As Record-breaking winter weather grips North-Eastern USA (and much of Canada as well) President Donald Trump has caused quite a stir with his latest Tweet.

There is nothing new in the current President’s tweets causing controversy. This is a hard-hitting one has highlights a point of real significance for AGW theory. After decades of human-caused global warming, record cold temperatures are more significant than record warm temperatures. Record cold can be accommodated within the AGW paradigm by claiming greater variability in climate resultant on the warming. This would be a portent of the whole climate system being thrown into chaos once some tipping point had been breached. But that would also require that warm records are
(a) far more numerous than cold records and
(b) Many new warm records outstrip the old records of a few decades ago by a greater amount than the rise in average temperatures in that area.
I will illustrate with three temperature data sets I looked at a couple of years ago – Reykjavík, Iceland and Isfjord Radio and Svalbard Airport on Svalbard.

Suppose there had been an extremely high and an extremely low temperature in 2009 in Reykjavík. For the extreme high temperature to be a record it would only have to be nominally higher than a record set in 1940 to be a new record. The unadjusted average anomaly data is the same. If the previous record had been set in say 1990, a new high record would only be confirmation of more extreme climate if it was at least 1C higher than the previous record. But a new cold record in 2009 could be up to 1C higher than a 1990 low record to count as greater climate extremes. Similarly in the case of Svalbard Airport, new warm records in 2008 or 2009 would need to be over 4C higher than records set around 1980, and new cold records would need to be up to 4C higher than records set around 1980 to count as effective new warm and cold records.
By rebasing in terms of unadjusted anomaly data (and looking at monthly data) a very large number of possible records could be generated from one temperature station. With thousands of temperature stations with long records, it is possible to generate a huge number of “records” to analyze if the temperatures are becoming more extreme. But absolute record cold records should be few and far between. However, if relative cold records outstrip relative warm records, then there are questions to be asked of the average data. Similarly, if there were a lack of absolute records or a decreasing frequency of relative records, then the beliefs in impending climate chaos would be undermined.

I would not want to jump ahead with the conclusions. The most important element is to mine the temperature data and then analyze the results in multiple ways. There are likely to be surprises that could enhance understanding of climate in quite novel ways.

Kevin Marshall