The official temperature in Sydney is measured at the Sydney Observatory.

enter image description here

This is located near the city side of the Sydney Harbour Bridge - which is a major highway.

My question is Does measuring the air temperature near a 21 lane asphalt highway impact measurements? What is the reasoning?

  • $\begingroup$ This question would perhaps fit better at the Earth Science sister site. $\endgroup$ – David Hammen Oct 9 '14 at 23:24

There are, by my count, 33 other official weather observation stations within Sydney. Here's a map:


With regard to the Sydney - Observation Hills (Station ID 066062) station -- That freeway didn't exist when the station was built. How could it? That station dates back to 1858. The station was moved 150 meters to its current location in 1917. That tiny move resulted in a rather significant temperature jump.

Meteorologists have seen this happen multiple times. They finally learned their lesson, that moving a meteorological station is a very bad idea. Moving a station is equivalent to removing an old station and adding a new station. Continuity is lost.

On the other hand, there's a problem with keeping some of those old stations. Buildings, parking lots, and multi-lane highways grow up around them. That Sydney station is rather poorly-sited. The problem isn't the highway so much as the extremely close parking lot. US standard call for a station being sited 100 feet (30.48 meters) from a paved area. There's a parking lot within 9 meters of the Stephenson screen.

There are a number of options with regard to stations that have over time become poorly-sited:

  • Use these poorly-sited stations as an excuse to ignore all evidence of climate change. This doesn't fly scientifically, but pictures like yours and even worse (and there are far worse) do work to cast doubt with regard to climate change in the minds of the non-scientific public and elected officials.

  • Adjust the recorded temperatures based on statistical analysis. When one site starts deviating markedly from other nearby sites, that's a sign of a biased site. The site data can still be useful if adjusted for that bias. This is the approach taken by most professional climatologists.

  • Throw out the data from those poorly-sited stations. This was the approach taken by the Berkeley Earth Surface Temperature project. This project was started to disprove the claims those professional climatologists cited above. Something funny happened along the way to proving those pesky climatologists wrong: They found that those pesky climatologists were right. One of the things the BEST team investigated was whether poorly-sited stations resulted in erroneous claims of climate change. They threw out all measurements from suspect stations. The result was a tiny difference between the warming claimed by climate scientists. That tiny difference was well within uncertainties. There is no statistical different between adjusting the temperatures recorded by those poorly-sited stations versus discarding all data from those stations.

  • $\begingroup$ You got to the question behind the question. Much appreciated. $\endgroup$ – hawkeye Oct 10 '14 at 1:37

It no doubt does. Solar radiation will heat surfaces, and the rate at which a black surface absorbs that heat is greater than for other colors. That heat is then partly lost by radiation, partly by conduction, and partly by convection. If you assume the last two terms will be the same regardless of the color, then you can see why a black surface will become relatively hotter than other surfaces (which reflect more heat).

These hotter surfaces affect the local temperature: maybe not by much, but certainly a measurable amount. If you have ever see the "shimmer" of the air above a hot road surface, you will have observed the impact of the heating of the air on the refractive index of that air - a visible reminder that sunlight on dark roads heats the air (image source). enter image description here

So if air locally gets heated, it stands to reason that temperature measurements nearby will be biased. A 21 lane highway in a sunny area is likely to bias the measurement.

There are many instances where local geography can affect temperature. The size of the artifact giving rise to a variation will affect the distance over which it can affect the measurement: a 21 lane road may affect temperatures measurably within a 50 m range, while temperatures at the edge of Lake Michigan can be lower than those measured further inland for several miles.

From the wikipedia article on "urban heat island" effect (my emphasis):

An urban heat island (UHI) is a metropolitan area that is significantly warmer than its surrounding rural areas due to human activities. The phenomenon was first investigated and described by Luke Howard in the 1810s, although he was not the one to name the phenomenon.1 The temperature difference usually is larger at night than during the day, and is most apparent when winds are weak. UHI is most noticeable during the summer and winter. The main cause of the urban heat island effect is from the modification of land surfaces, which use materials that effectively store short-wave radiation .[2][3] Waste heat generated by energy usage is a secondary contributor.

Replacing trees with tarmac will increase the apparent temperature in the vicinity.

  • 1
    $\begingroup$ The real question -- absent wind, does the temperature of the highway affect something a few hundred feet away laterally and 6 feet high? Even accounting for wind, perhaps the predominant wind blows the heat away from the sensor? $\endgroup$ – tpg2114 Oct 9 '14 at 22:02
  • $\begingroup$ Awesome reasoning - much appreciated. $\endgroup$ – hawkeye Oct 10 '14 at 1:36
  • $\begingroup$ Highways, roads, asbestos rooftops, countless air conditioners, etc. do make cities warmer than the surrounding countryside. One issue is how to best avoid exaggerating that real urban heat island effect unduly by siting a station right next to a parking lot or highway. Another issue is how best to avoid making that real but local effect from biasing climate estimates. These two issues are a bit at odds with one another. Downtown Sydney truly is hotter than the nearby New South Wales countryside. But incorporating that urban heat island effect into climate studies may result in bias. $\endgroup$ – David Hammen Oct 10 '14 at 6:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.