Seeing what one expects

By Kevin Kilty

This morning I awoke to a mid-April morning temperature of -11F. The 1981 to 2010 climate normals indicate our average daily minimum temperature per this date as about 23F, and the standard deviation as 8F. Thus, our morning low temperature is a 4-sigma event. Surely something to evoke comment. Yet, it did not so far as I know.

This caused me to ponder something I observed  two months ago. Ten minutes from my home, in the mountains to my east, is great nordic skiing. It was at one time home to what we call, the Norwegian olympics. There was unusually good snow this winter, and people came from near and far to enjoy it. What I heard often in conversation in the parking lot in February was that we were having an “unusually warm” winter. I thought not. I have lived in this area, off and on, for 40 years, and I thought this winter was pretty typical, even possibly slightly cool.

What the observations show

Combining data from Mesowest and from the NCEI office of NOAA, I produced Figure 1 for our inspection.

This figure is very revealing not only about what the present winter is like, but what typical winters around here are like. There is the “normal” winter, the smooth curves of weather averaged over 30 years time; and there is the current winter daily averages shown in bold orange variations.

In the current winter, which I view as typical, average temperature rises up to near the maximum daily climate normal temperature for a time. This occurs principally in the southerly flow of warm air a few days in advance of a cold front. After passage of the front there are days of cold air to contend with. Rarely is the daily average temperature at or near the climate normal. There is typical winter, and there is climate normal; rarely do they meet.[1]

This year the first two weeks of December showed days of variable temperature but the average remained between the climate normal average, and the climate normal maximum. One might call it warm. Suddenly the weather became more extreme with daily averages rising well above the maximum climate normal to well below. Out of the ordinary? No, typical; as the extremes still never exceeded one standard deviation above or below. Once averaged it seems like a couple of weeks of typical winter storminess; but possibly someone predisposed to “global warming” would have noticed and remembered the warmer days, but not so much the others.

In January the weather varied within a smaller range, but stayed largely within the band between the average climate normal and the average maximum normal. One would have been absolutely correct in stating this to have been a milder than normal January. February starts warm then becomes cold to the extent of 2-sigma below the average climate minimum, then proceeds with temperature extremes, again with the passage of successive storms, but generally seems to be a few degrees below the average climate normal. It hardly seems correct to label it a warm winter as the skiers were doing — especially while sweaty at the end of their circuit. February, from the data, looks to have been the coldest month of our winter season.

And even if viewing the graph of temperature gave them some doubt about their original thoughts, some could now argue with me that the climate normals from 1981 to 2010 are themselves warm because of climate change, thus providing hope of validating their original claim.

Data should settle arguments

Many people rightly say that data should settle arguments, but real data, filled as it is with noise, doesn’t often serve this purpose. Real data often contains enough variation that people operating with confirmation bias can find in it what they need to maintain a preconception. Unless the signal is very obvious, real data can actually fuel disputes.

Most of the skiers around here are university educated people who are more exposed to the current climate wisdom and probably more prone to a warm bias as a result. I observe this in their conversations about every wintertime observation they make — from how little snow they recall having moved from their driveway, to how often we have dropped below -20F this season, or to how warm they felt after their skiing jaunt. Thus, they can see a warmer than normal winter in those select warm days during February, but seem to neglect the numerous very cold days.

Irving Langmuir noted that there is a pathological side to science where bias can recruit even objective measurements to its side. It works especially well when the data are noisy and the signal  barely resolvable.[2] The ESP experiments done by Joseph Rhine were terribly biased, but so influential was he and his enormous collection of observations, that paranormal science became respectable during the 1950s. It is no longer. Many were once convinced about the Palmdale bulge; but it too vanished. Presently we can see the same workings of bias in our debates over climate change, imagined droughts, or COVID19.[3] Topics change, reason seems as frail as ever.

Notes: [1]-Thus, Earl Butz, one time Secretary of Agriculture, said that in 50 years of Minnesota farming, he had seen two normal winters.

[2]-Irving Langmuir and pathological science. Lectures presented at the General Electric company in 1953. There are many copies and synopses of this lecture series on the internet, check Columbia and Princeton Universities, but its ramifications are so universal that it has even invaded management. See for example, Pathological Science, Research Technology Management, Vol. 32, No. 5 (September-October 1989), pp. 11-17

[3]-Neil Ferguson, the Imperial College epidemiologist, who produced the scare projections now cited as evidence for tight, and tightening, social isolation, also made the following observation

 “Stopping mass gatherings is predicted to have relatively little impact (results not shown) because the contact-time at such events is relatively small compared to the time spent at home, in schools or workplaces and in other community locations such as bars and restaurants….”

Our officials are happy to cite selected evidence for their responses, but fail to notice advice arguing against some of them.