Does your weather station get the wind? How station siting compromises temperature measurements via wind issues

Guest essay by Bob Fernley-Jones

Well Melbourne (Oz) sure does, (and we are the butt of a few intercity jokes, a polite one being; we can have “four seasons all in one day”). The following graphic (source) of an extreme example when we had tragic bushfires in 2009 shows how wind direction, strength and duration can have a dramatic effect:


My interest was perked when our nice weatherman on ABC TV news recently announced two nights in a row that Melbourne had experienced the hottest July (midwinter) on record. That’s odd I thought, because me living in a NE suburb didn’t notice anything unusual. I Googled for: melbourne + “hottest July” and found that it was viral out there. And, repeating that search today there were 15,900 hits, although there were imperfections in the results by page 10 at least. Well anyhow, next stop was the BoM temperature records for this our smallest mainland State, and:


Link 1 for the graphic above is here, and link 2 here

[Note:  Correction; temperature difference should be (only 0.36F)  BF-J]

It seems that for our State, last July was rather ordinary for about the last four decades.

Here follows a tabulation of BoM daily weather data for Melbourne for last July with my highlighting added, (see footnote for the two locations used in this dataset, which strangely are not available for St Kilda as reviewed in figure 1 and vice versa):


Notice that very few days had cool winds between SW and SE; but just one in the morning and four in the afternoon out of the 31 days. So why would the city warm more than the State as a whole because of sustained north winds from the hot centre, (deserts)? So let’s take a peek at the city weather station, courtesy of Google maps and contemplate UHI (Urban Heat Island) effect as an additive factor, in this view looking west:


This is no longer a regional weather station office as it likely was when it was set-up, and here follows some demographic stuff according to Wikipedia:


Here follows an adapted Google satellite view in which Google confusingly “enhances” traffic routings. It appears to have been taken on a Sunday afternoon:


And, here follows a Google street shot looking east, including the edge of the park to the north.   [BTW, for NH readers the midday sun is from the left, (north).]


Is there a desert near you, or cooler ocean breezes?

0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
M Courtney
September 6, 2013 1:30 pm

What is the mechanism for defining a weather station as urban or rural?
Is it just lights seen from space?
Because there must be more development near where weather monitoring people live than where no-one does.
This could be a widespread issue.

Gunga Din
September 6, 2013 1:34 pm

Yet another little spot on the globe whose numbers can’t be trusted for determining a global average.
Reminds me of a post some time ago where the official high temperature for (I think it was Boston.) the day came from a site by a runway at the moment the wind shifted to blowing across it.

September 6, 2013 1:51 pm

for NH readers the midday sun is from the right, (south)
In the SH, the sun is from the north.
BTW, those highrises will give a nice Urban Canyon Effect.

Lil Fella from OZ
September 6, 2013 2:03 pm

I suspect many weather stations in Aus are similar in abuse! I am about to check on the ones in my region. Warmest winter everywhere in Aus this year!! I just about froze to death where I live but I probably imagined it. One thing I wont imagines… going to the polling booth today!

Brian R
September 6, 2013 2:14 pm

I live and work in the SW part of Metro Denver. My office is right against the hogbacks in the area. In the summer when the wind is from the NE I’ve seen the temperature at the office 2 degF(unofficial Scion tC thermometer) higher than just 1/2 a mile away despite being a couple of hundred feet higher in elevation.

September 6, 2013 2:32 pm

The wind indicators at airports are used to choose runways and as information given to pilots. I would expect them to be accurate for that task.

September 6, 2013 2:39 pm

Our host’s Surface Stations project has highlighted this, at least in the US.
And here, in Australia.
Where else?
Interesting to see that the – politely, less-than-studious – efforts to get accurate temperatures for Global Climate appear to be a bit hit-and-miss. Although if UHI and airport tarmac feature in an unrepresentatively high number of station’s readings, I guess not every reader of WUWT would be astounded.
If the theory doesn’t fit the data – the theory is wrong. [Feynman paraphrased].
And if the data can’t be trusted [and may be open to a measure of malleability] where do we stand?
Southern England has just had a lovely summer [not too-too hot, mostly dry, and we retained the Ashes], but rain swung through today – and my fear is global cooling. Shorter growing seasons, more old folk die from hypothermia, etc . . . .

September 6, 2013 2:47 pm

Please correct my silly, as pointed out by Philip Bradley. The text above the last pic should read:
BTW, for NH readers the midday sun is from the left, (north).
And, groan, please add a note under the second figure:
Correction; temperature difference should be (only 0.36F)
Sorry, too much Cab-Merlot maybe?
[Please verify the corrected notes are proper. Mod]

Steve (Paris)
September 6, 2013 2:56 pm

This is an excellent post. One pebble at a time.

September 6, 2013 3:37 pm

The best definition of “rural” I have seen came from a friend at Statistics Canada. If you can see a cow, hear a cow or smell a cow- you are rural.

September 6, 2013 3:44 pm

“we can have “four seasons all in one day””
I live in Wisconsin, USA. The line around here is if you don’t like the weather, wait 5 minutes.

Gail Combs
September 6, 2013 4:10 pm

Auto says: @ September 6, 2013 at 2:39 pm
….. Although if UHI and airport tarmac feature in an unrepresentatively high number of station’s readings, I guess not every reader of WUWT would be astounded.
E. M. Smith did a lot of work on the airport issue:
Here are other links to information on the temperature record.
Start here and go forward for lots of articles:
Examining Urban Heat Islands

September 6, 2013 4:11 pm

The large building directly south has a white (presumably concrete) facade and its albedo will be up around 0.5 reflecting plenty of solar energy on the Stevenson Screen.
Although, had they constructed that building on the north side of the street, Melbourne would be experiencing record cold temperatures in July..

September 6, 2013 5:02 pm

Gunga Din wrote:
”…Reminds me of a post some time ago where the official high temperature for (I think it was Boston.) the day came from a site by a runway at the moment the wind shifted to blowing across it.”
Yes, interesting! The reason that the anemometer on this site is only used to record average wind-speed, (not for gust speeds or direction like elsewhere), seems to be because of turbulence effects. I guess too that southerly winds would have a screening effect from the buildings whereas the north winds are more likely to have a “runway effect” across all that paving.

September 6, 2013 5:13 pm

They’re taking the p1ss, innit?

September 6, 2013 8:50 pm

A few observations about Melbourne RO. Until June 2013, the BoM’s Melbourne monthly summary at, always contained something similar to:
Located in the Central District at the head of Port Phillip Bay, Melbourne is Victoria’s State Capital. Here, overnight minimum temperatures were slightly warmer than those usually experienced and averaged 7.4°C (departure from normal 0.5°C). That the overnight temperatures in Melbourne are higher than those in most surrounding localities is a consequence of the city being under the influence of the effect of urbanisation (cities are usually warmer than their rural surroundings, especially at night, because of heat stored in bricks and concrete and trapped between close-packed buildings).Daytime maximum temperatures were warmer than those usually experienced and averaged 15.4°C (departure from normal 1.4°C). Total rainfall for the month far was 94.8 mm, this being much more than that usually recorded (normal 49.5 mm, percentage of normal received 192%).
Some 20 kilometres northwest of the Melbourne city centre, Melbourne Airport, is more typical of the suburbs. Here, overnight minimum temperatures were near to those usually experienced and averaged 5.8°C (departure from normal -0.4°C). Daytime maximum temperatures were slightly warmer than those usually experienced and averaged 14.1°C (departure from normal 0.5°C). Total rainfall for the month was 128.2 mm, this being much more than that usually recorded (normal 39.9 mm, percentage of normal received 321%).”
As of July (, the reference to UHI has disappeared … no point confusing the readers with fact, I suppose. The acknowledged UHI is unadjusted in ACORN.
Since Melbourne’s temps have little to do with reality, you might decide to go to the BoM’s Monthly Weather Review for each Australian state ( Good luck. All Victorian summaries back to November 2012 will inform you that “This issue of the Monthly Weather Review is not yet ready for publication.” All other Australian states and territories are pretty well up to date. Ah well, Victoria is only Australia’s second largest state so there’s probably no rush for the BoM professionals to update its figures.
Funny thing is, if you go to you’ll find an audit of all 201 weather stations in the ACORN dataset, detailing the number and proportion of rounded .0 Fahrenheit raw recordings taken before Celsius conversion in 1972, and rounded .0 C since then. Probably because it was a regional office and one of the most meticulous in Australia, Melbourne had one of the country’s lowest proportion of .0F rounding at just 14.1% for min and 12.1% for max. This compares to 30.1% of all raw temperatures logged by observers as whole .0 Fahrenheit degrees before Celsius metrication on 1 September 1972.
It can be argued that if half the observers rounded up and half rounded down, it makes little difference to Australia’s historic temperature records. Fair enough as long as all observers weren’t in a hurry, had good eyesight and were just as likely to round 68.7F to 68.0F as they were to round 68.3F to 69.0F, for example.

September 6, 2013 10:17 pm

R2Dtoo says:
September 6, 2013 at 3:37 pm
The best definition of “rural” I have seen came from a friend at Statistics Canada. If you can see a cow, hear a cow or smell a cow- you are rural.
Does not work for Phoenix Arizona.

Ken Stewart
September 6, 2013 10:26 pm

Specifications for weather stations in Australia state that the instrument enclosure should be at a distance from nearby non-complying surface e.g.pavement, roads etc, of at least 5 times the width of the non-complying surface. The siting of this station is laughable, surrounded by pavement, sheltered from cooling south winds by buildings and open to hot north winds. And Melbourne is BOM’s Australian HQ. What a joke.

September 7, 2013 12:32 am

I thought I could not be more shocked than I already was until I saw your comment:
I have a feeling that the BoM suffers from a lack of sincerity or something.

September 7, 2013 12:43 am

Ken Stewart,
I see that you too are not overly impressed by the Melbourne weather station siting. It’s almost tempting to enquire of Dr Karl Braganza who has been the “Head of Climate Monitoring at the Bureau of Meteorology’s National Climate Centre” since 2004, if he could clarify the situation.
This seems to be his current email address:

Ken B
September 7, 2013 6:02 am

Well we have a new Conservative Government under Tony Abbott, after a rout of the Labor parties and what looks like some reduction of the green voter influence!!
Hopefully the Climate Change Commission will be gone with Flannery’s salary a saving, and perhaps time for a few revealing reviews. Australians, particularly older Australians want honestly in Government and with the Carbon Tax on the way out, time to put some warmth to the debate on the science and the scandalous treatment of data and the propaganda produced in support of taxing regimes. Perhaps time for a campaign to exonerate C02 and bring some real scientific experts into the debate.
May this result be some comfort and lead to our American friends as they resist being dragged back into a carbon trading mess on the pretext of dealing with a beneficial trace gas.
More power to Murry Salby to move past the roadblocks and present his material to the new government, we should help and encourage this to happen as then there would be no need to play the advocacy games the Australian BoM and C.S.I.R.O. have been engaged in over so many years.
May science be placed in the forefront to advance rather than retard knowledge!!

Brian H
September 20, 2013 2:20 am

Edit: “my interest was perked …” piqued

%d bloggers like this: