Freezy hot climate change

Tamino who runs the blog Open Mind which is mostly about climate change, knows far more about statistics than I do.  One of my majors at uni was Mathematical Statistics, but that was a long--a very long--time ago.  In a recent post he discusses the whole issue of data analysis, and since he has simplified it down for idiots like me, even I understood it.

First he shows a chart of the daily minima and maxima for a place called Kremsmünster in Austria (You add the 'e' after the 'u' to show the umlaut when it doesn't exist in your language's alphabet):

What does that chart tell us?  Well, first, there is a clear seasonal cycle. (Duh)  Second, the gap between minima and maxima in summer is great than in winter.   Third, there are "spikes" when a single day is hotter or colder than surrounding days (that's called random fluctuation, and in economic analysis you reduce their influence by using extreme adjustment and/or moving averages).  Fourth, you might be able to discern a modest uptrend.  Maybe.

Now, to determine the trend in temperatures, what you have to do is to take the temperature for each day and compare it to the average for 30, 50 or 100 years.  So if it's always hot on 4th July, what matters is whether it's getting hotter over time.  This is called the temperature anomaly and also allows you to compare temperature trends from places which have different average temperatures, for example, the Arctic and Miami.  The trend in the temperature anomalies of both places can be up even if their average temperatures are very different.

Tamino did this and then averaged these anomalies for each year.  He then fitted a Lowess smoothing to these averages.  This is what the average annual anomaly looks like:

So what can you see here?  Well, in Kremsmünster, anyway, the trend of the temperature anomaly rose (according to the records) from 1880 to 1900, then was basically flat apart from small fluctuations until about 1980 and then started rising inexorably since then.  In fact, it looks as if temperatures have risen on average by 2 degrees C since 1980.  Note that  even if the average was flattish between 1900 and 1980, there were big fluctuations from year to year.  And some years the average anomaly was right outside the 95% probability bands represented by dashed lines in the chart.  In my economic analysis I tend to remove those extreme observations using an algorithm which for short I call "extreme adjustment".   I've written a handy little program which does that.

You might say well, big deal, what's 2 degrees?  After all, as you can see from the top chart, the temperature fluctuates from -10 C in mid-winter to plus 30 C in mid-summer.  2 degrees? Pfui!  But it's the extremes which move. 

Here's the number of freezing days for each year from 1880 fitted with a Lowess smooth and a "piecewise linear" trend:

Back in 1880, there were something like 80 freezing days a year.  Now that number is down to 40.

What about hot days(days when temperatures exceeded 30 degrees C)?

Well, from 1910 to 1980, hardly any.  Just a few days each year, often none.  But since 1980, a clear uptrend.  And now there are 20-30 days a year when the temperature exceeds 30 degrees.  So even though average temperatures are up by what seems to be only a little (2 degrees C) the number of hot days is up 20 fold.

[Read more here, all charts from that source too]

People who don't understand the statistics underlying our certainty that there is climate change get distracted by a particularly cold day in mid-winter.  How can there be global warming, they ask, when I am so freaking cold?  It's much easier to understand in summer, though.  Someone born in the 50s and 60 and 70s will be able to say that it's much hotter now than it was in their youth.  And they'd be right.  In fact, it was record temperatures this past northern hemisphere summer which crystallised Europe's belief that global warming wasn't just real but also a threat.

At Paris, the world committed to trying to keep the rise in temperatures to below 2 degrees C.  As you can see, in some places we are already there.  We need to do more to green our economies; we need to counter the lies spread by fossil fuel companies and billionaires and by the rabid right and Republicans.  Fortunately, the precipitous collapse in the cost of renewable energy and batteries has made our path easier and clearer.  The only thing now stopping us from a rapid and wholesale switch to renewables is politics.  And in democracies we know the solution to that, right?

[With thanks to Tamino, who is trying hard to lighten the world's ignorance about climate change, one chart at a time]