This is in a way complementary to a post by Paul Homewood
Figure 1. UK Sunshine hours according to Met Office areal data.
As I have been doing recently for rainfall, this is deannualised and normalised.. Filter is end corrected.
The result I get is slightly different because of the compensation for time of the year. December and January are dim months, sunniest is May. Using the simple meteorological mathematics and using the hard edged meteorological period of winter they found an extreme, sunniest winter in the data. See if that looks right given Figure 1. The Met Office also write “March has continued the trend”, Trend? That word is overexercised, they mean a run of four months, not unusual.
If I use the meteorological average of three months to make a winter figure I make it 5th sunniest since the 1929 start. Given the site and equipment changes over the whole UK, take it roughly.
I’ve included a compromise filter for Figure 1 so that the fast noise is suppressed reveals sunshine follows roughly the same path as temperature.
Looking at rainfall and sunshine combination files where I show the top 20 of wet/dry, sun/gloom, 2015 does not appear at all. Nothing to see here.
This is mentioned as particularly sunny, not in top 20 so where is it?
73rd but I have taken into account annual variation, so yes sunny, notable, no.
If time was handled properly the results would be somewhat different. I cannot do this unless the Met Office open up and supply much better data.
Figure 1 was produced as an experiment using different plot software, a ticklish problem since all methods have pros and cons. Any comment?
Recomputed from scratch as usual, no difference from before. Here they are anyway.
[update] I’d forgotten I had done Tmin, Tmean and Tmax too, added.
These are more interesting because they show more of a popcorn noise, episodes of hot or cold, implies more like 1/f noise[/update]
Post by Tim