Wednesday, 24 January 2018

Some season oriented weather material

In a weather discussion group in which I participate there is some discussion of seasonal comparisons, at present mainly on whether Summer 2017-18 is the Hottest OF ALL TIME.  The jury will have to stay out on that for a few more weeks but it has prompted me to have a look at the seasons of 2016-17 as I define them.  To remind readers of those definitions.  For the reasons explained here :

  • Summer is November (year x)  to February (year x+1);
  • Autumn is March - June;
  • Winter is July and August; and 
  • Spring is September and October.

I don't really have enough data on Humidity or Wind to say much so have restricted myself to commenting on Rainfall and Temperatures.  The period being looked here is from November 2016 to October 2017.

I have tried to make this post self-contained so I might repeat material from earlier posts.  Sorry if that is annoying but I felt it better than whacking in a lot of links.


In a word - very low.
 Every season is well below the average although none are quite the lowest recorded (and some of the minimums are really appalling).

I have shown two measures of minimum in this chart;

  • the traditional average or arithmetic mean; and
  • the median which is the value which has half the values higher and half lower.

I did this mainly to see if they told different stories.  As apparent from the chart they are very similar, so I will revert to just showing the mean.


The first measure of heat is the average maximum temperature in the seasons. The values for 2016-17 are so close to average in all seasons that the lines for average and the last year almost overlay each other.
 The other measure I use for heat is the number of heat wave days.  A Heat Wave is a period of 3 or more consecutive days with maxima above 30C and minima above 10C.  This is a bit complicated but is close to the B0M definition of a Heat Wave.  A Heat Wave Day (HWD) is a day in a Heat Wave.  

As might be expected, since I define Summer as the period with many maxima above 30C, Heat waves mainly occur in that season.  There are so few heat Waves outside that Season (and none in 2016-17) that I have decided not to show a graph in the format above but a comparison of HWDs in the years since 2010.  (I have doubts about some of the maxima in earlier years.)
Summer 2016-17 was a reasonable amount above average (which position is maintained even if the very low value for 2011-12 is removed.  It will be interesting to see how 2017-18 fits in: December was low and January high.


 The simplest measure of cold is the average minimum temperature.
All seasons in 2016-17 were above average, and except for Winter, were close to the warmest recorded.  This is not a good thing.

I use two measures of frosts.  The first is a Ground Frost (GF) where the temperature at screen height is below the equivalent of 0C at ground lever: for simplicity I take this as  +2C.  The second measure is an Air Frost (AF), a subset of Ground Frosts where the temperature at screen height is below 0C.
The average seasonal patterns are as expected.  The very interesting phenomenon is that while AF were all close to average for Autumn and Spring the number of Ground Frosts was much below average.  This suggests that we got a normal amount of very cold weather but when conditions were't suitable for an Air Frost we got warmer than expected weather.  I will not take this any further, but would welcome any comments..

No comments: