Roy Spencer starts his chapter in Climate Change 2017 with the comment that our met stations started out to record large daily changes that people actually noticed and now temperature tracking is focussed on changes of fractions of a degree over periods of years. Most of the data exaggerate any warming that has occurred because it is subject to the urban heat island effect, an effect which is not confined to urban areas because many country stations are poorly sited. See Don’t Sell Your Coat by Harold Ambler and some of the other chapters in this book.
Spencer is probably best known for his chart that shows how almost all of the 92 complex climate models in use to “project” the future have failed. This chart should be on every school noticeboard in the nation beside a chart that explains how CO2 fertilisation is greening the planet. He is a Principal Research Scientist at the University of Alabama and he previously worked with John Christy at NASA developing satellite-based recording. Christy has done a great Youtube video on the science, politics and morality of the climate debate. Just try not to be distracted by his uncanny resemblance to Ned Flanders in The Simpsons.
With the phasing out of old-fashioned mercury thermometers the three basic measuring systems are now: First, surface-based thermistors giving continuous readings with (apparently) great precision. Second since the 1950s there are radiosondes carried in weather balloons. Their coverage is sparse but they have the benefit of giving readings at different altitudes. Third, satellites systems including two independent groups that are routinely cited in warming studies. One is the unit at the University of Alabama and the other is a private research firm in Santa Rosa, California.
Global-average warming trends
The bottom line of Spencer’s analysis is a weak warming trend that is nothing to worry about.
Despite the recent pause of hiatus in warming it can be seen that there is a weak warming trend since the 1979 of about 0.14C per decade…If it continued it would be about 1.4C in 100 years which would probably be easily adapted to by humans and the natural world.
Of course a warming trend is just what you get when you are emerging from a mini-glacial and it takes a lot of ingenuity to relate that to CO2 and justify the catastrophic and divisive war on coal and other hydrocarbon fuels.
The urban heat island effect quantified
In a section on the comparative accuracy of satellite and ground-based readings he reported on an worldwide exercise to compare nearby pairs of thermometers where the population density differs. The population density stands as a proxy for the heat island effect and eliminates the need to classify locations into urban and non-urban. The results in Figure 12.4 show a rapid rise at the bottom of the population density range followed by a steady upward trend to reach 1.5C at the top of the range.
UPDATE. Helpful supporting information from Bruce of Newcastle for people who are too busy to read comments.
The version of the graph that John Christy used in his testimony to Congress is the clearest and simplest. Worthwhile bookmarking.
While I’m at it here is Roy Spencer’s blog post on his UHIE analysis. Also worth bookmarking. The final graph is the best one: it shows that a population density of only 20 people per sq km is enough to equal the amount of global warming last century (~0.75 C). Calling it UHIE is a bit of a misnomer, it is due to land use changes like agriculture, rural roads and etc.