Monday, 20 July 2015

The length of the Cool

At some point recently in discussing the cold weather being experienced around Canberra, Sean Carson - the BoM go-to man for quotable quotes - explained that on some days there is more time with the temperature below zero than above.  His example was the temperature falling below 0  by 7pm and not rising above it until about 9am the following morning.

Some detail on how I have assessed this follows but the big result is that days with extended periods of air frost (ie temperature below 0oC) are most frequent in August.

This is quite a reasonable outcome since:
  1. August is preceded by more cold months than June or July:and 
  2. the angle of the sun is still low so doesn't warm the atmosphere as quickly as it does from September onwards.
This logic follows equally well for explaining why, in the Northern Hemisphere, February seems the coldest month.

Now for the detail.  Unfortunately the time-frame cited by Sean is very difficult for a bear (possibly a polar bear) of little brain, and less SQL programming knowledge, to reflect in an exploration of my weather data.  What I have used as a substitute is the number of days with more than 'x' readings below 0, classified by month. I have converted these to percentages (to adjust for having 3 years of observations for May, 2 years for June and a single month for the other cold months) giving the result in the following table:

% of days with >n hours below 0oC

>6
>9
>12
May
22.58
11.83
7.53
Jun
38.33
31.67
20.00
Jul
41.94
32.26
19.35
Aug
48.39
48.39
38.71
Sep
16.67
6.67
0.00
Oct
6.45
3.23
0.00
This is illustrated in a graph:


2 comments:

Peter Quinton said...

Makes sense to me - the drop to 0 is the most important marker for me, cooling and drying. Can I ask what period your data was over?

Flabmeister said...

Thanks for the comment Peter. The series is not long - 2013-15: but one has to start somewhere. I will hopefully remember to redo the sums in 2016!

Martin