The temperature data is updated every month.
So that gives you twelve data points per year.
Over ten years that's 120, over twenty years it's 240 etc.
So that means that you've got thousands of comparisons
that you can make between any two months.
Why was 2 years chosen for this report?
Because it gave a significant result.
But what happens if you use 23, 25, 26, etc months?
And what if you use a different starting month?
You'll get a different result.
In effect, it's a non story because you're using a single data point
and ignoring all the others.

Hou toch eens op met HOKJES denken

Ik ben nu een tijdej