How I spent my Covid 19 Social Distancing Time

They claim Sir Isaac Newton invented Calculus during a plague. I may have been that bored but certainly not that smart. I did spend my time playing games with long term temperature data from six weather areas in the Great Plains. This is Les Henry’s fault for the article in Grainews in January and again in March where he first looked at 100 years of data for the Swift Current Research Centre and then included 100 years of data for Fargo and Bismarck. This post is kind of wonkish* so if you are not into weather or numbers feel free to skip it.

Finding long term weather data for Canada is not easy, especially for Western Canada (SRC in Saskatoon goes back to 1966). But once I learned how to use the NOAA website (https://w2.weather.gov/climate ) I could find any number of locations with data going back over 100 years. Except South Dakota, too much missing data, even for a slap happy person like myself, so I skipped it.

I learned several things. First, I have a huge amount of respect for Climatologists. Never mind predicting the future, just making sense out of today’s weather in the greater scheme of things is not easy. Second, rolling averages should be limited to five years, max eleven years, with the average temperature for each period assigned to the centre year: e.g. 1882 is the average of 1880 to 1884, 2017 is the average of 2015 to 2019. Third, to compare temperature change among geographically distant points, one needs to take the difference between the actual measurement and a baseline, usually 30 years. I chose 1951-1980 for no good reason and could have chosen 1961-1990 or even 1981-2010. Next year we will have 1991-2020 to work with. 

I need to thank Raven Onthill and Elaine Wheaton for their assistance. Raven is a fellow blogger with experience in using statistical analysis on meteorological data. Elaine is a retired Climatologist from Saskatchewan Research Council who has contributed to the IPCC Assessment Reports.

In Figure 1 below, 5 year rolling averages were calculated for annual mean temperatures (all in degrees F, sorry) for six weather areas from North to South. NOAA weather areas include several weather stations not just one. This provides a more accurate temperature than just relying on one station, as every station likely has its own little micro-climate. Trend lines were calculated using 4^th order polynomials for no good reason except it made nice curved lines and the largest R². The larger the R², the more variability is explained by the trend line. The equations and R² from left to right match the trend lines for the weather areas from left to right. The upper left is for Fargo etc.

Figure 1, Five Year Rolling Average of Annual Mean Temperatures

That the annual mean temperature increased from North to South was hardly unexpected, but it also increased from west to east for Cheyenne, North Platte (Ogallala) and Omaha weather areas. Since I know nothing about these areas that was unexpected. However, the baseline temperatures from 1981-2010 (Figure 2) illustrated that nicely. The three weather areas may be side by each (Figure 3) but have quite different climates.

Figure 2, Annual Average Mean Temperature 1981-2010

Figure 3, Three weather areas showing number of weather stations

Plotting the annual mean temperature difference from a 1951-1980 baseline and calculating a 5 year rolling average gave me a chart like this (Figure 4), with R² ranging from 0.77 to 0.40 with four between 0.5 and 0.6 and a marked dip between 1960 and 1990. I sent this to Raven Onthill for comments and he sent me the link to this chart (Figure 5) from page 187 in IPCC AR5 Chapter 2 (reference below) noting that both charts showed a similar dip in temperatures between about 1950 and 1990. Figure 5 is based on world temperature differences from 1961-1990 from four different data sets. I thought “Holy cow, look what I did”.

Figure 4, Five year rolling average of annual mean difference from 51-80 baseline

Figure 5, see reference below

Although I had the data by month, I calculated by quarter as I was lazy and not sure if I would learn much for the work. I may go back and do it anyhow. The following charts are for Q1-Q4. There are certainly seasonal differences in the rolling average trends and R². Most notable in the 1^st 3rd and 4^th quarters.

 

Figure 6, Q1

Figure 7, Q4

Figure 8, Q3

Figure 9, Q2

Elaine Wheaton asked what question(s) I was trying to answer. That was a tough one as I mostly use the Thomas Edison approach which is try something and see what happens. But the question really was whether, if the global climate is changing, can we measure it at our own backdoor. I concluded that it is not that easy, and more than temperature must be considered – precipitation and timing, wind and weather events also come into play. What is happening in one part of the world is not necessarily happening in other parts.

Hartmann, D.L., A.M.G. Klein Tank, M. Rusticucci, L.V. Alexander, S. Brönnimann, Y. Charabi, F.J. Dentener, E.J. Dlugokencky, D.R. Easterling, A. Kaplan, B.J. Soden, P.W. Thorne, M. Wild and P.M. Zhai, 2013: Observations: Atmosphere and Surface. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

* not wankish, you Brits.