The Air Conditioner Effect – Part of the Urban Heat Island Effect

I would point people at this article from 2010 as a starting point.

Think about this:

We wouldn’t put meteorological stations (met stations) right by the exhaust of smoke stacks, would we?  Of course not.  And why not?  The air coming out is HOT.  And what is the air coming out of an air conditioner’s outside compressor unit like?  HOT.

The smokestack effluent is hot because of some heating process.  Normally there is some process in a plant that uses heat and that heat is disposed of out the stack.

Similarly, an air conditioner has a process going on within IT that warms up its exhaust air.  That process is the operating of the compressor unit.  The compressor unit essentially forces a building’s inside air through nozzles, and as the air passes through the nozzle it expands rapidly – which cools THAT air.  But that cooling doesn’t come free.  Compressing that air THROUGH the nozzle means heating it up FIRST.  It has to have a certain velocity in order for the cooling to happen as it exits the nozzle.  So pressure is needed.  And when that coolant is compressed, it heats up and up and up.  Compressors run quite hot, in fact.  And they need to be cooled down.  That is the other side of the air conditioning equation – if you want cool on one side (inside) you have to deal with the heat produced on the other side, and send that heat outside.  The last stage is to run the heated fluid through tubes in front of a fan which blows over them, cooling the fluid in those tubes.  And where does that heat end up going?


(From Notice that orange/yellow arrow representing a heated air plume being sent out into the external atmosphere.  Now do that tens of thousands or hundreds of thousands of times within a city area.  The insides of buildings get cooler, and the outside – what happens to IT? The heat gets dispersed and blended with the ambient air, warming it.  That warming is not zero. Not individually nor collectively.  Collectively those non-zeros adds up.

That is why you don’t put a met station right by the exhaust of an air conditioner – because the air there is HEATED.  So that we can be cool inside, the outside has to get warmer.  In short, the air in our cities is cooler inside and warmer outside.

Now guess where the met stations are – inside or outside?  Right – outside. <i>Where the heated air conditioner exhaust goes.</i>

I do not propose that air conditioners are the <i>major</i> factor in the urban heat island effect (UHI).  However, I do think it is certainly A factor that contributes to the UHI effect – an unrecognized, under-appreciated and under-addressed effect.  After all, if we spend X amount of energy removing that heat, even the greens should be wondering what the effect is.

This blog post by Steven Goddard points to lost heat in the wintertime. Certainly that is also a factor, and I applaud his observation.  Both lost wintertime heat and air conditioners should be addressed as part of the situation with global warming as it is measured.

Recent news articles and blog posts have pointed very accusing fingers at the adjustments that continue to be made to past temperature data.  Most of us think that once data is collected and is part of the collected history that it stays at the values it had when measured, don’t we?  But that isn’t the case.  In climatology they oddly use the present as a fixed value and CHANGE the values of the past relative to that present measured temperature.  They do this instead of having a fixed point in the past that is the basis for changes.  (This would mean, e.g., that a 50.7°C month in 1934 in Asunción, Paraguay would always be 50.7°C.  Or that, once adjusted, that the adjustment would be there forever for everyone to see and would be constant.  But that is not what is done.  Instead, the climate guys compare 1934 to whatever the present year is and devise a NEW adjustment amount for that year.)  What ends up happening is that it gives the climatologists free rein to adjust past temperatures EVERY YEAR.  If you think that is a sound practice, I sure don’t.

So, with all of that going on, I am going to throw air conditioners into the mix, too.  The BIG Kahuna is those adjustments, but the NON-adjustments of UHI and the Air Conditioner Effect are also important in the overall picture.

By their very design, air conditioners take the heat out of buildings and expel it out of doors. We all know the effect if we have ever stood near the exhaust of an air conditioner. (<i>That is the reason that met stations are not supposed to be sited near air conditioners, so they obviously recognize the problem.  And if ONE nearby air conditioner can have an effect, it is foolhardy to not consider the accumulation of – the density of –  air conditioners in any city.  Placing the met stations away from the direct blast may not be sufficient, not if there are air conditioners everywhere.</i>)

So, there already IS an Air conditioner Heat Island (ACHI) effect, and they at least nominally try to adjust/allow for that – simply by trying to not have the met stations close to air conditioner exhausts.  They do this knowing that though those ACHIs are small in area, they do exist.

So what I am talking about is merely a wider acceptance of this ACHI as a part of the UHI.


With each air conditioner, the total heat in the indoor-outdoor system is more or less the same. (Only the waste heat of the compressor is added to the system overall.) But the heat is REDISTRIBUTED from the internal space to the external space, thus adding heat energy to the very outdoor atmospheric volume being measured by thermometers in met stations. The indoor-outdoor balance is disturbed, even if the overall is not. But the excess heat is now outdoors.  What used to be is now outside.  The net gain/loss is pretty much zero.  But nobody puts met stations INDOORS.  They are all outdoors.  Where the heat has been moved to.  One air conditioner at a time.  Now add them all up, and what do we have?

The more buildings (and floors), the more air conditioning. The more air conditioning, the more watts added to the external atmosphere. In other words, the bigger the city, the more heat energy pushed out where the met stations can measure it all.  Is it any surprise that bigger cities have on balance higher UHI values?

Notice though, that the displacing of this heat energy has nothing directly to do with CO2 levels in the atmosphere. It is an element of “Land Use”, not CO2.

No, this is not the only “skewing” effect of land use, and it is not the major portion of all land use effects, but it does exist, and it is one that gets ignored. <b>And it is a FAR higher percentage of land use UHI than CO2 is as a percentage of the atmosphere. If CO2 at 0.000400 of the atmosphere is important, then air conditioners are important, too.</b>

I worked in offices in two sweltering cities that had NO air conditioning back during the 1966-1972 period, and those offices were more the norm than the exception. Try to find a non-air-conditioned office in a city now! I think it is fairly clear that the amount of air conditioning since around 1970 has multiplied by at least ten-fold, if not 100-fold. What percentage of air conditioners exist in cities? 75%? 85%? 90%? More? That, I think, parallels the perceived rise in global temps since 1970.

One other point: The warming has been attributed more to nighttime lows being not as low, rather than to daytime highs being higher. Thus, this “Air Conditioner Effect” is addressing the OTHER part of the warming. What does that mean? That cities might (according to the met stations) be cooler now than they might otherwise be? That cities overall balance might be globally cooling? With more efficient buildings in the first place, perhaps they reflect more heat energy than buildings in 1970; that is not unlikely. Without quantifying this, all I have is my engineering sense that that heat energy being measured at the met stations is not being measured and that it is an unknown. (I’ve only heard of one paper that addressed this, and I subjectively don’t agree with its conclusions. With only one paper, to me there has been no replication.)

You heard the term here first: “<b>The Air Conditioner Effect</b>”.

We are intentionally pushing heat energy out where the met stations can be affected by it. The effect out in that outer atmosphere can NOT be zero.

In 1800 there were only one or two cities in the entire world with 1,000,000 people – Beijing and London (and those just barely).  In 2005 there were 336 such cities. See here.  Some of them have urban populations of over THIRTY million.  It is also well known that over 50% of the world’s population now lives in cities.  In 1800 that percentage was well 10%, unless I am totally wrong about it.  This urbanization is essentially totally ignored by climatologists in their haste to convict CO2 of being a criminal element (actually criminal compound).  The IPCC has, I believe, assigned a 0.01°C value to the UHI – for the entire history of the global warming (since 1900).  This is a farce value.  With 243 cities with one million people, and with ALL of them having met stations (many at their airports), this is intuitively impossible to have essentially no warming.  Read the UHI article at the link above.

Yet, even with all of that population, the slope of the warming curve didn’t do much until the post-1960 period, before it began to accelerate after about 1980.  Why did the acceleration begin at the same time as the air conditioner became ubiquitous?  Why at the same time that cities in tropical areas and ones with very hot summers in mid-latitudes began growing as air conditioning made living in those places more palatable?  With 1970 being before air conditioners were everywhere, and with 1990 being after the advent of them, I think it is a fair argument to consider that it is air conditioners PLUS increased urbanization rather than CO2 that is behind the apparent warming. Was urbanization – land use – expansion of cities greater after 1970?  Was land use alone the cause of the warming?

The Great Dying of the Thermometers

The real task should be to ask, “What is the warming in all of the RURAL areas?” After all, if global warming is truly global, then the warmth should have spread everywhere and thus should be also be reflected in the rural met stations’s readings, far away from the effects of UHI, right?

Well, just like with the “post-dated check” issue of post-adjusting, there is a problem with the rural stations.  What would that be?  Well, you see, in 1989 they essentially erased most of them them from the record. In climate skeptic circles that is known as “The Great Dying of the Thermometers”.

With the scandal breaking about the ever-adjusted values, let’s make it even WORSE and rub it in about The Great Dying of the Thermometers, too.

We all are familiar with the advent of the PC and the vast, VAST increases in their memory capacities, right?  Do we remember that in about 1990 the hard drive capacities went up about double every 12 or 18 months, right?  That the processor speeds and the RAM chips were capable of more and more every year, yes?  And more and more data could be processed faster and faster?

Well, can you guess what happened in climatology?  Right when they became MORE capable of dealing with MORE data, in 1989, they decided to reduce the number of met stations reporting to the GHCN network (the one which is the foundation of all the other climate databases).  And they didn’t reduce them by a little bit.  Here is a map of <b>1987</b>. Notice the density of stations in China, Australia, Canada, Turkey (!), and Europe: Map 198703 GSCN Thermometer Map

You have to admit it: That is a lot of met stations in the world.  (It was also the real take-off pint for computers.  That latter is a hint of things discussed below…)

<b>Here is AUG 1990.</b> Notice the BIG dropoff in Asia and Canada, as well as Africa and S America.Map 199008 GSCN Thermometer Map

<b> Now look at 2001.</b>   Australian thermometers, too, has been decimated, as well as Europe.  And Asia and Africa, in particular, are nearly devoid of stations, relative to the mid-1980s.  And stations – especially CHINA.  Compare this to 1987.  ODDLY, THEY SEEM TO KEEP ALL OF THE ONES IN THE USA.  (<i>Does anyone else wonder why that sounds weird?</i>)Map 200103 GSCN Thermometer Map

<b>And here is late 2014.</b> Except for the USA, the world’s met stations have been decimated. Canada has almost none.  China is down by 90%, as is Australia.Map 201410 GSCN Thermometer Map

Now, one question that arises is, “WHY, WHY, WHY?”  Why – at the very moment when computers began to be able to handle VASTLY MORE data points – did the GHCN network people decide on having LESS data points?

You  the reader may think that I am asking this hypothetically, but no.  Heck, I didn’t even think of this myself.   In 1997-2001 I worked with a (now deceased) licensed meteorologist, who had done many climate models of the atmosphere himself, who was an early and eager champion of computers in climate studies, and HE asked ME that question.  I had no answer for him, back in 1998.  I still don’t.

How Do They Do It With Fewer Data Locations?

Since we don’t know the answer to that question, let’s discuss what HAS happened instead of having more data points…

Le’ts look at the arctic areas, in Canada and in Russia.  Look at 1987 above and look at 2014.  In 1987 except for perhaps Greenland there were met stations every 200 miles or so.  See how they were spread out along the north shore of Canada, with at least 8 stations on Baffin Island alone.  And in Siberia there are any number spread out near or on the north shore in 1987, and though not so many less in 2014, there are less.  This arctic area since about 1990 has been specifically the area that has been treated – vocally and loudly and intentionally – as the “canary in the mine shaft”.

Skeptics like Anthony Watts have pointed out time and again how the very few met stations left in the Arctic are used to represent the entire arctic area.  How do they do this?  By their adjustments and homogenization techniques, which peek at the numbers of “nearby” met stations in order to interpolate what the thermometers are doing in un-covered areas.

This they do INSTEAD OF USING THE OTHER ACTUAL DATA FROM THE MET STATIONS THEY DO NOT USE NOW.  Yes, most of those not included are STILL existing and taking readings – except GHCN ignores them.

<b>Instead of USING the actual data, GHCN estimates them!</b>

Think about that for a moment!  GHCN would rather use interpolated data – calculated data – than real measurements.  The real question is WHY.  It isn’t necessary.

  1. Those stations still exist.
  2. The computers since the FIRST Great Dying of the Thermometers have increased their storage and computing capacities.
  3. The amount of WORK necessary by climatologists is no more with all of the tens of thousands than with the current number.  The computers do all the crunch work.
  4. Had they chosen to, the climatologists could have RE-INCLUDED many or all of those now excluded stations, in recognizing that the computers could handle the load.
  5. Others have looked at the data from the now excluded stations and had no problem dealing with the volume of data; one must conclude that such inclusion would, then, be just as easy for the climatologists.

Notice that I haven’t even gone into the rural-vs-urban met station thing.  Anthony Watts did a great job with that and the work doesn’t have to be repeated.  But I WILL point out that the great majority of the met stations that “died” – were removed from the GHCN database – were rural, and rural stations have no UHI.

Overall, then, the foundational point in my own thinking is that we should not be using met stations that might have a UHI that needs to be adjusted for.  Rural stations SHOULD be able to tell us if GLOBAL warming has happened.  If UHI exists AND if global warming exists, then the UHI heat should be well represented, even out in the boonies.

And if they do NOT show it, then what?

Let’s not be measuring temps near relatively small heat source regions and arguing that that is global.  Let’s SEE if it is global.  If that heat spreads out and affects the entire globe, then the rural stations should be showing it.  Global is global, after all.  So why did they keep all the urban met stations and dump the more rural ones?  No reputable and empirical scientist would look at the places of extremes and claim that those represented the average.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s