Sydney's historic weather station: 150 meters makes all the difference

Here is an interesting story about the weather station in Sydney at the Astronomical Observatory (well sort of, it got bumped off). It seems the astronomers and meteorologists at the observatory got into a tug of war in 1912 over a cottage and the resulting move in 1917 ended up skewing the entire temperature record irreparably. First, here is a photo of what the observatory and grounds looked like “back in the day”. While I can’t be certain, the latticed canopy at the left may be where they had the early thermometer exposure. Early attempts at sheltering the thermometer from sunlight were often large affairs like this. It may also simply be a lounge area, but the design doesn’t look relevant for that.

Photo circa 1874, more photos and history here

Now here is the article of interest from The Sydney Morning Herald:

How hot heads caused climate change

Richard Macey

July 2, 2008

NINETY years ago Sydney’s temperature took a leap. However, it had more to do with rising political, rather than global, heat.

The instant climate change was remembered yesterday as Bureau of Meteorology staff, and astronomers, gathered on Observatory Hill to mark 150 years of Sydney’s weather observations.

On July 1, 1858, NSW’s government astronomer, Reverend William Scott, noted 12.7 millimetres of rain had fallen in the previous 24 hours. “He was the astronomer, the timekeeper and the meteorologist,” the current astronomer, Nick Lomb, recalled yesterday.

In 1908 the role of recording rainfall and temperatures was taken over by the Commonwealth’s new Bureau of Meteorology, which was allowed to share accommodation in the observatory’s grounds.

However, Dr Lomb said, it was a tense relationship. William Cooke, appointed in 1912 as the new astronomer, wanted the bureau out so he could move his family into the observatory’s occupied residence. “A battle between the two sides” ensued.

The astronomer triumphed and in 1917 the bureau was moved 150 metres south, to an old cottage across the hill.

The seemingly insignificant shift triggered a sudden spike in Sydney’s temperature records.

A 1972 study by meteorologists Rosea Kemp and John Armstrong found that since 1918 Sydney’s average annual maximum temperature, as recorded at the new site, was 0.7 degrees warmer than the average at the old site. Winter averages were up 1.6 degrees.

Dave Williams, a senior bureau meteorologist, said the new site was sheltered from cold winds by buildings, creating a micro-climate. “It’s quite noticeable in the winter months,” he said.

While some of the increase probably resulted from warming caused by Sydney’s urban development, the artificial spike has meant Observatory Hill’s statistics cannot be included in climate change studies.

In 1989 a young bureau observer, Mike De Salis, inherited the job of making daily visits to collect rain and temperature readings. On cold, wet days it could be a miserable task. “The worst part was trying to measure the rainfall accurately, within 0.2 of a millimetre, while it was pouring rain,” Mr De Salis said.

For yesterday’s 150th anniversary, Mr De Salis returned to again make the manual observations. The rain gauge had detected just 0.2 millimetres.

An exhibition of meteorology equipment opened yesterday at the observatory.

(h/t Barry Hearn)

I’ve located the station on Google Earth. I’ve posted three annotated views below but you can also locate it here at this live link.

First the wide view with the Google Earth measuring tool showing the distance from the Astronomical Observatory. Note the depression in the earth and the proximity of the freeway toll plaza.

Click for larger image

Now here’s a closer view showing the complex of buildings mentioned in the article. The Stevenson screen with the thermometer is in a small yard on the right side of the complex, closest to the freeway:

Click for a larger image

 And finally the close-up view where I measure the distance to the asphalt cul de sac to the screen, not also the concrete walkway:

Click for a larger image

Based on distance to artificial heat sources such as the asphalt and buildings, this station would rate a CRN4, and could be argued to be a CRN5 due to proximity to the walkway, which appears to be concrete, but I can’t verify that from these photos.

As mentioned in the article, the Australian Bureau of Meteorology wisely chose to exclude this station from “climatic studies”.

However that doesn’t stop Dr. James Hansen of NASA GISS from using it, as it is in fact part of his GISTEMP database, see the plot below from GISS:

See the source plot and data here at NASA GISTEMP database. Note the lat/lon matches that of the lat/lon labels in lower left of the Google Earth images I posted above.

It makes me wonder; if the station is not good enough for BoM to use, why would GISS use it?


newest oldest most voted
Notify of

The Sydney Morning Herald is a newspaper read avidly by Mr Gore whenever he jets into Sydney!

Leon Brozyna

Thanks for that Google link. Zooming out I see how close the site is to the bay. So the pre-move location was freshened by wind off the bay resulting in stable temp readings. After the move the data becomes more volatile. I would suppose some years it would be more sheltered from wind while in other years it would happen to catch more wind in its depressed site. I also noticed that despite this volatility, the temps were fairly stable and even declining slightly until the post WWII era. With its depressed location and all that post-WWII construction this qualifies as a poster child for the UHI effect.
Good to see that the Australians had the good sense to exlude this site from data on climate studies.
Maybe the folks at GISS should use some of their supposed computer savvy and learn how to exploit Google Earth.

It’s obvious isn’t it. Don’t let the facts get in the way of a good story.


As this a thread about Australia, I’d like to throw in something that astonished me, frost is forecast for Monday and Tuesday next week in Perth. This may sound unremarkable, but Perth’s city weather site has never had a frost in 150 years.
BTW, in the last 2 years we set 2 consecutive record low temperatures. From memory the current record low is 0.6C set last winter.
REPLY: Lut us know if it comes to pass.

Amazing how such little changes can skew results.

Daniel Rothenberg

I think you’re asking the wrong question. What you’ve got here is a station with an extremely detailed history as to observation times, instrumentation, siting changes, and all sorts of different variables. Because of this, it’s a prime candidate for use in climate change studies and datasets. When a station has a very detailed history, it is much easier to cut out the extraneous noise from the data; in other words, it is much easier to calibrate a station when you definitively know its source of biases.
Compare this to the Tucumcari station you recently posted about. Have we yet identified the exact cause of the spike towards the end of the data set? In a way, that uncertainty as to the nature of the bias in the dataset makes it less than favorable for inclusion in a temperature record data set. Here, though, we very readily know that the new siting of the observation is in a micro-climate with a warmer trend than that of its previous siting. Hence, we know the bias of the station, and we should more accurately be able to adjust the data.
The article doesn’t specifically say what sort of studies the BoM omits this data from. However, for the GISS dataset, as many stations as possible is the modus operandi. Even if those data sets need to be analyzed and scrutinized for bias, it’s better that this process be done than exclude potentially valuable data.


If that’s meteorological equipment in the front yard of the observatory (or anywhere nearby, actually) in the 1874 photo, then wouldn’t the surrounding vegetation, buildings and driveways have made the station a CRN4/5 at the beginning?

steven mosher

hansen uses it because he never took the time to check. he was too busy saving the world. snark off.

Harold Vance

All your hockey sticks are belong to GISS!


The BoM in Australia has an array of “high quality” stations, more than 100 in number, that it uses for climatic trends. Sydney is not one of them. All of them are still actively reporting, yet GISS stopped colecting data from most of them in the early 1990s, while keeping Sydney and various airports.
You can see that GISS also stopped collecting data from Sydney in 1992, but it still collects data from Sydney airport, which has a huge UHI.
(I’m not sure how high quality the BoM network really is. Many are towns of various size or small regional airports.)

Scott Covert

The cherry tree is full of AGW. Maybe Hansen cut down the cherry tree.

Wanna bet that temperature rise tracks the AADT of the adjacent motorway almost 1:1?


Fanatics do not possess rational reasons for doing the things they do. The end justifies the means.
No matter how crazy those means may appear to outsiders.

James Lane

The site is not in a depression. The “ring” around it is a loop ramp that joins traffic from the Cahill Expressway to the Harbour Bridge.
If the modern site is sheltered from the West, that would explain the sharp increase in winter temperatures at the time of the move, as cold westerlies are the prevailing wind in Sydney in winter.
The UHI impact on this site must be enormous, as it has transformed from a parkland location to being a few meters from the busiest piece of road in Australia. There was no adjacent roadway until the Bridge was constructed in the 1930s.
I believe Phil Jones (HADcru) dumped Sydney observatory in favour of Sydney Airport many years ago (the airport having its own UHI issues).
There is one splendid site in sydney – Fort Denison, which is an island in the harbour. Unfortunately, it no longer collects temp data (if, indeed, it ever did).


And speaking of crazy, I read on ScienceDaily how nutsedge grows better in an artificially CO2-enriched environment and are we ever going to be sorry when our CO2 levels rise.
I would have thought that there would be an adult somewhere on the staff that would have said “And the point is? All of the plants will be in an elevated CO2 environment and will grow better, so the nutsedge won’t be any more invasive than it already is” but, sadly, I was expecting too much.

Mike Bryant

I can almost see the gears turning in hansen’s mind… “We keep and accentuate the data from airports, sewage treatment plants, parking lots and anything that smacks of progress. I can push the world back to the dark ages using their own progress against them…”

The bias in this whole tale, did they actually think no one would figure it out, and report on it? Or do you think they were depending on the media to hold the line and make the story stick? While that has happened, the story has far from stuck.
It’s as if no one understands what is meant by 1918 levels of CO2 production. Energy use has contributed to man’s life extension, that fact cannot be denied. Aren’t humans to be treated as part of the ‘natural environment’? Humans aren’t aliens afterall.


“the artificial spike has meant Observatory Hill’s statistics cannot be included in climate change studies”
This is the problem with ALL the station data. I bet if you had detailed history on every site, all of them would have some major instrument/siting change that would significantly contaminate the data. And any attempt to “adjust” the data to remove that contamination is just a guess. The effect of siting, the adjacent tollroad, general UHI, the sidewalk, the parking lot, etc etc, are not a one-step or linear process.


Sydney’s surface temps it can be argued teleconnects with the strip bark bristle cones in California /sracasm off.

Russ R.

GISS uses urban sites because they need to be “adjusted”. I don’t see how you can manufacture an emergency scenerio, with the clean temps. They fall well within the bounds of moderate warming, we have enjoyed since the end of the LIA.


Russ R, moderate warming since the end of the LIA will not get you a million in funding from environmental activist organizations and George Soros.
It’s all in the money.


According to what i can gather from CA, GISS adjusts the upwards of half their urban site warmer rather than cooler.
Just how does GISS choose their sites, anyway? (Not the first time THAT question has been asked.)

I’vé made a couple of comments on the Australian RCS in the past. A lot are at airports, though most not at all major. Some of the stations appear to be excellent, others might not quite make Anthony’s 1 rating! Have a browse through the BOM site photos at
For instance Canberra’s RCS is at the airport (Google Earth Lat long):
35deg 18min 17.45sec S 149deg 12min 04.78 sec E
You can see the new car park under construction next to it (now finished) on Google Earth, and it is down (prevailing) wind from the RAAF Tarmac. FWIW, Canberra just had its second hottest June on record…


All Australian state capitals, with the partial exception of Adelaide, are located on large bodies of water about 10Ks from the ocean.
The water will create a micro-climate that rapidly decreases with distance.
Over time 2 things have happened. One is that as CBDs get built up weather stations get moved to less urbanized locations further away from the water. This happened in Perth in 1992 and the move took the station about one K further from the water.
The second is that the water gets moved further away due to land reclamation. I’d have to check old maps to be certain, but at some point Perth’s river shoreline has been moved at least 300 meters away from the CBD.
scott, most of the airports are not airports in the sense a North American would recognize. I am familiar with a number of them. Most are 90% grass or bare earth with a single concrete strip and a couple of modest buildings.
One of the airports in the network, Halls Creek, doesn’t have any buildings, at least it didn’t when I was last there 10 years ago.
I very much doubt there has been UHI creep at any of the airports I know. I’d say they will be some of the best stations in Australia. Except, I think some of them are irrigated. I recall Geraldton Airport had nice green grass in the middle of summer.


Geraldton Airport is definitely irrigated. You can see the contrast between the green grass at the airport and nearby brown and reddish areas which will be unirrigated at the link below.
Sigh. Changes to the irrigation schedule will swamp any natural temperature trend. And changes will have occured in recent years because the government has mandated irrigation restrictions that even apply to bore (well) water, which is likely used at Geraldton.
Carnarvon airport doesn’t look irrigated, although there is extensive agricultural irrigation nearby.
I know that Halls Creek airport isn’t irrigated, nor is Meekathara airport and neither have nearby agriculture.
I’ll try and put together a comparison of these 4 airports to get a feel for the effect of irrigation.,GGLL:2008-22,GGLL:en%26sa%3DX


Stand by for significant Global Warming in Melbourne! The Laverton air force base that is the site of the nearest Australian reference site has just been sold for medium-density housing, so that all the building/paving/heat source addition should cause the temperature to rise markedly. Might almost be enough to make up for global cooling.


Phillip_B, agree re contrast between AUS and US airports, but there is a lot of potential bias at sites like TVL and WAG and other RAAF/RAN sites (like CB or NAS Nowra).
I’d go for the inland sites, or places like Cape Leeuwin or Cabramurra which don’t show much UHI normally!
On the whole they may be OK, but changes like those I’ve noted at CB are quite clear, but may not be tracked any better than those exposed by

[…] the fact that the inconsistency of the data made it no good for usage in long term climate studies, Anthony Watts claims that noted “global warming expert”, Dr. James Hanson used the data when making one of […]