Mark Johnson, Chief Meteorologist of WEWS in Cleveland writes:
A friend of mine, Justin Berk, a local TV Met in Baltimore, MD had this story to tell today:
“There’s something fishy going on at BWI (Baltimore International Airport),” he says. Hourly obs at BWI airport (April 12, 2012) never went higher than 59 degrees.
(See the obs from BWI below – Anthony)
But, he noticed the official high temperature was listed as 62 degrees.
“There’s no way a jump of 3-4 degrees occurred and then fell back down between obs,” he added. Why the discrepancy? Justin called the local NWS office.
For a brief 10 minutes, the steady NW wind that persisted all day at BWI shifted to a westerly direction. That allowed the HEAT from the nearby runway to provide a quick 3 degree warm-up between hourly obs. Once the winds shifted back to a NW direction, the temperature fell back to 59 degrees.
The NWS employee concurred that the extra warmth came from the runway.
Global Warming is real (thanks to poorly-sighted thermometers)! This is the second time Justin observed a false high temperature reading this week at BWI.
========================================================
I followed up on this, and his story checks out.
First the table of high/low for the day from BWI:
Source: http://www.nws.noaa.gov/climate/getclimate.php?wfo=lwx
And here are the hourly observations for the day from the BWI airport ASOS station:
Source: http://www.nws.noaa.gov/obhistory/KBWI.html (downloaded at 2AM EST 4/13)
Here in what the BWI ASOS station looks like along with compass points added:
Source: http://binged.it/HC0WPG
And here is a view looking to the west:
Source: http://binged.it/HFgGmM
Seems an open and shut case. This is not the first issue with weather observations at BWI, they also have issues with measuring snow, and I reported here:
BWI snow record rescinded: Another reason why airports aren’t the best place to measure climate data.
=============================================================
UPDATE: 3PM PST 4/13/12 The Capital Weather Gang tries to avoid the siting issue with an alternate explanation. This from comments.
I’ve posted a different perspective on this at washingtonpost.com: The case of the curious temperature spike at BWI airport: asphalt or the sun?
REPLY: Anything to avoid UHI it seems with you guys. As for sun/wind debate. It could very well be both. Asphalt absorbs sunlight pretty well. More sun coupled with a shift of wind to the asphalt area can easily make a quick 3 degree jump. Sunlight by itself on grass, not so much. You didn’t mention albedo in your article so I’ll assume you don’t understand it.
Bottom line – airports are a poor place for temperature observations used for climate purposes, as they aren’t representative and are very dynamic with land use changes, and, see this detailed analysis.
http://wattsupwiththat.com/2012/04/13/warming-in-the-ushcn-is-mainly-an-artifact-of-adjustments/
Airports are part of USHCN.
– Anthony
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
David H at 3;25AM
The problem is related to “representativeness”. Our concern is just how much of the observed temperatures are caused by the increase in GHG concentrations. Temperatures that represent or can be used to calculate that effect should not be affected by local conditions which can change irrespective of changes in the atmosphere. As Anthony has been trying to point out, airport measurements poorly represent temperatures to be used for GHG attribution calculations and this example shows definitively that local conditions at airports affect temperature.
There is another problem illustrated. The new meteorological sensing systems collect one-minute data averages and those data are being used to set the daily maximum values. At some point in the past, the data were collected with manual observations. Clearly there is a much greater change of higher maxima and lower minima now that we have continuous data.
Would the new weather stations with the wind screens have prevented that problem/false reading?
Steve Mosher
are you saying that if it’s windy then the UHI doesn’t exist? In very simple terms surely the heat is spread round over a larger area with a lower increase in temperature. I think of it like a fan blowing on an auto’s radiator the amount of heat (or energy or whatever you want to call it) remains the same in the whole system, its just that different things and volumes are warmer/cooler. If you only measure the temperature of the radiator then you’re mislead, no?
To my simple mind on windy days UHI is spread further down wind, the windier the day the further down wind it goes and, perhaps, becomes harder to measure.
Thanks
thailandtours says:
April 13, 2012 at 2:26 am
It’s the abroad tour that everyone could afford, and the Thailand tour package offers it all. To be precise, it is the travel package for everyone, and ThailandToursInc.com arranges teh deal. The link at http://www.thailandtoursinc.com comes with the complete details.
———————————
Looks like the mods approve comment adds.
[Thanks. Spam deleted. ~dbs, mod.]
I thought this set of paired temperature data does a bang-up job of showing how airports are rotten for gathering the accurate temperature data
The city is on the North Carolina/Virgina border and right on the ocean. Not that far from Baltimore (a 230 mile drive). Take a look at the city vs the airport! Norfolk City and
Norfolk International Airport
Other Coastal Cities in North Carolina:
Elisabeth City
Wilmington NC
Here is the raw 1856 to current Atlantic Multidecadal Oscillation Amazing how the temperatures follow the Atlantic ocean oscillation as long as the weather station is not sitting at an airport isn’t it?
BWI, was originally called Friendship airport and dates from the early 1950’s. It’s expansion has been huge, now having more traffic than Washington National. With all the asphalt and concrete that has been laid, it’s got to have it’s own mini UHI issues. Wiki says that the expansions started in the 70’s. Interesting coincidence.
I really, really like having weather stations at the airport. They provide the information that can be used to have the pilot take off and land their aircraft safely. However, they are only indicative of the “climate” at the airport. To use them for anything else is foolish, but sadly we have many fools among us.
Sorry, I’m not buying this. Theres a more logical reason for this. Notices in those obs that the winds were 10 mph gusting to 18. That is PLENTY of wind to mix out that heat coming from that tarmac. This article acts if while the ASOS station had NW winds the tarmac to its west has no/ calm winds. It doesn’t work like that. Those NW winds were blowing over that tarmac and mixing out. So what caused the jump?
If you have ever looked at a topographical map for the BWI area there is the MD fall line about 3 miles to the west of the airport. Its a fall of 300-600 feet over 2 miles. I’ve noticed even at my house (which is 9 miles away and in a wooded area) that when the wind blows from the west you get an increased jump in temperatures due to downsloping from the fall line. Thats NOT an ASOS placing problem, it’s natural and has been there way before the airport ever existed.
This happens at RDU as well. An interesting consequence of this is appearing. I subscribe to Weather Underground and keep its page open in a browser panel all day. Recently I’ve noted an interesting discrepancy between the NWS forecasts for RDU and reality. First, the forecast is almost always high compared to what the day turns out to be, on average in the area. Second (and perhaps more interesting) it often remains high even well into the day of the forecast as the temperatures are obviously not going to make the mark, perhaps because there is unexpected cloud cover and the temperature is in no way going to make 84, it will be lucky to make 78. Third, weather underground actually has a regularly updated local forecast one can click through too with the expected temperature every four hours, and this will often have reasonable predictions that aren’t aligned with the NWS forecast. Fourth, the minimum temperature predictions for a day are often absurdly incorrect, but in a way that again almost always makes them higher.
For example, on the Weather Underground page for Durham, NC right now the low temperature prediction for Friday is given as 41F. The temperature outside right now is 41F. However, at 5 am, the temperature outside was 35 F — a few short degrees away from a disastrous frost (that for some reason they failed to warn of yesterday). The Weather Underground’s hourly report (that you have to click through to see) clearly shows the drop to 35 around 5 am, but the NWS forecast is what they put up on their main page and it is already absurdly wrong. What’s up with that?
One thing that’s “up” is that somebody who just looks at the NWS numbers and doesn’t actually track temperature is left with the impression that the minimum temperature of the day is some 3K/6F warmer than it really is/was. When they do the same thing on the warm side with the high temperature, the whole day appears warmer than it is, was, or will be. Since most people do not track the actual temperature but instead look only at the forecasted temperatures in the newspapers or on weather sites, by posting consistently high forecasts the NWS effectively shifts reality several degrees warmer everywhere in the minds of the people. After all, nobody posts the maximum and minimum temperature actually observed yesterday so nobody knows how accurate the forecasts were or what the high and low really were unless they look!
I would bet that even a cursory examination of NWS forecasts would reveal the high bias — it is quite pronounced in Durham and I have no reason to believe that it isn’t as obvious elsewhere. Sadly, today isn’t even an unusual case — a low forecast higher than the already observed low for the day.
The big question then is, why? I’m not one for conspiracy theories — I find it difficult to believe that there is a fiendish person in the NWS who has coded “add 3F to all forecast numbers” into their computers. Perhaps it is just this phenomenon at work. The “official” Raleigh/Durham temperature site is generally taken to be RDU (the airport). Never mind that the airport is flanked by three roads that carry several million cars past the airport every day, maintaining a blanket of local CO_2 and H_2O from the car exhaust over the airport itself. Never mind that (as a busy airport hub) it has jets idling and landing and taking off every few minutes all day long, dumping equally huge amounts of CO_2 and H_2O from burned jet fuel out in the sky directly overhead (in case the cars alone weren’t enough). Never mind that 40 years ago when I first arrived in Durham to go to Duke, the entire airport was the size of a small-town bus terminal with a dozen planes a day and a tenth of the car traffic going anywhere nearby. Never mind that today Raleigh, Durham, Morrisville, and Cary have grown out to surround the airport with city and stores so that the haze and greenhouse gases of the entire urban area continue to blanket it no matter what direction the wind blows. And then yeah, it has a lot of runways, all black tarmac, and I’m quite certain that when the wind blows just right it spills all that extra heat right onto its weather station(s). It could well be that its peak temperature is several degrees warmer in spikes, that it stores enough heat in its runways and traps it with the local greenhouse blanket at night to keep is lows several degrees higher than they are ten or fifteen miles away on the outside of the urban ring that surrounds it. Maybe NWS forecasts are just (correctly) forecasting spurious temperature readings caused by siting the area’s official temperature at a flawed site.
The big question at this point is — why bother? We don’t need to make any single weather site “the” official site for an area any more. Weather Underground also accepts input from citizen-owned weather stations and makes them available in a standard web format to anybody that clicks through to them. There are dozens of them in Durham and the surrounding area. There are rural sites and urban sites. All of them record the temperature all day long (and the wind and rainfall) so that graphs of them are instantly available, and all of them have history automatically stored so you can see the record of any station on any given day since the station was established (or various averages).
Of course this is horribly unscientific. People site their stations (I’m sure) wherever they like on their property, more often at a location dictated by esthetics or a need to cable it up than by weather measurement perfection. Some are in the woods, some in open fields, some on the sunny side of the house, some on the shady side, some are urban and some are rural, some sit on hills and some in valleys, and all of these things will cause the local readings to be a few degrees warmer or cooler than the readings that would have been obtained as close as the other side of the house. Still, doing a flat average of all of these sites, “sight unseen”, within a given radius around RDU would beat the hell out of using RDU alone simply because one could hope quite reasonably for these variations to correctly average out to a meaningful average temperature.
Better still would be to locate not a single official station but a web of official stations at selected urban, suburban, rural farmland and rural woodland sites in a 30 mile circle around RDU and use an average of them all. The record of them all would also provide meaningful data on the UHI effect and quite aside from “UHI” per se on the actual spatial granularity of temperature variation on the ground. My own belief is that “weather station measured” temperatures vary by degrees F over lateral distances as little as tens of meters, and that much of that variations is highly systematic and consistent.
Still, I think it would be very interesting to write a top article on the possibly inadvertent “NWS Prediction Fraud”, if one could obtain the records that would clearly demonstrate it. Given that the NWS predictions are the reality for most people, a systematic bias in those predictions is all that is needed to provide the illusion of warming in the United States.
rgb
A few years ago, WUWT did a piece on the same subject. I took it to heart and drove to BWI on a sunny day with a slight NW breeze. There is a road that completly circles the airport. I drove the circle twice and observed that that temp was ~3 degrees warmer downwind of the airport compared to upwind. On the downwind side (near the park on Dorsey Road) the runway is only a few hundred feet from the road. This spot had one of the highest recorded temps according to my car thermometer.
Accurate for landing a Jet, not so good for recording high and low regional temperature records.
Here you will notice the temp for the Baltimore Science Center, (in the heart of downtown), was also 62F:
http://www.weather.gov/data/obhistory/KDMH.html
But…take a close look and note that it is a reasonable climb from 61.
And, as I have written before; BWI normally runs a few degrees cooler than Balto city. BWI expanded a few years back and made things even hotter.
Here also is the UMBC station, 6 miles away, near my office.: http://weather.umbc.edu/
Unfortunately, their site often malfunctions, and I can’t get it to give up yesterday’s plot. Maybe someone else here could? I watch it all day, and it never got out of the 50s yesterday…unless it sneaked by me.
I have never understood why average daily temperatures are not calculated by averaging readings made every ten minutes throughout a 24 hour day. In Southeastern Virginia, my home, we have many instances of warm air at night ahead of cold fronts in the winter that yield a daily high temperature in the 60s even though the temperature was in the 60s for under one hour and was under 40 for the following twenty three, yet the daily average temperature is listed as 50. This happens at least three times each winter, and has happened more than once in the same week. Since there are no warm fronts that are as starkly temperature contrasted as cold fronts, nor as numerous, There would always be an overall warm bias in these averages.
Gail Combs says:
April 13, 2012 at 5:21 am
I thought this set of paired temperature data does a bang-up job of showing how airports are rotten for gathering the accurate temperature data.
The ASOS stations are for reporting conditions at the airport. Period.
Using the readings as proxies for the surrounding area is foolhardy. I’ve seen temperature differences of +5°C between what the tower was reporting and what I was reading on my free air temp gauge at my location just outside the control zone. And at my altitude (500 feet AGL), the adiabatic lapse rate only accounted for one degree of the difference.
Mosher,
“Its not a big mystery why UHI is real and why it doesnt show up in the record.”
Didn’t it just show up in this record? It may not show up all the time in the record, but it looks as though it does show in some of the record. As you say, ‘conditions have to be right’…and they were.
Statistical analysis shows that this is not an “open and shut case.” The following are some standard deviations (2000-2010) for daily high temperatures:
BWI: 18.292°
DCA: 17.949°
MD Science Center: 18.653°
National Arboretum: 18.651°
Vienna: 17.095°
If there were a problem with the siting of the ASOS, one would expect a much larger difference in the variability of high temperatures between the Maryland Science Center and BWI. If one expands the area to include the Washington, DC area, one finds greater variability among sites in and around Washington, DC than between the Maryland Science Center and BWI.
In the end, what very likely happened was that downsloping, not heat from the airport runway led to the brief bounce in BWI’s temperature.
Don- If the temperature was consistently higher there would be no impact on deviation. Only if it was a sporadic event would you see statistical evidence. On non-windy days, heated jet exhaust will at some point waft past the station, yielding the temprary high temperature. This is why airports are lousy station sites.
Corey S. says:
April 13, 2012 at 7:02 am
Didn’t it just show up in this record? It may not show up all the time in the record, but it looks as though it does show in some of the record. As you say, ‘conditions have to be right’…and they were.
======================================================================
If the above example is correct, would you actually call it “UHI”?
This is just a siting issue
Similar problem at nearby DCA (Reagan National). The area around the ASOS station is covered in dark gray gravel. Anthony noted it in 2010 when it was running hot: http://wattsupwiththat.com/2010/06/01/hot-air-in-washington-dc-more-asos-failures/
Believe it or not, they have added more gravel since then. I took this google capture about a year ago (it’s still the same currently) http://i433.photobucket.com/albums/qq51/palmer2/national-asos.jpg Left is from Anthony’s post, right is current.
Steven Mosher says:
April 13, 2012 at 2:35 am
omnologos says:
April 13, 2012 at 12:24 am (Edit)
UHI, as it’s well known, is a phenomenon that happens locally everywhere but, as if by magic, globally nowhere.
#############
its rather easy to understand
1. Its a function of windspeed and direction. Note the temperature changed for 10 minutes
when the wind changed.
2. Its a function of timing. If that wind change happens at 19:54.. no change to TMAX
Its not a big mystery why UHI is real and why it doesnt show up in the record.
And the cloud cover at each recorded event, and the moisture content in air, the physical change(s) for different recorded events, and more.
It is entirely possible that every improperly sited station would need it’s own unique highly tuned adjustments as no two recording events would be under the exact repeated circumstances.
We know there is a UHI/LHI effect but has any site been shown to have been properly adjusted, let alone the entire network?
However, to be succinctly correct, one would have to say that in this instance, man-made warming was caused by the burning of fossil fuels.
Instead of carbon exchanges and hidden taxes, let’s shut down transportation and see how the greenies like THAT one! How will they get to their next conference in Cancun?????
mfo says:
April 13, 2012 at 3:37 am
“It seems very odd to locate a weather station just south of that part of the runway from where aircraft are taking off, . . .”
It might be odd if the purpose was a long term study of climate. For the pilot about to lift-off along that ribbon of concrete the “real close” measurements are just fine. Weather data from a irrigated green park a few miles away are not superior.
————————————————–
Anthony placed a table from BWI in his report. Notice it does begin with the heading “CLIMATE REPORT.”
The “climate” seems to be a lot colder this Apr. 13th than in 2011. !!
Seems to me if a thermometer is anywhere near pavement of any sort, then it will manifest UHI. The severity of this UHI would depend upon the prevailing winds. Trying to correct this thermometer data would be pure guesswork.
I concur with rgbatduke that this has been a problem at RDU ever since they moved the location of obs equipment from the NE sector of the airport to the current ASOS site in the NW sector near the north end of the 10K ft runway in 1994.
Standard deviation is fluctuation. If the Airport had a siting issue with the kind of problem alleged in the opening message, one would expect larger variability from high temperatures due to runway-related spikes. That isn’t the case. BWI’s high temperature variability is similar to that of nearby sites. It is highly unlikely that the Airport is suffering from runway-induced temperature spikes that distort its overall temperature record.
Consistent warmth e.g., UHI, would lead to a higher temperature average and particularly less variability in low temperatures.
Also, just so my point is clearer, I’m not saying that the runway has no impact. I’m suggesting that its impact is so small that it cannot explain the temperature bounce in question. Downsloping winds can lead to such scenarios and I believe that is what happened at BWI.