The Importance Of Proper Temperature Sampling

On University Challenge a few nights ago, the answer to one question was ‘Stevenson Screen’ [pictured].

This reminded me of the following article on climate change which I wrote nearly ten years ago. It remains relevant today.

When people challenge me for not being a scientist, I can reply that I was a miller and corn merchant, so I understand the importance of proper sampling – something with which the ‘climate community’ appears to have problems.

In these days, when heads of state and government appear to defer to the words of a mentally deranged teenager who believes she can see carbon dioxide in the air, a return to common sense is long overdue.

A thumb (or two) in the scales by Edward Spalton (2010)

In the early ’60s, I went to learn the corn trade at a firm called Lamprey & Son in Banbury. The old office and shop building still stands next to the town hall and looks much the same today although it has long been converted to other uses.

One day the boss showed me a beautifully made Victorian brass balance fitted into a polished wooden case that would slip into your pocket. On one end of the beam was a cylindrical pot about as big as a good-sized egg cup.

The other side of the beam was milled with serrations and graduated with a sliding weight which moved along it.

If you filled the pot up with a sample of grain and struck it off level, you could slide the weight along until it balanced with the contents of the pot and read off the bushel weight from the scale.

Bushel weight (these were pre-metric days; since joining the EEC, the calculations have been done in kilograms per hectolitre) is a good indicator of quality. Plump, full grains weigh heavier than thin ones.

A bushel of reasonable quality barley would weigh 4 stone (56 lb or half a hundredweight) and a bushel of good wheat 4½ stone (63 lb). The sample might represent a parcel of grain which could be anything from 5 or 6 stones up to over 100.

The boss let me try this out and, in two or three goes, I was getting a very consistent reading. He then did the same with the same sample and got a considerably heavier bushel weight.

Eventually, he showed me the trick. The strike or straight edge, which was used to level off the contents of the pot, had two sides.

One was like a ruler and the other had a piece of dowel along with it. If you used the dowel side, it pressed a few more corns into the pot than the straight edge.

With the effect of scale, this made the sample look considerably heavier and better quality. Even with a correctly drawn sample, a small change in procedure or instrumentation could significantly bias the result.

He was warning me about the sharp practice and possible carelessness in apparently objective tests.

‘That’s how they did it the old days. Buying or selling, you see, boy,’ he said with a wink. I should add that this was shown to me as an antique curiosity and was not part of the trading practices of that fine old firm!

The kit which is used to ‘sample’ the temperature of the climate is remarkably unchanged and about the same vintage as that rather splendid little balance.

It is called a Stevenson Screen and was designed by the father of Robert Louis Stevenson, the author of Treasure Island.

It is a standard-sized wooden box with louvered sides to allow free circulation of air around the instruments and keep them out of direct sunlight. Hence the expression ‘in the shade’ when referring to temperature.

The thermometer might be a traditional maximum/minimum type or more modern sensors. Stevenson Screens were traditionally painted with whitewash.

It is doubtful whether a character like Anthony Watts could exist in state-controlled Britain. He is an American meteorologist and weather forecaster for commercial TV and radio stations.

For his living, he depends upon his customers’ satisfaction with the accuracy of his forecasts. He also supplies custom-built weather stations, TV graphics systems and video equipment to broadcasters all over the world.

So he is an expert who makes his living from the weather but is neither a civil servant (who can be made to toe an official line) nor dependent on tax-funded grants (which require applicants to be politically correct).

So he has a certain independence of mind and demonstrates that rugged individualism and tenacity of purpose which used to be the stuff of all-American heroes in many films of my youth.

He noticed that the US National Oceanic and Atmospheric Administration’s National Weather Service (NOAA/NWS) – roughly equivalent to the UK Meteorological Office – had made a small change to its Stevenson Screens.

He wondered whether this change would affect the temperatures recorded. Back in 1979 the NWS had stopped using whitewash and started painting its Stevenson Screens with white, semi-gloss, latex paint.

Whitewash essentially gave a coating of calcium carbonate whilst latex paint used titanium dioxide, which has significantly different reflective properties.

In 2007, having a little time on his hands, he set up a trial to see what the differences might be. He used three Stevenson Screens, one unpainted, one painted with latex semi-gloss used by the NWS and one painted with historically correct whitewash.

He used a modern stacked plate aspirated thermometer as an additional control.

His results showed that the latex paint raised the maximum recorded temperature by 0.3 degrees Fahrenheit and the minimum recorded temperature by 0.8 degrees Fahrenheit when compared with the whitewashed Stevenson Screen.

That is an upward bias of 0.55 degrees Fahrenheit. Not very much you might think – but the whole scare about global warming is based on a claimed, observed temperature rise of only 1.2 degrees Fahrenheit in a century.

Anthony Watts then decided to have a look at the NWS’s Stevenson Screens in his locality to see if they were painted to the official specification.

What he found was disquieting. In one case, heat-generating radio equipment had been installed inside the screen near to the temperature sensors.

In other cases, the weather stations were near to the outlet vents of air conditioning systems or close to other heat sources, all of which could bias the recorded temperatures upwards.

He conducted further investigations, eventually recruiting a team of volunteers to observe and photograph as many of the 1,221 weather stations all over the United States as possible; 867 of them were visited.

Nearly nine out of ten weather stations proved to be outside the specifications laid down by the US authorities.

They were near to heat sources, on concrete or tarmac surfaces, close to buildings, in the steamy warmth of sewage farms and so on. All the observed faults would tend to raise the recorded temperatures.

It is a fascinating story of one man’s determination to get at the truth and can be read in full on www.SurfaceStations.org.

Anthony Watts also has a regularly updated blog that is one of the most widely read, independent sources of climate information. I particularly like the fact that contrary views are welcomed.

Whilst they are vigorously debated, they are treated with respect and normal courtesy, unlike some blogs pushing the official line.

To return to my analogy of the corn merchant’s balance – the few cubic feet of air inside a Stevenson Screen stand proxy for a huge amount of the earth’s atmosphere.

Weather stations are often hundreds of miles apart. So those few cubic feet are proportionately much smaller than that eggcupful of grain representing a parcel of many tons.

Any change, such as a different coat of paint, a heat-radiating transformer inside the screen or a nearby heat source can have a disproportionate effect on a tiny sample which is claimed as representative of hundreds of cubic miles of atmosphere.

Probably unintentionally, the official methods seem to have had an effect, not unlike a thumb or two being pressed on the side of the scales indicating a warming rather than a stable or cooling climate.

Yet the tax-funded ‘climate community’ was not at all grateful to Anthony Watts for looking into the basic data and methods used to measure it. For them, ‘the science is settled’ is the whole of their faith.

Forget accurate measurement. They have computer programs to adjust things in ways that only they understand. Watts is a heretic and that’s that.

How the Canadian thermometers were taken out and shot

When the US authorities began monitoring temperatures at the Earth’s surface, 1,221 US weather stations were part of a worldwide total of some 6,000.

But something strange happened in the last few years. The number of stations used to record temperature dropped dramatically. Figures were still shown for all the areas of the world but they were calculated by reference to far fewer actual observations.

They were ‘adjusted’ and ‘homogenised’. The Canadian blog Small Dead Animals reported as follows on January 16, 2010, under the heading ‘The Sound of Settled Science’:

‘In Canada, the number of stations dropped from 600 to 35 in 2009. The percentage of stations in the lower elevations (below 300 feet) tripled and those at a higher elevation above three thousand feet were reduced to half. Canada’s semi-permanent depicted warmth comes from interpolating (don’t you love the word!) from more southerly locations to fill northerly vacant grid boxes, even as a pure average of the available stations shows a COOLING.

‘JUST ONE THERMOMETER REMAINS for everything north of latitude 65N – that station is Eureka. Eureka, according to Wikipedia, has been described as “The Garden Spot of the Arctic” due to the flora and fauna around the Eureka area, more so than anywhere else in the high Arctic. Winters are frigid but summers are slightly warmer than other places in the Canadian Arctic.’

The same has happened to US thermometers south of the border. The computer expert E. Michael Smith joined forces with the Certified Consulting Meteorologist Joseph D’Aleo and appeared in a TV program on www.kusi.com on 16 January 2010.

They reported that the number of weather stations used as a starting point for measuring the world’s average temperature calculations had been reduced from about 6,000 in the ’70s to about 1,500 now.

‘That leaves much of the world unaccounted for. The greatest losses were in areas where NOAA and other data centers claim the warming was greatest like Siberia and Canada.’ D’Aleo added: ‘In these regions, NOAA “estimates” temperatures based on stations that may be 700 miles away.’

Smith noted:

‘When doing a benchmark test of the program, I found patterns in the input data … that looked like dramatic and selective deletions of thermometers from cold locations … The more I looked, the more I found patterns of deletion that could not be accidental. Thermometers moved from cold mountains to warm beaches, from the Siberian Arctic to more southerly locations, and from pristine rural locations to jet airport tarmacs…’

Even as a new trainee corn merchant, I knew better than to draw a sample from only the best part of the bulk. It would lead to the actual delivery being rejected and to great extra expense for my boss. He never liked that sort of thing!

There seems to be no similar sanction for scientists drawing bad samples of climate and temperature with equipment and methods known to be faulty. If the results are what the politicians want to hear, they are fine.

The enormous bill to supposedly remedy global warming will simply be passed to the taxpayer and consumer, even if it isn’t actually happening.

Read more at Conservative Woman