Sunday, 1 June 2014

A month of three halves: May weather report

I hope the title isn't too upsetting to any OCD enhanced linguists (or mathematicians).  It refers to the way May started off rather cool then got unseasonably warm for a couple of weeks before the cool was reborn.


The point in the preamble is shown by the following chart.
Up until the 9th maxima (top of the green bars) were below 15oC and the minima (bottom of bars) were below, or close to 0oC.   For the following 5 days the maxima rose a little while the minima stayed down.  For the period 15 May to 27 May the maxima stayed higher while the minima also rose.  28th to 31st reverted to the pattern earlier in the month.

This shows an interesting imbalance between the proportion of days in which the temperature at 23:30 hours was higher than the temperature at 0:30 (yellow boxes) and the converse situation (blue boxes).  That seems to show the effect of the sun in warming things up and then slowly cooling after sunset.

This led me to look at the timing of minimum temperatures.  (It initially proved a tad difficult to get a sensible vertical axis for the following graph until I realised the "counting unit" was proportion of the day,rather than hours.)  I think it shows an interesting pattern, with most minima occurring in a small time band around dawn..
I am having trouble working out some words to explain in a simple fashion what is going on when the minimum occurs close to midnight!  It is something to do with a cool change coming through during the day!


This section should more properly be called precipitation as there were sufficiently heavy fogs on 9 days to give a recording of 0.2mm.  We only recorded >5mm on two days and thus had to do a little watering.  Overall, we got more precipitation (21mm) than last May (12mm) so not great but OK.
Taking a long term view - always a good idea with weather  - here is a graph of  my series of 12 month moving average rainfalls.
The trend line is beginning to stop going down.  That is probably an obscure way of saying we aren't out of a dry patch yet.  But we're certainly better placed than we were in December last year.


The first part is to look at the maximum gust per day.
That is pretty much all over the place (I won't drop into idiom about challenged person's faecal distribution) so lets look at the daily average of gusts per 30 minutes.
At a glance it seems that May had its moments but was a bit less windy than other months.  We have numbers here to help with that.

Month ave Std Dev
March 5.67 2.94
April 4.99 2.24
May 5.07 2.93

This shows that it wasn't as strong a wind as March (quite close to April on average) but much more variable than April as shown by the Std Dev.


The chart which follows shows the 5pm relative humidity for April and May.
In view of the  absence of rainfall it was a little surprising to find that May was basically relatively moist.

To take a bit of a deeper look at this I created a chart showing the average rH by hour, together with the confidence intervals (at 1.96 Standard Deviations) of daily observations.
As expected the values drop through the day.  It is however surprising (but quite sensible) how stable the night-time readings are.

No comments: