People Send me stuff. I get pictures of weather stations from all over the world. Here we have Henderson Field, serving the capital city of Honiara, Guadalcanal, Solomon Islands.
Hi Anthony, I thought you might be interested to see this weather station.
It is the main one for the Solomon Islands, good situation for the airport, but as you can see in the photos, not so good for accurate temperature measurement.
The planes land then turn down the road to the Terminal apron, as they turn onto the apron, the jet exhaust washes over the weather station.
I’m not a technician, but I’ve repaired enough damaged equipment in my time to think that the exhaust heat may cause some problems with calibration over time.
Google earth -9.430025° 160.047393°
I’m trying slowly to get some full size photos loaded into google earth at present, but internet here is sporadic at the best of times, and down right miserable the rest.
Thanks,
Warren Nash
Solomon Islands
Here’s the closeup view of the weather station at the airport, the instruments are inside the fenced in enclosure.
As weather stations go, it isn’t bad, as the Stevenson Screen is 30 meters from the taxiway asphalt. That would make it a CRN2, acceptable by NOAA siting standards.
Here’s a ground level view of the station taken from the terminal:
Here comes a plane!
Coming into the terminal…
Hey, park it over here!
Uh, oh, look where the jet exhaust is pointed:
Hmmm, a new high temperature today?
Back to normal.
Now in the defense of the global climate record, it doesn’t look like this station gets updated at GISS very much:
And here’s the plot of data from NASA GISS:
Source here
Hmmm, pretty crappy data set dontcha think? Why have it at all?
If only GISS could use Weather Underground:
Eh, but that would require a thousands of man-hours, and millions of dollars in grant money to pull off.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Sorry. The UK has over 1700 miles of motorway. It has around 200,000 miles of road.
Stephen Skinner:
Ya I know. But the video was funny as heck.
Thanks for educating people on the high bypass ratio. I’m not sure everyone appreciates that. I’m pretty sure with a little work one could explain why the airplane exhaust will not impact the temperature. I look a little bit more at airports from a different perspective entirely.. here
http://stevemosher.wordpress.com/2010/09/29/cooling-stations-a-uhi-hint/
simpleseekeraftertruth says:
September 29, 2010 at 12:23 am (Edit)
Two definitions of anomaly;
1. any occurrence or object that is strange, unusual, or unique
2. a discrepancy or deviation from an established rule or trend
So plotting temperature anomalies against time would appear to (in)conveniently encompass both types. The question has to be: do anomaly temperature graphs show deviation from an established rule/trend or indicate events strange/unusual/unique? The surfacestations project is looking for noise in the signal but it could end up looking for signal in the noise!
(Or have I just embarrassed myself by revealing that I am the last to realise this?)
##############
reporting the “anomaly” of a temperature is a method of ‘standardizing” the measure according to a rule that does not distort the trend signal.
very simply:
Temperature: 15, 15, 15, 10, 10, 10, 15, 15, 15
To turn that into an “anomaly” you select a “period” That period can be any period or the whole series.
lets pick {10,10,10} You take the average = 10.
you substract the anomaly from the original series
Anomaly 5,5,5,0,0,0,5,5,5
And there you get Deviation from the normalizing period. take the slope of either the temp or the anomaly and you get the same answer. And we care about trend not absolute temperature
By taking the anomaly we dont have to be concerned about adding or dropping stations that are cooler or warmer. As long as you take the anomaly of the station you have no issues. you do have issues if you add or subtract stations with different TRENDS.
So dont get frazzled by the word anomaly. understand the basic math and you will see exactly why they need to be used in data series were you have data drop outs ( like a station that stops reporting )
How about putting a data logger in the screen, measuring temperature every 10 seconds, and then seeing if you can tell when a jet goes by.
Snowlover123: I beg to differ: I happen to enjoy mockery very much. You been drinking sour Maple syrup? Let this kid have his fun, and more power to him. If WUWT readers go over there give the guy a pat on the head in the spirit of Voltaire, “I disapprove of what you say, but I’ll defend to the death your right to say it”.
Stephen Skinner: Thanks for your comments
I agree that the jet exhaust issue is highly unlikely to be a cause for concern here.
For the following reasons:
1. The plane in the photograph comes in once a week as you note.
2. As you note the high by pass ratio means that most of the intake air is passed
around the combustion chamber.
3. The plane would have to arrive or depart in certain time windows to impact either Tmin or Tmax. if its 80F outside and the plane raises the temp 7-8 F that will not necessarily raise the Tmax for the site. Simply, if its 80F at 9am ( see the WU chart) and the plane raises it to 87F, thats really not a problem. Why, cause later that day the temp hits 89F at 2pm, when the plane effect is long gone and 89F gets recorded as the TMAX. For the climate we only care about tmax and tmin. So an airplane that raises the temp briefly will have no effect on TMAX. The only way it can impact TMAX is if just happens to exhaust the right amount of excess heat AT THE RIGHT TIME and PLACE. so, if it blew hot exhaust at 2PM, then you might have an issue.
( looking at the flights looks like there are a couple 2:10 flights )
Tmin is almost immune from the kind of effect a single jet like this could have.
Tmin occurs between 5 and 6 Am. somewhere from a temp of 72F at 5AM over the course of an hour it drops to 71.6.
Ok. Walk out to the thermometer at 5:30 and heat it artifically to 85F .
That Wont impact Tmin. cause Tmin happens at 6am. so if the temp goes from 72 F at 5AM to 72F at 5:29, to 85F at 5:30, you’ll be back down to to 71.6 by 6AM. It would take prolonged heating around the times of the day when Tmin and Tmax occur to creep into the records. And if it did infect Tmax by 1 degree, on one day of the week. That would effect Tave by only 1/2 a degree for that day, and when you average those 4 days with the 26 other days, the impact of the chance ocurrance is even smaller.
D. Patterson says:
September 29, 2010 at 2:25 am (Edit)
Look at the air temperatures for the airports in the United States in the time period 10, 11, and 12 September 2001.
That’s been done in the study on contrails. very famous study
Stephen Skinner says:
September 29, 2010 at 7:09 am
Just digging around it seems the plane in the photograph only comes once a week. There are a handful of other airlines but some are smaller turboprops. One airline operates the bigger A330 and the runway is about long enough to get one in and out but I can’t find anything that says they go in there.
The plane in the photo lands every day Monday to Friday at 14.00 approx, departs at 1500 approx.
Arrival times are staggered around 1400 for all airlines flying into Henderson Field from Brisbane.
Pacific Blue arrives approx 1410, departs 1445.
Solomon Air has a morning run, and an afternoon run.
There are also the weekend flight schedules, usually 3 – 5 planes landing Saturday and Sunday.
The inter island services use turbo prop aircraft, Solomon Air are upgrading to A330 from reports in the Solomon Star, and I believe Pacific Blue have A330 aircraft operating now on the Brisbane/Honiara service.
If I can convince someone with the Met Dept here I will try to get permits to go airside and check for any effect from the exhaust with a digital thermometer. The hardest part will be finding a thermometer in Honiara.
Steven Mosher says:
September 29, 2010 at 1:14 pm
I enjoyed your apparently successful attempt at self-delusion.
Go stand behind a modern jet. The relatively well-mixed air a hundred yards behind a high bypass engine is still warmer than the air that went in the intake.
The jet is 180 metres away from the weather station at the point it starts turning in front of the airport buildings.
Billy.
You still dont get it.
I’m granting that the air may be warmer. Here is what you miss.
Tmin for this location appears to happen on this day at 6AM
from 5AM to 6AM the air temp drops from 72 to 71.6.
Suppose your plane raises the temp 10F. suppose that. question, for how long? 5 minutes? Lets say 15 Minutes. lets say that blast of thust gives you a lingering 15 minute warm spot.
5AM 72
5:15 82
5:30 72
5:45 72
600 71.6
And when the climate guys get the data what do they look for Tmin. NOT THE BLIP.
they dont integrate temps, they take the max and the min.
To “infect” Tmin the jet would have to land at just the right time to infect Tmin.
Same goes for Tmax.
Now, if you have a bunch of different landings landings thats another story. I still think it doesnt matter in those cases, but in this case its just plain stubbornness to hold the position that you can tell from the photo that the plane will impact the temps. Every bit of evidence we have argues otherwise in this case.
if you are going to raise individual cases as examples of contamination, then you have to accept the fact that sometimes you cannot make a good case on the facts. The video is a good clue. but when you assemble everything you know, the grounds for concern diminish. You do better by focusing on the good cases, rather than throwing up weak cases just to muddy the water
I dunno but with arrival times and departure times and with the weather underground chart of hourly temps it would seem a easy matter to look for any blips
Rememeber how ghcn calculates an average Tave = (Tmin +tmax)/2
go find the blip in the hourly reports and you got a case to look further.
otherwise there are BETTER cases to look at and other more pressing issues.
Gday Anthony:
You said: As weather stations go, it isn’t bad, as the Stevenson Screen is 30 meters from the taxiway asphalt. That would make it a CRN2, acceptable by NOAA siting standards.
That may be fine in America, but Solomon Islands aviation regulations and meteorological standards, I am fairly sure, would have been set up with Australian or perhaps New Zealand assistance. The Australian specifications are:
A3.5.2.3 Clearances on airports:
A3.5.2.3.1 Turning areas and aprons 150 metres
A3.5.2.3.2 Runways 120 metres
A3.5.2.3.3 Taxiways 75 metres
So the siting is abysmally non-compliant!
Jet exhaust wash would have a minor effect compared with the heat from nearby tarmac.
Ken
For the folks saying the high bypass fan will be cool air:
1) It’s going through a hot engine and mixing with hot exhaust. We’re worried about a 1/2 C “anomaly” as Global Warming, so it doesn’t take much. Far less than you would feel on your skin.
2) Don’t forget that the “cold” air being sucked in is coming right off the tarmac. FOD (Foreign Object Damage) is from turbines sucking up bolts and stones off the runway… plenty of suction to pull in surface air. So that ‘cold’ air is both tarmac and compression heated, just not by 500 F …
3) As has been pointed out: Pilots NEED to know the temperature over the tarmac. that is where the wing goes, and to get off the ground your density altitude calculations must have that over the runway temp, not some ‘in the forest nearby temp’. To use these stations for “climate” records is exactly wrong given their design goal of accurately reporting RUNWAY temperature.
4) Aiports ARE hotter than non-airports. (I’ve been at enough of them…). Between snow removal, de-icing sprays, fuel burn, ground vehicle heat, tarmac heat, building heat, and surrounding traffic and parking area heat, they are just darned hot places.
5) Airports consistently have grown over time. We went from nearly no commercial traffic and small prop planes in 1950 to the Jet Age and a gazzillion flights a year with everyone and their cousin flying somewhere. They are NOT a step function heater.
JJB MKI says:
Out of interest I thought I’d take a look at the weather stations in England and Wales used by GISS to see if the temperature trends might be contaminated by the ‘Airport Heat Island’ effect. Post 1995, all of the temperature data for England and Wales used by GISS comes from stations located at international airports (5) or large RAF airbases (3). […] Besides this, all five international airports now used have undergone expansion over the last 20 years.
This is only a casual observation and I’m no climatologist- perhaps there are good reasons why GISS would only use data from these locations to help demonstrate the late 20th century ‘warming’ period, and I don’t know if they make any adjustments to data coming from stations located next to or on top of asphalt runways, but it might be an interesting starting point for further investigation..
Yes, it was…
http://chiefio.wordpress.com/2010/05/22/an-easy-airport-heat-island-audit/
Using SJC and 95050 that gets a Santa Clara station (San Jose Airport has one side basically against Santa Clara) gives SJC at 74 F with Santa Clara at 71.9 F (Santa Clara is measured 43 minutes after SJC, so there is a chance of a TOD error). Barryessa, California, on the other side has 72 F and the same TOD as Santa Clara.
North Willow Glen, just south of the airport past a couple of freeways has 76.3 at the same 8:35 pm time as Santa Clara but SJC has a S wind indicate with “calm”, so my guess would be that we are getting our usual evening ‘cool’ drift in from the bay sending the SJC heat island breeze toward North WIllow Glen. Sure, it could be plain old UHI added in from downtown San Jose (right on the other side of the airport…)
But what it’s not is that SJC is pristine and accurate while Santa Clara and Barryessa have just decided to get cold for no good reason.
Oh, and GIStemp treats many (most?) airports as “rural” and often uses them to “correct” for UHI in other stations (which may explain why half the time it has UHI “correction” going the wrong way…)
http://chiefio.wordpress.com/2009/12/15/of-jet-exhaust-and-airport-thermometers-feed-the-heat/
http://chiefio.wordpress.com/2009/09/04/most-used-rural-airport-for-uhi-adj/
http://chiefio.wordpress.com/2009/08/23/gistemp-fixes-uhi-using-airports-as-rural/
http://chiefio.wordpress.com/2009/12/08/ncdc-ghcn-airports-by-year-by-latitude/
Yes, it’s that bad. Our major global temperature data set basically is measuring the UHI and growth at Airports and in their surroundings.
Steven Mosher says: By taking the anomaly we dont have to be concerned about adding or dropping stations that are cooler or warmer. As long as you take the anomaly of the station you have no issues. you do have issues if you add or subtract stations with different TRENDS.
This would be true for the simple model you present, but you know darned well that the temperature series codes like GIStemp DO NOT do anomalies by comparing a single thermometer against itself. They do a host of machinations and create a fictional “Grid/ Box” temperature with one set of thermometers in the “baseline” and a different set in the present. This is, IMHO, broken. There is no ‘thermometer trend’ in this case, just blended box average temperature fictions.
It is similar to me saying that the average top speed of my high school era VW and Ford Fairlane is slower than the average of my Mercedes SL and BMW 325is today, and that clearly there is a trend of cars speeding up over time.
While I appreciate your giving a trivial anomaly example (and saving me the trouble of doing it) you really DO need to avoid giving the impression that this is how the temperatures are actually handled when they are not.
So dont get frazzled by the word anomaly. understand the basic math and you will see exactly why they need to be used in data series were you have data drop outs ( like a station that stops reporting )
And it would be really nice if the temperature series DID use simple self to self anomalies instead of what they do. When the temperature is looked at with a simple thermometer to itself only anomaly, a much different pattern emerges, one where the temperatures have a clear “hockey blade” in about 1990 when all the station dropouts happen and the airport percent skyrockets to near 90%.
http://chiefio.wordpress.com/2010/04/11/the-world-in-dtdt-graphs-of-temperature-anomalies/
So unless all the CO2 did nothing through the entire industrial revolution and suddenly got switched on in 1990, it’s not a CO2 problem, it’s a data and processing problem.
And part of that problem, IMHO, is the use of “Grid / box” anomalies with varying box contents from year to year (with boxes often having very few, sometimes only ONE thermometer in them; so sometimes you really are comparing ‘apples to oranges’ as ONE thermometer is in the baseline and some OTHER thermometer is in the present “grid / box”. VW fastback vs Mercedes SL so cars today are going much faster…)
Oh, and per the “does too, does not” on can the airplane “infect” TMIN and TMAX, for many airports it certainly can.
At SJC, for example, TMAX comes mid-day, as does a boatload of flights. TMIN comes early morning just before sun up fairly often, right when maximum departures are scheduled for business travelers.
Does that happen at Solomon? Don’t know. But I do know that the tarmac and concrete will hold heat from the daytime into the night (helping to raise TMIN) and will be much hotter mid day (and definitely raise TMAX on sunny days) and having an airplane deliver that heat to the box with a big fan will not make it cooler…
I’d not be so quick to dismiss that as a possibility.
Oh, and it doesn’t have to be EVERY day. Remember we are dealing with a MONTHLY MEAN in most cases, so depending on how that monthly mean is calculated it could be only one per month that needs a little ‘help’ to warm the planet… So that ideal timing of the airplane need only happen once a month to shift the whole thing. (I’m pretty sure the USA does average of daily (min-max) average so would only get a small ‘lift’ from one day; but other nations data may be handled differently. I could find no statement of a standard, but did find discussions of possible variations. But even raising just a few days a month by a few degrees can get us to our 1/2 C AGW Panic Point.)
Steven Mosher doesn’t know what a stable nocturnal layer is.
EM:
“Oh, and per the “does too, does not” on can the airplane “infect” TMIN and TMAX, for many airports it certainly can.”
I am not arguing that it CANNOT. I am merely pointing out this.
1. certain conditions have to obtain for this to happen. The right amount if excess heat, at the right time. Clearly, if Tmin happens ( for THIS LOCATION) at or around the early dawn and if TMAX happens at or around 2PM FOR THIS LOCATION, then this seems obvious:
A. if all the planes took off and landed between 5pm and 10PM you wouldnt even have a CHANCE to infect the measurement.
Do you agree or disagree. Simple question
2. For the “infection” to show up as a significant factor in monthly means, you have to have relatively frequent flights. For example, If you had one flight a month that only infected one day’s Tmax, the impact would be 1/60 in the monthly figure.
Simple question: the more flights, the higher the probability of infection and the larger the impact. True or false.
3. For the infection to be significant the temperature of the air exiting the aircraft
needs to be higher than the ambient temperature around the thermometer.
True or false
4. An jet engine with a high bypass ratio, generally puts out cooler air than an jet engine with a low by pass ratio? True or false
5. on a taxi way into the gate, the plane is typically at idle and the EGT is lower than at full power. True or false.
6. The temperature increase created by a single pass of jet will diminish with distance from the jet? true or false?
In short, if your looking for a GOOD CASE to show, to demonstrate the impact, this case is not a good one. That doesnt mean you cant demonstrate or find a case. Just this: this airport, this airplane, these conditions, you aint gunna find something CONVINCING. and when you throw up UNCONVINCING CASES, you destroy your credibility.
Paolo M. says:
September 30, 2010 at 6:46 am (Edit)
Steven Mosher doesn’t know what a stable nocturnal layer is.
#########
sorry Paolo, You might want to look at a couple things first. the collapse of ABLs at coastal locations at nightfall and the critical windspeed for mitigating UHI.
Who knows you might be able to make a case, but that would require a hypothesis, and a test that you would be willing to accept as definitive.
“Does that happen at Solomon? Don’t know. But I do know that the tarmac and concrete will hold heat from the daytime into the night (helping to raise TMIN) and will be much hotter mid day (and definitely raise TMAX on sunny days) and having an airplane deliver that heat to the box with a big fan will not make it cooler…”
I’ll suggest you do some calculations on the heat capacity of the material and how quickly it starts to give up that heat after sundown. It would make a better case than ‘i dont know” Also, glad you mentioned sunny days. When you look at UHI you find that it is mitigated by two factors windspeed ( do you know the critical windspeed) and cloudiness.
In any case, there is one study I know of that did compare a pristine station with a nearby busy airport. go read that.
You should be able to figure out very easily if the solomons uses a (tmin+tmax)/2 figure. In any case, you don’t make a better case for this location by pointing at things you don’t know. As the GHCn document you obliquely refers to mentions there are many methods for calculating a mean. you should list them.
Many of you fail to get my criticism of this case. You want strong cases. cases where you have all your facts lined up and all objects/explanations answered. You want better science than the science you are criticizing. Not conjecture and speculation and hand waving. Those types of appeals ONLY WORK on the people who are already convinced. If you want to get the attention of the people who need to be convinced, if you want your words and numbers to matter, then more needs to be done than pointing at the possibility of errors. You have to find real errors and document them. completely and thoroughly. and learn to take constructive criticism. In other words, raise your game above the game of climate science.
Anthony showed the way with surface stations.
Oh, EM,
that station looks like it is in GHCN daily with Tmax and Tmin, ( need to confirm differing numbering systems) so GHCN would calculate a (tmax+tmin)/2. but check
if the stations are really the same or just in the same location.
And who knows maybe more data will come from these folks, where it looks like the solomon islands is on the list.
http://www.wmo.int/pages/prog/gcos/aocpXV/A_13.4_wcdmp_climate_data_rescue.pdf
Steven Mosher says:
September 30, 2010 at 1:05 pm
“In any case, there is one study I know of that did compare a pristine station with a nearby busy airport. go read that.”
Can I have the link please Steven?
And one final thing
In the above example you have two sources that will be combined into one longer record. ( see the dup flag )
And The solomon islands would be surrounded by SST measurements.
In a global temperature average that takes land area into account, the solomon islands figure would be scaled by the land area in the grid. So that particular 5 degree cell has .0178 of land in it. If it was all Land the value would be 1.
BUT, since it is .0178 Land, we calculate the average for the cell like this:
First we get a land mask using land area
d <-getMask(5,"land_percent2_qd.asc")
Then we find the cell that conatins the lon/lat of the solomons
solomon <- cellFromXY(d,c(160.1,-9.4))
then we pull the fraction of the cell that is land
fract 1-.017875
[1] 0.982125
So, to calculate the temp for that cell, we do something like this
TEMP + (SST*0.982125)+ (Land*0.017875 )
So, if SST was 75 and Land was 75, the average would be 75.
And, it was asserted that a 1 degree difference was all it would take…..
well “infect” tmax by 2 degrees every day of the month and you end up with a
final corruption for that grid cell of…… 0.01788F
TEMP + (75*0.982125)+ (76*0.017875 )
Which is why, this is a bad case.
But if you like I can take the global map, I can change the solomon islands down a few degrees and you can see that this is a bad case to argue. or you could compare it to UHA and see how the sat looked over the same period.
Basically, all of those will show you this. On the vanishingly remote possibility that there is an effect here, the effect is not measureable, and if it were measurable, would have no measureable effect on the average.
Stephen,
sure. There are also a bunch of studies on using Lidar to look at aircraft plumes
and temperature profiles of plumes and exhaust ( for example, in the plume you drop from 950F to 500F in the first 40cm)
But the study you want is this: Its very limited but does give the only figures I know of. there could be more but I dont have access behind every paywall ( lots of stuff on the verticle heat profiles over different materials.. building science stuff )
http://www1.ncdc.noaa.gov/pub/data/uscrn/documentation/research/Sun.pdf
toward the end of the article as I recall.. errr lemme look
section 6, last paragraph.
Theoretically, since ASOS data is near continuous you could go get some and look specifically at the temperature profiles as planes take off and land. As it stands, this the only paragraph I have ever read in the past three years that said anything of substance ( thin as it is ) about airport/non airport differentials. FWIW.