurd and simply a scripted response typically found in pro-AGW propaganda blogs. There is no way without comprehensive studies such as above that UHI can be quantified.
Alan the Brit (09:26:16) :
Thanks for that. What I was aiming at was that there were several predictions in 2006 that it was about to start back then, & it was supposed to be fast & furious, & worse than its predecessor Cycle 23!
I know, but not everybody was predicting that. The forecasters with the best track record [such as it is] have been predicting a weak, ‘sputtering’, ‘lowest in a 100 years’ cycle 24, e.g. http://www.leif.org/research/Cycle%2024%20Smallest%20100%20years.pdf
coby (10:28:04) : said,
UHI effects are well known and accounted for.
If the “it’s just UHI” folks were correct, then the global hotspots would by and large be over areas with lots of development. They are not.
YES THEY ARE,
Urban heat island and population growth in Addis ababa.
Both minimum and maximum temperature trends examined together with urban increase during the period between the late 1960`s and 2000. The total population increase in the 18 years (1967-1984) was 739,581 and the annual mean maximum temperature in the same period became warmer by 1.7 degrees C. The annual mean temperature attained it`s peak in 2000, the urban population
was also at it`s highest in that year.
http://www.geo.uni.lodz.pl/~icuc5/text/P_6_11.pdf
UHI in Saskatchewan’s two largest cities, Saskatoon and Regina.
Nocturnal air temperatures in the city centres of Regina and Saskatoon are, on average, 3–4°C warmer than in the surrounding countryside; and under ideal nocturnal heat island conditions, urban-rural temperature contrasts, measured at the same time, can reach 6–8°C. In spite of the abundance of parks and open spaces in Regina and Saskatoon, both cities exhibit a discernible heat island influence. It is expected that heat islands of lesser intensity may be detectable in smaller cities and towns throughout Saskatchewan.
http://esask.uregina.ca/entry/urban_heat_islands.html
I suggest you read this and look at the thermal image in pic 4.
http://www.urbanheatislands.com/
Dee, I agree with your assesment. I was only refering to the temperature classifications (CRN) given to sitings made at surfacestations.org. That seems to me to be a conservative estimate as the CRN number is the minimum bias. Recalculating, I was in error, the average minimum bias is 3.5 F or 1.95 C. This has to account for some of the ‘warming’ we have seen.
Pretty Cool Pic there Anthony; how does one get to inherit one of these field kits so that the necessity of driving is put to some use ?
Now This: “j oshv (06:06:45) :
A simple question that perhaps Anthony can answer based on his meteorological experience.
Why don’t we calculate average surface temperature using the techniques meteorologists use to produce those nice temperature gradient maps (isobars? I am not sure of the exact term)?
Use the known temperature points to interpolate a temperature field for all lat/lons, and then sample that field at regular grid points. This corrects for UHI quite naturally, as the vast majority of grid points in a regular sampling grid would be rural. Sure, UHI would influence the overall average, but urbans areas would probably represent just a few percentage of the sample points. Lacking a good way to remove UHI from the temperature signal (other than ignoring urban areas altogether), this would seem to be the best way to handle it.
Now you have me bamboozled JoshV. You want to take some actual measured “known” temperature points; and then you want to “interpolate” (based on what scientific or mathematical justification?) to get a grid of made up “data” which we now pretend is real, and then you want to make up some more data for some more points and somehow you have a cascade of increasing information content ?? And from that you are going to compute a really accurate average. Simply wonderful; can you really do all this if say you only have two real starting temperatures ?
The whole problem with GISStemp and all lookalikes, is that the sampling process; which is supposed to be a gathering of actual real data from the real planet; is faulty.
You cannot increase information by interpolation, unless the function you are interpolating is bandlimited to that appropriate to your sampling rate.
Anthony’s little cross town jaunt points out the enormous amount of information that is just thrown away by the climate modellers.
I only have to watch any of the SF Bay area’s 6PM news weather to see that temperature fluctuations, are wild compared to the numbers that are reported as typical. You can ignore the highs and the lows if you wish, but the planet does not, and its various surfaces radiate in some way roughly proportional to either the fourth or fifth power of the absolute temperature (all the time). Fourth power if you are only interested in the total flux emitted, but fifth power if you also are interested in the spectral peak which is important for figuring out GHG absorptions.
It is easy to show, that if at some point the temperature goes through some sort of cyclic change (say daily) about a mean (integrates to zero about the mean), then the integral of the 4th (or 5th) power of the cyclic function, which would roughly correspond to total radiation emitted (4th power); that the total is ALWAYS larger than taking the 4th power of the average temp and using that to compute a radiation total for the cycle.
When you examine this for the annual change during a comlete solar orbit, or even on a 24 hour daily basis; not to mention the -90 to +60 c global temperature extremes(surface), the total emitted radiation is somewhat greater than if you just take the global mean temperature and use that as your total radiation baseline to apply “forcing” to.
I did put out a short essay on that to some people, but5 you can easily do the math for yourselves.
GCMs may be climate models; they just aren’t the models of any planet that is of interest to us. No planet I know of has the sun directly overhead at every point on the planet continuously throughout its whole yearly cycle, so that it uniformly illuminates the entire surface all the time at some constant insolation rate. (I have devised a ZEMAX model of a spherically enclosing sun with a surface that emits uniformly over a very restricted solid angle (0.5 degrees) from the normal at every point on its surface. From any point on the internal planet surface, you can look directly at the zenith and see a 1/2 degree angular diameter sun all the time.)
Hullo, Steve Goddard.
How do you like the rumors going ’round that we are the same person? (Or maybe we are both just mythical virtual constructs of El Reg?) #B^1
Further evidence: Steven Goddard = St. God?
For that matter, how many John Philip(s) ARE there . . . ?
Just wait till the Red King wakes up!
So, common then somebody, just when is Solar Cycle 24 going to start in earnest?
We’re still waiting for action from the House of C’mons.
Anthony,
Thanks for providing first hand, the chance to see science in action…it’s fascinating to witness things like this unfold.
Nobody beats the Rev!
The question is: does urbanisation affect trends over large periods of time for a significant number of stations.
Good question. The answer would appear to be “yes”.
See McKitrick and Michaels (2008) and LaDochy et al. (Dec. 2007).
Increase in trend [sic] is exaggerated by around a factor of two.
That’s how heat sinks work: If there is a warming trend, it is exaggerated. The converse is, of course, that a cooling trend is exaggerated, as well, as the heat sink effect “undoes” itself.
Also what appears to be happening is that suburban and exurban creep (not to mention microsite violation) are overtaking the rural stations. SHAP is obviously not accounting for this (SHAP, believe it or not, is a POSITIVE adjustment–no lie!) This causes an entirely spurious inflation of trends.
ared (07:47:32) :
“So in order to prove the IPCC wrong, you have to prove that UHI’s have affected trends, not just that they are real, …”
The Ipcc advanced the theory, therefore they have to prove it. The ipcc has to prove that uhi’s have not affected the trends and that anthropogenic emissions of CO2 have adversely affected the global environment in a way.
ared (08:27:31) :
“Dee, I know all this. I’m just making the point that knowing the size of one UHI at one moment in time does not prove the IPCC wrong. You need to know how UHI develop over time and what effect many of them have on long-term global trends to say something usefull about whether 0,05 per century is wrong or not.”
What data proves the Ipcc right?
ared (08:46:44) :
“stan, the IPCC doesn’t say that the UHI effect is small. It says that after correcting for the effect, the impact on global trends is small.”
The ipcc is forced to correct for the effect to prove their foregone conclusion.
all of this discussion about global warming/climate change is beside the point. The original al gore/ipcc attack was on anthropogenic co2. we have allowed the discussion to be diverted to global warming/climate change, which would occur if there was no human life on the planet. It is the political attack on the cycle of life with the intent to control, through taxation, our freedom, to live life to its fullest, that must be confronted. Why be defensive about the “straw man” of global anything. The attack by al gore/un is on the american way of life. we are too prosperus and must be restrained by the elites of the world.
I would appreciate some response to the visuals I posted here, and the complete lack of correlation between temperature anomalies and urbanization. If UHI is significantly skewing the global warming signal, then there would have to be a correlation betweeen warming regions and urbanized regions. There is none!
Gary Hladik :
Regarding this page: http://www.ncdc.noaa.gov/oa/climate/research/ushcn/#urbanization
I think it is poorly worded but it does in fact say that the greem line is the result of the processing algorithm used in HCN version 2 data. They say that no individual correction is applied because the general algorithm detects the step change and false warming from urbanization.
So, if my reading is correct, it is green line, not red.
Here is a graph of the NOAA GHCN data fro Reno showing unadjusted (blue) and adjusted (red). http://www.appinsys.com/GlobalWarming/climgraph.aspx?pltparms=GHCNT100AJanDecI188020080900111AR42572488000x
NOAA adjustments remove some of the older warming but not the recent warming.
Here is a graph comparing Reno with the 3 closest stations listed as rural in the NOAA GHCN database since 1970. http://www.appinsys.com/GlobalWarming/climgraph.aspx?pltparms=GHCNT100XJanDecI197020080900410AR42572488000x42572488001x42572488003x42574501004x
Reno shows substantially more warming than the rural stations.
Coby, you are way behind the curve. Start out by reading the old threads on this blog, starting at the beginning. Biases are not limited to urban settings. In fact, the term “UHI” is a misnomer. It should be “AHE” – anthropogenic (direct) heating effect. Such effects can be seen in locations that are “lights = 0”
Go catch up, then come back here.
are people not aware that the satellite records show a similar amount of warming as the surface records over the past 30 years?
bob, people are quite aware of the satellite temps. But upon closer inspection the satellite records show a much more pronounced cooling trend while the surface trends like GISS temp now only show a leveling off of the temperature rise