Today has been electrically charged, first my interview with Andrew Bolt and Steve Price on their radio show in Melbourne, then rush to the airport, trying to do a radio interview on a cellphone while checking in, make a quick post on the skeptic blacklist, off to Hobart, then discovering your luggage never made the plane.
All was not lost though (well it might be if I don’t get my luggage). The first stop my trusty guide Alan made was to the Hobart Weather Station next to the Anglesea Army Barracks at Battery Point. He said, “you have to see this”.
Looks pretty nice doesn’t it? It is very picturesque looking out over the bay and the Wrest Hotel/Casino (the tall building).
This station was showcased by the ABC radio network om 30 May 2008, here and ABC supplied this photo with the story:
Notice anything interesting? I did, but nary a mention from the ABC. They write:
“The Battery Point site being the official site is where temperatures are verified against” said Malcolm.
Malcolm went on to explain that one of the limitations with the Battery Point site is that due to its location it will be affected by the sea breeze thus the temperature may not climb as high as places further inland such as Glenorchy or Brighton.
The major change in the operation of the site over the years is that where once a person would be on hand to measure the temperature, these days everything is done electronically.
In regards to the future of the site Malcolm said it is important from the Bureaus perspective that conditions remain as constant as possible as any major changes to buildings or even trees surrounding the site can influence the temperature readings.
Whilst such changes may only alter the readings by a tenth of one degree it is just such evidence that the Bureau need to establish long term trends in climate change.
They didn’t mention the most important feature – air conditioners. Lots of them. Here’s more of my photos:
Here’s a reverse angle, a composite of two photographs to take in the wide angle:
Here, David Archibald poses next to the Stevenson Screen and the A/C heat exchanger units:
The industrial sized unit in the foreground was working so hard it had iced up its coils. David scraped about an inch of frost off of it.
Here’s the aerial view.
Note that just measuring the distance to asphalt and the nearest building, the station is less than 10 meters away, making it a CRN4 station, which would be considered unacceptable by NOAA standards. It would fail by either the old 100 foot (30 meter) rule, or the new Climate Reference Network siting rule.I don’t know how much of the building built up around it or when, but it clearly fails.
Here’s the temperature data, via NASA GISTEMP:
The jump around 1970 may be of interest related to siting, but without more time to research that metadata I can’t speculate if it is related or not. Note the plunge though the last two years. Quite a drop.
And here is what the data looks like after GISS finishes with their “homogenization” adjustment:
While I don’t have time right now to do a full analysis as I’m due for a metting shortly, I can say it appears that GISS flattened out the could snap in the 1940-1960 period, making the long term slope more positive. I’ll look at that later.
The point here is, it seems no matter where I go in the world, I seem to find siting issues with official weather stations used for climate monitoring. Stations that are long period records of historical importance suffer the most from such siting issues, because their record is valuable. Worse, when the data is adjusted, it seems to add to the warming.
This station is NOT used in creating the Australian climate record as it is regarded as Urban. But the 5 stations that are used by BOM have a 40% warming bias in the homogenized record. Watch for my next post on Tasmania today or tomorrow at kenskingdom.wordpress.com
Ken
wayne said: June 22, 2010 at 11:41 pm
“Yep, everybody added A/C in the 60′s, check the graphs! Coincidence?”
Nope! Tasmania staunchly remains a decade or two behind the “civilised” world 🙂
Given the paucity of hot days here I have no intention of installing A/C in the foreseeable future. But then again, Gits are not everybody 😉
Soren Rosedahl Jensen (1:46am):
You downplay the fact that “homogenization” of the Hobart RO record has increased the fitted linear trend from 1893-1992 by a factor of ~1.6 by appealing to mysterious “confidence intervals.” CIs ordinarily refer to the sampling uncertainty of the available estimate relative to the true (population) value. What is that “population” in the present instance? Is it other century-long stretches of (nonexistent) temperature record at Hobart, or is it other realizations of “Hobart” over the same years, but in (nonexistent)parallel universes? Please explain.
I am probably more on top of this site than most, because I live practically on top of it. Well, 200m away.
Interesting site. The green rectangular area intruding into the image at top right appears to be a bowling green (for lawn bowls). To the west, is the Anglesea Barracks, with a large area of asphalt (parade ground) as well as a lot of car parking. I have been in this area to a function at the Officers’ Mess, but have limited knowledge of it. The function of the barracks has changed considerably with time, and I do not think it is used as actively now as in the past. To the south is a large building that was once a school and is now a number of residential apartments.
On air conditioning: these would definitely be what we call heat pumps. Hobart has some summer days when air conditioning is desirable, but these are relatively few. We use these more in winter for heating than in summer (they are about 300% efficient), but have only done so extensively for 20-odd years. Low electricity prices historically and pricing tariffs such as off-peak mean that forms of heating such as ‘heat banks’ or off-peak heat storage were popular before then, but we are now interconnected with the Australian mainland and sell our valuable (predominantly hydro) electricity into a national market.
I was sorry to miss Anthony’s meeting last night, if only to meet him. (I’m already familiar with his work). But I’m just back from Europe, and badly jet-lagged, and it was past my bed-time!
“…I can say it appears that GISS flattened out the could snap in the 1940-1960 period…”
British warmist propagandists are given to calling the worst winter in x years a “cold snap.” I see you’ve bought into their terminology. But it’s wrong.
Cold snap: a sudden, brief spell of cold weather
http://www.yourdictionary.com/cold-snap
So the Ellerslie Road site is still bedevilled by warm air falling and flowing downhill across the lawn? I suggest calling a priest.
Here is a really humorous thought.
Why don’t we work on the house and senate to reinforce a new regime in temp taking and give it back to the amateurs rather than the governmental idiots that run things now?
After all temps across the world used to be taken and transposed to the local weather beureaus for a very long time before they hosed it with government intervention.
Nick said @ur momisugly June 23, 2010 at 9:31 pm
“So the Ellerslie Road site is still bedevilled by warm air falling and flowing downhill across the lawn? I suggest calling a priest.”
I suggest we call the priest: witch doctor 😉
My point was only to show what I stated:
That the trend estimates for the two dataset where within the confidence intervals of each other.
Note the big uncertainties on the trend estimates.
Anthony,
See also http://www.bom.gov.au/cgi-bin/climate/hqsites/site_data.cgi?variable=maxT&area=aus&station=094029&period=annual&dtype=raw&ave_yr=T for this site as part of the Australian high quality climate site data
I am glad you enjoyed the visit to the station, and perhaps you will do a story on the the banner you photographed, as we left the Barracks, of a local radical green group (part of a Climate Action coalition) opposing renewable energy from biomass in the form of harvesting residues and sawmill offcuts from Tasmania’s sustainable forest management [about half the island is still forested with half managed for conservation, a third privately owned and 20% managed for timber production by a government agency.
Soren Rosedahl Jensen (4:12am):
Confidence intervals are meaningful only in the the context of statistical sampling from an ensemble of time-series or a population. When there’s no missing data, the “trend” fitted by linear regression over a set period of time at a particular station is not a statistical sample, but an entirely deterministic calculation! It is EXACT in the same way that an exhaustive census of the population is exact. No matter their range, confidence intervals based on the usual iid ASSUMPTION of regression analysis are meaningless in that context. They obscure the issue at hand, which is that any “homogenization” MANUFACTURES a time-series with a SYSTEMATIC bias. It is a transparent attempt to mislead with pseudo-statistical nonsense.
In Burnie they take the readings 5kms east of town where the sea breeze is very pronounced. Some days in summer are 25C + yet on the news they say “Burnie had 19C”. Great for tourism right!
Wouldn’t all those building block the “cooling” sea breeze?
ahhh Tassie. a weather station in the direct line of at least five a/c units! gotta love it
I concur with “Sky” regarding inapplicability of standard “statistical regression analysis” to time series from individual stations. These data are not an ensemble (which would be true if the data can be repeated under the same starting conditions), and not independent members of temperature “population”. The time series at a station is a realization of a completely deterministic dynamic process (if one includes presumably known land use change in the vicinity). The variance in individual samples is not a result of measurement errors with known Gaussian random noise. At most, the measurement error might come from a local thermometer. Assuming periodical calibration of equipment, error of each individual data sample does not exceed 1C. For a selected metrics of “yearly average”, the result comes from 730 samples, such that the error in this average is about SQRT(730)=27 smaller than the 1C individual error, or about 0.03C. Therefore, for a given time period the slope of linear fit is a very precise characteristic at that particular location. If you wish, it is a unique “signature” of it, and if a station’s reading are drifting down monotonically over 100 years span, they ARE drifting with very high confidence.
P.S. This post is in partial response to authoritarian opinion of one “professional statistician” who wrote about Sky’s post: [Response: Well, I’ve already used “LOL” and “ROTFLMAO” — what more is there to say?]
and my other observation about stations in Texas:
[Response: He hasn’t thought things through at all. Nor has he applied any statistics. He’s just thinks he knows everything.
He gives the example of different warming rates for Albany, TX and Haskell, TX. Ask him if that difference is statistically significant, and at what level of significance. After he answers, ask him how he compensated for red noise.]