Record Breaking High Temperatures – What’s Really Behind It?

Reader Paul Homewood shows how PBS cooks the books to give the appearance that the USA is having a hot year when in fact it’s just another run of the mill year.

By Josh

Record Breaking High Temperatures – What’s Really Behind It?
By Paul Homewood

America’s Public Broadcasting Service (PBS) is keeping track each day of the number of high temperature records being set in the USA. As at 30th August they reckon the year-to-date total was up to 4779. The clear message is that “temperatures must be getting hotter”. But is this so? Are matters as simple as they appear? We need perhaps to start by looking at the methodology employed.

PBS are using data collected by the NCDC from official stations only (not volunteer stations). They go on to say: “In order to make it into our count, a recorded temperature must either tie or exceed a previously recorded temperature on that date in a previous year”. In other words they are not just recording all time highs or even highs for a particular month, just highs for that particular day. So, for instance, a heat wave that lasts 2 days could set 2 records even if the second day was not as hot.

PBS also state – “NCDC records go back for many stations as far as the mid-19th century”. To the extent that this is true, very few record low temperatures would be set given the fact that the 19th century was a particularly cold era. However, as we shall see later, this is such a misleading statement that it borders on being dishonest.

USA 2011 35th warmest – nothing unusual

Before we look more closely at PBS’s numbers we should have a quick look at the overall temperature record for the USA in 2011. From January to July, 2011 ranks only 35th warmest in the last 100 years according to NCDC data.

However this year has perhaps been slightly unusual, with the Northwest being relatively cool and the South and East being warmer than unusual, so averages do not tell the whole story.

Out of the total of 4779 record highs so far, 718 occurred in July, a figure slightly above the average monthly figure of 597. I have broken down this July number and find that 42 record highs are in Kansas. Again this seems to be a representative sample. According to the NCDC map, Kansas July temperatures were “much above normal” but not as hot as Texas and Oklahoma, see here.

Double counting in Kansas

The mean July temperature for Kansas was 84.8°F this year. This ranks 5th warmest in the last 100 years with 1934 the hottest at 86.7. 1936, 1954 and 1980 also had hotter Julys. In other words, although this July has been very hot in Kansas, it is certainly a long way from being the hottest, see here.

What do the 42 record temperatures for Kansas tell us when we look at them more closely? There are 20 official weather stations in Kansas of which 19 appear to be at airports. Record temperatures, however, have only been set at 8 of these. All of these 8 stations are airport based. Two of these are in Topeka and so are arguably double counted! These are listed below:

Location – Year first operational
Dodge City Regional AP – 1942
Emporia Municipal AP – 1950
Garden City Regional AP – 1939
Hill City Municipal AP – 1948
Olathe Johnson Co AP – 1978
Topeka Billard Municipal AP – 1948
Topeka Forbes Field AP – 1942
Wichita Mid Continent AP – 1953

Many “record setting” stations do not even go back to the 1930s

We can see immediately that the claim that “many stations go back to the 19th century” certainly does not apply to Kansas. More importantly none of these date back to the mid 1930s when July temperatures were much higher than now. (Topeka Billard is listed as starting in 1932. However the actual records start in 1948), see here.

Although none of these stations go back prior to 1939, there are earlier records available at other locations that are very close to some of these stations. These records show that in most cases higher temperatures were recorded either in 1934 or 1936. These alternate stations are typically less than 4 miles from the current station and the metadata show similar conditions such as altitude.

Out of the 23 “record highs” set at these 5 stations in 2011, 16 were lower than previous highs set before the station became operational. Only 30% were genuine highs or 1.4 per station. (It is possible of course that some of these records were also exceeded in years prior to 1934).

When we analyse the July temperature for these stations further. It is clear from this analysis, regardless of the parameters used, that July in both 1934 and 1936 were significantly hotter than July 2011.

Siting effects

There is one further matter to look at. What effect have airport siting or other UHI effects had on the long term temperature trend at these sites? Taking Topeka as an example, I have identified 4 other weather stations nearby which are not based at airports and are in smaller towns. They are all certified USHCN COOP stations and have continuous records dating back to 1936. Their metadata and geography are summarised below.

No station changes since 1932.
Station situated 410 ft from runway
Topeka has a population of 127,000 but the airport is 4 miles out of town.

The station is based at a water plant and has been in the same location since 1983.
Population of 2000.
Horton is 42 miles from Topeka.

The current station is within a mile of its site in 1936 with no significant change in altitude.
Population of 1000.
77 miles from Topeka. (Pawnee City is actually in Nebraska, but is closer to Topeka than most other comparable Kansas sites).

The current station is 2 miles from the 1936 site. Both sites are within the town.
It is also now at an altitude 145ft higher than in 1936 which will have a small effect on temperature differences.
Population is 66,000.
Lawrence is 40 miles from Topeka.

The current station is within 1 mile of the 1936 site and both were situated in the centre of town. As with Lawrence there has been a slight altitude increase of 80 ft.
Population is 38,000.
41 miles from Topeka.

Temperature comparisons between 1936 and 2011 show:


In other words, Topeka has become significantly warmer relative to the “rural” sites at Horton and Pawnee City. Both maximum and mean temperatures show a divergence of between 3°F and 4°F. Interestingly the medium sized towns appear to be consistent with Topeka in terms of maximum temperatures (particularly when the slightly higher altitudes are factored in), which suggests that the UHI effect in these towns is similar to the airport effect at Topeka. (As the airport is 4 miles outside town the UHI effect of Topeka’s larger size is probably not significant). There does not appear to be any logical reason, though why mean temperatures at Leavenworth and Lawrenceare relatively cooler.

So what can we conclude?

1) Over the US as a whole, 2011 has so far been barely warmer than average. In no way can it be remotely described as “hot” as implied by PBS.

2) As forKansas, July, although much warmer than average, is only the 5th warmest in recent years and 1.9F cooler than 1934.

3) The only stations in Kansas where records were set in July are all airport based and date back no earlier than 1942. Alternative stations confirm most of these “records” are lower than previously set in 1934 and 1936.

4) Where records are available back to 1934, they show that temperatures in both 1934 and 1936 were considerably higher for the month as a whole than 2011. This is true for average maximum, average mean and high maximum.

5) Comparison with rural stations suggest that airport siting and possibly UHI have had an upward effect on temperatures since 1936 of as much as 4F.

Paul Homewood

Note and References

1) The following alternate stations were used to establish temperatures in 1934 :-
a) Emporia1S – 4 miles N of airport.
b) HillCity1NE – 2 miles SW of airport.
c) Olathe3E – 4 miles NE of airport.
d) Topeka2 – 4 miles W of airport.

2) All individual station data has been obtained from the Individual Station Original Forms available from NCDC.

3) The comparison between rural and airport stations is unaffected by any TOBS considerations.  Temperatures at all stations used for this exercise were logged around sunset in the 1930’s and all 2011 ones at 7.00am. Therefore any TOBS adjustment will be essentially the same at rural and airport.

4) Station metadata has been checked for each station at NOAA to ensure compatibility.


5 responses to “Record Breaking High Temperatures – What’s Really Behind It?”

  1. Paddy

    This post mentions the significance of weather station location. Locations on airports leads to a distinct warm bias. Two posts By Dr Cliff Mass, professor of meteorology at the U of Washington, are instructive:

    It is well known that the NCDC official sites have been switched to airports of the past decade or so. Rural sites and those with colder norms have been dropped.

    Anthony Watts’ well know assessment of improperly located sites is well know. Frankly, the NCDC land temperature data is compromised as a result of theses systemic flaws.

    1. Paul Homewood

      I have been told by the Hadley Centre that an allowance of 0.2C is made to adjust for UHI as far as CET is concerned. This barely addresses the problem.

  2. R. de Haan

    Same subject From Lubos Motl:

    Austin Texas, all daily trends since 1938

    His conclusion: COOLING

  3. GregO

    All these “records” do really need to be kept in perspective – I can recall reading papers all the way back to the ’70s and there is always the breathless report of all time highs in summer and all time lows in winter. Only now days, I suppose to sell more papers, it’s all caused by Man-Made Global Warming from Man-Made CO2.

    This last August here in Phoenix was suppose to be the hottest ever. Hardly. This last summer and particularly the spring was mild by Sonoran desert standards. It was just beautiful here. I hacked around the internet just to see what was up, and not a single local piece I came upon mentioned global warming; a couple mentioned La Nina and one casually mentioned that the high was based on a single temperature station – the one at the airport – and that the city was usually up to 5 degrees F cooler than the airport (!).

    I’m not including the links because I did the search a week or so ago, you’ll have to take my word for it – mind you I wasn’t looking for confirmation one way or the other I just wanted to see what the local news guys made of hottest August ever. Ends up, not much. You know why? It wasn’t exceptionally hot here last August. Nobody really noticed. It was just a normal hot August.

    September has definitely cooled off. Here in Phoenix, the heat seems to either break hard in early September, or stay warm until late October. This year it looks on track to cool off early. It clouded up this afternoon, cooled off, and looked like rain.

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this. More information at our Data Privacy Policy