Saturday, 4 October 2014

September 2014 weather report (and seasonality)


What is this thing called rain?  We had effectively one day of rain in the month, as shown by this chart.
Overall we totaled 38mm.   This is 66mm less than 2013 and 17mm below the average for the previous 7 Septembers.
Despite this aridity Whiskers Creek has run throughout the month, although down to a trickle by the date of writing this. 


A very pleasant month, being a bit like Goldilocks porridge - not too hot and not too cold.
The green lines join the maximum and minimum temperatures for the day and the boxes show the difference between the temperature at 00:30 and 23:30 (yellow has 23:30 higher and blue 23:30 is lower).  I noted that we had maxima above 20 degrees for the first time in a while.  This inspired me to look at temperature data for the year as a whole.

This is a first look at this and I don't have even have one full year of data yet, but it does seem to suggest that the definition of seasons on the Anglocentric basis of three months each could use a bit of work.  I shall make it so, in the near future.


I set up four classes of maximum temperature, using 10 degrees as the class interval because it seemed simple!  I then counted the number of days in a month in each class.
While it might have been more precise to calculate the percentage of days in each class (to allow for days per month varying a bit) the pattern is clear enough:
  • all maxima >30 in January or February ; and
  • all maxima less than 10 occurred May to August


The same approach was applied to minima, although I decided to reflect the definition of frost in the class intervals.  This has a frost occurring when the temperature in a standard screen falls below 2.2 degrees (I rounded this to 2 degrees as my screen is not  quite 1.2m above the ground). When I was growing up my Dad differentiated between: 
  • a ground frost (temperature probably 0o at ground level   - or ~2o at screen height); and 
  • an air frost (temperature 0o or lower at screen height).
and I have used those values in the two lower class intervals.
All minima after April are below 10o and all air frosts occur in May or later.  July and August are the really cold months with >60% of minima being air frosts.

Interim Summary of seasonality

After I first published this I realised that I didn't come to any real conclusions.  Having chatted about it with Frances on our dog walk I will now offer the following:
  • Summer is defined as the months with maximum temperatures over 30o.  Looking at some other data I have, this would include November and December as well as January and February.
  • Winter is defined as the period when >50% of days have an air frost.  Thus July and August.
  • Spring is between Winter and Summer (thus September and October). 
  • Autumn is between Summer and Winter.  This is the remainder: March, April, May and June.
Possibly looking at weeks would give some fine tuning to this but at present that would be stretching the capabilities of:
  1. the dataset available to me; and
  2. my skills with ACCESS.


Eyeballing this chart suggests that the humidity reduces through the month.  This isn't reflected in a significant trend (too much variation between days) but does suggest a look at the nine months might be useful.  OK: interesting rather than useful.
This does show that average 17:00 humidity is dropping compared to the peak reached in June.  There is probably a world's more exploration on this topic but the post is getting too long: perhaps later, 


There were a few windy days in September as (dry) fronts moved through.
I have tried to make some sense out of differing monthly wind records but has proved a bit tricky, and will be passed over until mor edata becomes available.  About all I conclude is that, for 2014 at least, adages about the equinoxes being the windiest periods are not suported by evidence!

No comments: