Air conditioning trends in the USA

Lately, it seems that I’ve been finding air conditioners juxtaposed with temperature sensors for USHCN climate stations of record all over the USA.

Bainbridge, GA USHCN Climate Station of Record

That got me to wondering; what sorts of trends are there for air conditioners in the United States? And, could there possibly be any correlation between surface temperature measurements to the number of air conditioners in use in the USA?

Some research led me to a Dept. of Energy’s “Energy Information Administration” (EIA) website which had some interesting facts, some of which I’ve graphed to show trend.

The EIA website only had data as current as 1997 for some reason, but it did go back to 1978, though apparently surveys weren’t done every year.

By 1997, nearly three-quarters (72 percent) of all American households had air-conditioners. 47 percent of all households had central air-conditioning systems and 25 percent had window/wall air-conditioners. 1 percent had both central and window/wall air-conditioners. I would expect that number to be at 80 percent or higher by 2007, in part due to the availability of very inexpensive a/c units manufactured in China, Taiwan, and Korea, some of which can be had for about $100 US.

By 1997, over nine-tenths (93 percent) of the households in the South Atlantic Census Division had air conditioners. Over half (54 percent) of all the households in the division had air conditioners and used them all summer.

From 1978 to 1997, the total amount of electricity used in the residential sector increased from 2.47 quadrillion Btu in 1978 to 3.54 quadrillion Btu. Over the same period, electricity used for residential air-conditioning rose from 0.31 quadrillion Btu to 0.42 quadrillion Btu. Among the reasons air-conditioning electricity use did not rise more is the increasing efficiency of air-conditioning equipment. You can see the electric use trend in the graph below:


This correlates somewhat with the number of cooling degree days, but surprisingly, the trend went downward over the last few years of the study:


The gap between the “haves” and the “have nots” when it comes to air conditioning has been steadily closing. Interestingly, the trend of households with a/c units looks similar to some of the surface temperature trends that have been published:


But it makes you wonder, what effects do the millions of air conditioners dumping waste heat into the near surface atmosphere have on temperatures measured at about the same elevation as the waste heat is dumped? Is there enough atmospheric mixing to distribute it so that it becomes part of the entire UHI bias, or does it dwell in “pockets”? Sounds like a start for a study for somebody.


newest oldest most voted
Notify of

Some places, like the west coast of Washington, don’t really need AC. It doesn’t get nearly as hot for nearly as long as in the central and eastern part of the state. So people only really use AC for a few days a year, unless they’re just gluttons for high electric bills.
Last year when we had a “heat wave” (stayed in the 90s for a few days) people went and bought a bunch of portable AC units, only to try and return them when it was over.

Jeff Wood

I live in the countryside, and when I go to the city I really notice the difference in temperature between the two. Go into the back alleys where the shop air conditioning units vent and one soon appreciates the cause.
Follow the money. I was puzzled that for the last few years there has been no summer drop in oil prices in the UK, indeed often a rise. My newspaper explained a year ago that as air conditioning spreads, there is a greater summer demand for oil than in the winter. A remarkable datum.

Steven Jones

“Air Conditioners juxtaposed with temperature sensors for USHCN climate stations of record all over the USA.”
I just wonder if the A/C units being so close to the sensors are having a direct impact on the sensors.


This same data could be used to make an argument in the other direction. I could see Al Gore in front of a similar chart saying that producing electricity to run all those air conditioners is causing global warming.

David Walton

I think Will is on to something, “Al Gore in front of a similar chart saying that producing electricity to run all those air conditioners is causing global warming.”
All those air conditioners in close proximity to temperature measuring equipment and corrupting the data could be the cause of global warming.

David Walton

I don’t think air conditioners alone would be make for a reasonably broad Urban Heat Island effect study since there are so many other significant sources to also consider. But focusing on their bias effects would likely spill over into how heat from other local sources can effect temperature measurements.
Why not a simple aggregate study of a set of sensors with known heat island problems against a control group not influenced by such effects?
Just a thought. The comparison could be made entirely with already existing data by simply categorizing and separating data sources into groups.


Since we’re talking about tenths of a degree celsius, I would certainly think that hot air from an AC unit would bias the sensor’s output by measurable degree. The same way a particularly cloudy day would, but for the opposite reason (natural vs mechanical)


Is “power expended” a good measure of AC output? Doesn’t one have to measure the “coldness” in order to offset the heat?
Is it possible that a modern AC can affect the readings more than an old AC–while actually using less power?


While modern air conditioners are probably more efficient that the older ones, power expended is still an accurate measure since that power has to go somewhere, in this case either to moving heat indoors to out, or as waste heat.