In the previous WUWT entry, Willis posted his data on the Svalbard weather station, noting that it is spliced data and that the station data has been merged in nearby station history:
0 km (*) Svalbard Luft 78.2 N 15.5 E 634010080002 rural area 1977–2006
47 km (*) Isfjord Radio 78.1 N 13.6 E 634010050010 rural area 1912–1980
425 km (*) Bjornoya 74.5 N 19.0 E 634010280003 rural area 1949–2006
Svalbard Luft is the airport. As we’ve noted time and again on WUWT, the stations in GHCN have a propensity for airport migration. Here’s what Svalbard Airport looks like to visitors:
Lots of nice black asphalt and buildings there to absorb the feeble sunshine at that latitude. But where’s the weather station?
No help from NCDC’s metadatabase, they have no clue either, all they know is that it is “at the airport”:
The equipment tab gives no clues, and the lat/lon is too coarse to pinpoint a location within the airport complex.
But, the great thing about Svalbard is that it is now a tourist destination. Regular jet flights are available. Fortunately for us, tourists take photos, and upload them to Panoramio. Here’s one photo likely taken right off the plane:
Here are some additional views:
I found another tourist photo on Panaramio, that shows the characteristic metal legs and struts of the Stevenson Screen visible beyond the plane that tourists are boarding:
More tourist submitted Panoramio photos of the airport show just how much asphalt tarmac there is around the station, such as this one:
And when you zoom in on that panorama image at far left, sure enough, there’s the Stevenson Screen again:
It is clear that there’s a lot of asphalt around the station, but there’s also a lot of snow too. What happens when it snows at the airport? They clear the runway and tarmac, of course:
Where’s the Stevenson Screen? Right at the edge of the tarmac.
So, aviation snow removal makes a nice black year round albedo, right next to the weather station. Plus jet exhaust, generators, steam driven de-icers and other tools of the aviation trade are also nearby. Even if the Stevenson Screen has been abandoned in favor of an automated sensor, as often happens at airports, both would still have some locally measured effects in the record.
In the Arctic and Antarctic, aviation is the lifeline of humanity. A warm pocket of energy use in a sea of snow and ice. It would be interesting to plant a few of my portable USB logging thermometers around Svalbard away from this pocket of humanity to see what sort or temperature readings we get. By bet is that we’ll see a local AHI (Airport Heat Island) at Svalbard. It’s a busy place. In 2009, the airport had 138,934 passengers. Source: ^ Avinor (2010). “Årsrapport Passasjerer” (in Norwegian). http://www.avinor.no/tridionimages/2009%20Passasjerer_tcm181-109035.xls.
In Willis’ previous essay, he notes RC calls it a 5 sigma outlier event in April 2006. I had surmised it might be due to a tarmac resurfacing changing the albedo. I could be right. From the Wiki article on Svalbard airport:
In 1989, parts of the runway were re-insulated, giving these areas that previously had been the worst an acceptable solution. In 2006, this measure was conducted on the remaining parts of the runway.
There was construction going on in 2006, Oddly the source of that metadata is from a paper on gauging the airport performance under the “duress” of climate change:
Svalbard airport runway. Performance during a climate-warming scenario. (PDF)
In a study initiated by the Norwegian Airport Authorities in 1995, insulation of the whole runway in a manner similar to the 1989 procedure was deemed the most favorable long-term runway maintenance strategy (Instanes, D. and Instanes, A., 1998). This has so far not been carried out, and a new reconstruction is planned for 2005/2006 to improve the runway. The average global surface temperature is projected to increase from 1.4 to 5.8°C between 1990 and 2100 (IPCC, 2001). Warming at higher latitudes of the Northern hemisphere may be greater than the global average, as high as 4 to 7°C between 2000 and 2100 (ACIA, 2004).
They don’t seem to realize anywhere in the paper that the temperature data they are relying of for input to their models used for permafrost thaw comes from the little white box at the edge of the tarmac. Talk about positive feedback and polar amplification. Let’s build a new runway; hey look it’s warmer we were right! Sheesh.
People like Jim Hansen and Gavin Schmidt who sit up at the top of the climate food chain and take data from these weather stations at face value and then use it to extrapolate to nearby grid cells because there are no other nearby stations in the Arctic really need to get out more and see what the measuring environment is like. Maybe somebody can convince them to get off their taxpayer funded butts and away from their computer screens someday and do some field work.
Of course given what they did to censure Willis at RC when he brought up the issues at Svalbard, I doubt they’d believe their own eyes if it contradicted their expectations.
UPDATE:
The UHI at Svalbard airport has been measured, using the driving technique I first wrote about back in fall 2008 to study UHI in Reno, NV. This study at:
http://climate4you.com/LongyearbyenUHI%2020080331.htm
…was pointed out by commenter Ibrahim and is reproduced below:
=============================================
Longyearbyen UHI experiment, March 31, 2008
Longyearbyen March 31, 2008 16:15 PM (not corrected for summer time), looking WNW from the northern end of the lake Isdammen (see map below). The sky was almost clear, with a few local clouds forming over the fjord. The wind was weak from southeasterly direction, 0.5-3 m/s. The large building in the distance to the right is the main hangar at Svalbard Ariport. Compare with map below.
The general weather situation, measurement equipment and measurement route
The air temperature was about -20oC, and the wind weak from southeasterly direction, 0.5-3 m/s, but with local deviations (see map below). The sky was almost clear. The ground surface was covered by snow. The nearby fjord was ice free, with the exception of a 5-50 m wide zone with new icew along the coast. A thermistor was attached to the roof of a car (c. 1.5 m above terrain), and temperatures were logged at 2 sec. intervals. The time given in the diagrams below are not corrected for summer time. The measurements were carried starting at Svalbard Airport in the upper left of the map below, driving SE along the coast to the town, making a roundtour here, before proceding SE into the lower part of the major valley Adventdalen.
Longyearbyen is the worlds northernmost town and is located at 78o17’N 11o20’E, in central Spitsbergen . The present number of inhabitants is 2,001 (January 1, 2007). There is no official meteorological station located in Longyearbyen at the moment. The official meteorological station is located at the airport, about 4 km northwest of Longyearbyen, close to the coast (see map below).
Topographic map showing Longyearbyen and Svalbard Airport (Svalbard Lufthavn). The red line shows the measurement route March 31, 2008, starting at the Airport and ending in the lower part of the valley Adventdalen to the SE. In between, a detour was made in the central part of the town as shown. The wind was weak, 0.5-3 m/s, from south easterly direction, but with local deviations (blue arrows). The fjord was ice free. The map section measures c. 11 km west to east.
Results
Result of temperature measurements along the route Svalbard Airport – Longyearbyen -Adventdalen, March 31, 2008. The official Svalbard meteorological station is located at the airport. Se map above for reference. Time (not corrected for summer time) is given in hh:min:ss format along the x-axis.
Interpretation of results
The whole area was snowcovered. The sun was below the skyline formed by the mountains, and albedo effects caused by buildings and roads for that reason presumably not very important.
The registered air temperatures show an overall falling trend towards SE along the main measurement route. Near the airport, where the official Svalbard meteorological station is located, air temperatures are relatively high (about -18oC), which is interpreted as the result of the onshore airflow from SE across the ice free fjord. Further towards SE, this local warming effect diminishes, and colder air (about -25oC) draining out of the valley Adventdalen dominates. The temperature difference between the Airport and Adventdalen is about 8oC, representing the open water effect (OWE) at this particular time. In between, the local heat island effect of the town Longyearbyen is only weakly developed. The maximum UHI effect appears to be about +0.5oC at the time of the experiment. The local cold trough recorded within Longyearbyen (16:04) corresponds to the position of the main valley axis, where cold air masses is draining NNE from the glacier at the valley head.
The existence of an urban heat island effect in a relatively small settlement as Longyearbyen may come as a surprise. This is, however, not the first time this has been observed in the Arctic; see, e.g., Hinkel et al. 2003.
=============================================
UPDATE2:
A commenter asked if satellite and surface data deviated here. Willis provides the answer.
I just looked at the MSU versus the NORDKLIM/GISS record, and the surface record shows much more warming than the satellite warming, almost twice as much. The surface record shows warming at 0.10 °C/decade, while the MSU record is warming at 0.06 °C/decade … here’s the graph:
w.
UPDATE3: From comments, we have a close up photo of the Stevenson Screen near the edge of the tarmac. Thanks to commenter “Oslo” for finding it at the Norwegian Metorological Institute website.
I wonder if the cinder blocks are a permanent feature?
UPDATE4
Erik Kempers writes in comments that he has found a Panaramio photo that was misplaced on the map that shows the weather station in perspective with the Svalbard airport runway and taxiway/tarmac. I’ve provided a zoomed and annotated version below. The original is here.
Rod Smith says:
May 13, 2010 at 10:38 am
On the other hand, as aircrew, your experience would be travelling from aircraft to hanger. My experience going the other way, away from the hanger was the pan, (flight line,) was warmer in winter.
DaveE.
Here are some actual air traffic stats for Svalbard.
From Avinor:
http://www.avinor.no/avinor/trafikk/10_Trafikkstatistikk
Airport movements July 2008:
Passenger/Freight/Charter: 646 (including 41 from abroad)
Other planes: 294
For a total of 940
Average: 30 a day
Airport movements December 2008:
Passenger/Freight/Charter: 361 (including 5 from abroad)
Other planes: 135
For a total of 496
Average: 16 a day
Airport movements 2008:
Passenger/Freight/Charter: 6451 (including 211 from abroad)
Other planes: 2460
For a total of 8911
Bjorn
David Alan Evans: “On the other hand, as aircrew, your experience would be travelling from aircraft to hanger. ”
Absolutely untrue. Aircrews in my day had a lot of work on flightlines. We didn’t have a hanger big enough for B-36’s at our home base, nor a lot of places we deployed. Your experience may be the way it works now, but NOT the way it worked in the LeMay’s SAC. I remember sleeping under the wing of a B36 during an extended “get this thing off the ground as soon as it can fly.” We were there about 20 hours. And a gunnery mission took 8 hours minimum to install, preflight and load. And so on.
I could continue, but I won’t bore you.
Rod Smith says:
May 13, 2010 at 6:53 pm
I suspect you’re talking a concrete pan, not tarmac. big difference.
DaveE.
Willis,
Your trends look like they should be in units of C/year, not C/dec. (cf Update2)
Relating to the meteorological observations done at Svalbard Airport, you may find some additional data and observations relevant for this discussion on:
http://www.climate4you.com/SvalbardMetPhenomena%2020080902.htm
and
http://www.climate4you.com/UrbanHeatIsland.htm
Best wishes, Ole
Nick Stokes says:
May 14, 2010 at 12:16 am
Willis,
Your trends look like they should be in units of C/year, not C/dec. (cf Update2)
Nick’s right. This means that the much wider region covered by the MSU readings is warming at 0.6 deg per decade. I doubt the trend at Svalbard has got anything to do with the airport. The graph is pretty much what you’d expect when comparing a single station with a wider region – apart from some of the 2006 readings, that is.
I’ve commented that Svalbard temperature is airport temperature a while ago here at WUWT, what’s also interesting is that there has also been a weather station in Longyearbyen proper – which probably has a very different micro climate from the airport closer to the fjord. Isfjord Radio is at the ultimate west of Isfjord, i.e. it’s probably even more dependent on SST than the airport.
Most Norwegian GISS stations are now airports. They used to have a station in Oslo city, but now it’s (just north of 60N) Oslo Gardermoen Airport which, since it’s labeled as “rural” probably adjusts the GISS temperature in opposite direction of “UHI correction” in a large number of Norwegian and Swedish towns:
http://data.giss.nasa.gov/cgi-bin/gistemp/gistemp_station.py?id=634013840003&data_set=1&num_neighbors=1
I’ve often wondered where the Gardermoen weather station is – is there anyone here who knows?
If you think the Gardermoen data linked above show a step change just before 2000, you may want to know that in October 1998 Gardermoen was opened as Norway’s international hub – it used to be just a smaller airport before that.
With regard to the “station is not on the tarmac” ‘argument’ by Nick Stokes:
With reference to the close-up photo you cite: notice the day-glow red poles attached to the leading edge of the wooden stairs: those are plow guides, used to show snowplow operators where low-lying objects may be concealed under snow.
Airport plows don’t plow on the tundra, only on the tarmac. There’d be no need for those guides if the stairs and therefore the Screen were not immediately adjacent to the tarmac.
This of course being beside the fact that the third photo in the original article clearly depicts the Screen as located immediately adjacent to the tarmac.
Tilting at windmills is generally indicative of a lack of a better argument, don’t you think?
I wonder wether these AHI and UHI effects also explains the warming of the permafrost that has been observed at Svalbard…
http://met.no/filestore/temperatureseries_30m_janssonhaugen_no.jpg
REPLY: It might, if that temperature measurement is taken near the town or airport. Do you have a location for this measurement? – Anthony
Anthony: More on Janssonhaugen here: http://www.cicero.uio.no/fulltext/index.aspx?id=2059&lang=en
Note the following: “The closest meteorological station to Janssonhaugen that has been keeping long temperature series is Svalbard airport, Longyearbyen. Here a homogenous temperature series was constructed back to 1912 (Førland et al. 1997). The series show that air temperature increased by about 4 oC from 1912 and up to the end of the 1930s. The temperature dropped about 2.5 oC up to the mid 1960s, and then increased by about 2 oC up to today. The linear trend for the entire period of 1912–1999 shows a temperature increase of about 1.2 oC.”
The current warm period on Svalbard is not that much different from the previous one about 70 years ago.
Nick Stokes says:
May 14, 2010 at 12:16 am
Thanks, Nick, you are correct. I have recalculated them and updated the graphic and the text.
w.
Vidar says:
May 14, 2010 at 10:10 am (Edit)
According to this document, the borehole is about sixteen miles from the airport.
Vidar says:
May 14, 2010 at 10:10 am
I wonder wether these AHI and UHI effects also explains the warming of the permafrost that has been observed at Svalbard…
http://met.no/filestore/temperatureseries_30m_janssonhaugen_no.jpg
REPLY: It might, if that temperature measurement is taken near the town or airport. Do you have a location for this measurement? – Anthony
I don’t want to keep labouring the point but it’s obvious from the graph that AHI is not responsible for the Svalbard trend. In the early 1980s blue(MSU) and orange(station) lines are coincidental. Similarly between 1994 and 1998. In the last couple of years they’ve been within a whisker of each other. There is a 0.5 deg difference between the trends. If there were a spurious warming trend there would be a difference of 1.5 deg between the orange and blue line in 2010. The difference in the trends is almost certainly due to the greater variability in the station trend i.e. the dip in the late 1980s and the spike which peaked in 2006. This is to be expected. The greater area covered by the MSU measurements will result in some smoothing. The reason the surface data produces the greater trend is due to the least squares method of calculation.
Vidar says:
May 14, 2010 at 10:10 am (Edit)
“I wonder wether these AHI and UHI effects also explains the warming of the permafrost that has been observed at Svalbard…
http://met.no/filestore/temperatureseries_30m_janssonhaugen_no.jpg”
Can it be possible that this time, compared to 1920-50, a multi million dollar buisiness is backing up scientists LOOKING for signs of global warming? It is hot allright, but is it warmer than earlier dekades? Looks like most of the climate “scientists” have hypochondriasis. They are looking for diseases where no such can be found.
Pardon my english, I know it’s not good;-)
Gail Combs:
Svalbard is emphatically not the gunslinger’s paradise you think it is. Polar Bears are a fully protected species there. It is true that at least one person is required to carry a rifle in each party moving outside Longyearbyen town limits. However you may ONLY use it for self-defence, and only when you are positively being attacked by a Polar Bear. You are also required to fire a warning shot to scare the bear away before shooting it (this almost always works). If you do shoot a bear you must immediately report it, and there will be an investigation of the circumstances. If you are found to have shot it wantonly you are in big trouble. Most of Svalbard is actally a national park where all hunting is prohibited.
Incidentally nobody has been killed by a Polar Bear in Svalbard for almost fifty years.
The point I wanted to make, was that while Anthony criticized Gavin Schmidt and James Hansen for not knowing anything about the stations at which their data are gathered, it might also benefit Anthony’s work to know a bit more about the areas he’s investigating. I totally understand the point on uncertainty at observation stations, but also Anthony’s conclusions depend on the very same single station.
If Anthony had been familiar with the area, he would have known that Longyearbyen and Svalbard Lufthavn have no sun between 27th October and 15th of February (3.5 months, or almost 1/3 of the time), with the result that the albedo part of the AHI-effect would be zero in that whole period, in comparison to summer months, where there will be an albedo effect 24 hrs a day. Thus, the first thing to do should be to check for different trends in winter data and summer data.
However, anyone familiar with the nearby Barents Sea, know that measurements made by Russians since 1900 and Norwegians since 1977 show that the heat content of the Barents Sea is currently (peaked in 2007) at all time high since measurements began. This implies that the whole region is actually in a relatively warm period, and this is based on measurements that are totally independent of the weather station at Svalbard airport.
The station data should also be compared with data from Hopen, which is closer to Svalbard than Bjørnøya, and it is also, perhaps, to a lesser degree dependent on the atmospheric situation and temperature in the Norwegian Sea and the Barents Sea Opening, due to it’s geographical position.
Also, Svalbard is situated at the end of the “Highway” on which hot and moist Atlantic air is advected northward by the NAO, that is the high-pressure located above the Azores and the low-pressure located above Iceland. This has a huge influence on the Svalbard temperatures – far more than a tiny effect from the increased albedo at the airport.
One extreme example: A few years ago, warm air masses was brought all the way from the North Atlantic and northward to Svalbard, due to a strong high-pressure system above Europe and a deep low above Iceland. This was in December, the sun was below the horizon all day around, the sky was more or less clear, but still the temperature in Longyearbyen was above zero! That’s the potential of advection!
NCEP and ERA reanalysis data are freely available on the internet, and could be used for checking whether the anomalies pointed out by Anthony had anything to do with anomalous advection of air masses.
All of this is obvious and well known to any meteorologist or climate scientist, but still, it seems to be missing in this discussion..
REPLY: I mainly concern myself with siting issues, which has been the focus of the surfacestations.org project. But you make a good point about the albedo and Arctic circle endless winter night and endless daytime in summer issue. My point mainly was that when snow is on the ground (as seen in the photo provided) the AP albedo becomes black in a sea of white due to AP runway maintenance. AP’s aren’t static environments, they are bustling centers of activity and change, which makes them unsuitable for climatic purposes. Read some of my other entries on weather stations at AP’s and you’ll see what I mean. Honolulu is a good example where the AP mission of aviation monitoring trumps climate monitoring, and when things go bad, they won’t even correct the record. http://wattsupwiththat.com/2009/06/19/more-on-noaas-fubar-honolulu-record-highs-asos-debacle-plus-finding-a-long-lost-giss-station/
I didn’t mention ocean currents in that story about Hawaii either, focusing instead on the siting and instrumentation issues.
You do make another good point about the NAO “highway, and if I were writing from that perspective (instead of siting) would most certainly include it. This issue makes the move of the data from the radio station to the AP even more important, because it puts the data closer to that stream of air. The question then becomes: what is really being measured there? – Anthony
I would like to add that most glaciers in Norway(mainland) was gone about 8.000 years ago and where reestablished about 5.000 years ago and since have had an cyclic increased mass with a maximum , last 11.000 years for mainland Norway around 1750 AD.
Svalbard had its glaciers maximum last 11.000 years around 1910/1920…
So what are we comparing todays climate in Svalbard with?
The cooldest period last 11.000 years?
tty says: ‘nobody has been killed by a Polar Bear in Svalbard for almost fifty years.’
Not so. Four people have been killed by bears since 1971, the most recent fatality being in 1995. I’ll be in Svalbard again this summer, and will be cautious about bears as always.
tty says:
May 15, 2010 at 5:26 am
Gail Combs:
Svalbard is emphatically not the gunslinger’s paradise you think it is. Polar Bears are a fully protected species there….
____________________________________________________________________
I was being sarcastic. I have seen city types in Yellowstone with no sense of self-preservation get between a mother and her cub to take photos of the cub. Unfortunately if the cub squeals the guy earns a Darwin award and the bear is put down.
I was just hoping we could send some of my least favorite congress critters there. Now you have to tell me someone loaded with common sense and a rifle will keep the polar bears from cleaning the gene pool. SIGHhhhh