Hi

A bit late on this one today, I didn't realise it was St Swithin (or St Swithun’s) day till late this afternoon. I knocked up some codes to analyse the daily UK precipitation data set made available by the Met Office on their HADOBS site. The series only extends back to 1931 but I didn't think that I was going to find a past event in that time but how wrong I was!

It turns out that if you choose the England and Wales series there have been three St Swithin events in the last four years – 2009, 2010 & 2011, I know that sounds incredible but its true. The fact is it all depends on how you classify a wet day. Because the daily UK Precipitation for each of the 11 regions are a composite, dry days are not as common as you would think, and there are many days with less than 0.05 mm of rain. But if you’re messing with St Swithin you have to play it by his rules, and as far as I see it a day with even 0.01 mm of rain should be classified as a wet day compared with a day with 0.0 mm of precipitation – a completely dry day.

I know I’m stretching a point, but the fact is using Met Office data for three out of the last four years when it rained on the 14th of July there was then measurable rainfall for the next forty days - proving St Swithin correct!

I do have some radio button tucked away in a pop-up menu so that I can select a rainy day threshold of 0.0, 0.2 or 1.0 mm and Obviously the number of St Swithin events then drops to zero if I use anything other than 0.0 for wet, but interestingly if you swap to the wettest region – Northern Scotland – St Swithin events have still happen twice in the last 80 years or so for days 0.2 mm.

Bruce.

PS I think we can safely say that 2013 will have 0.0 mm of precipitation for St Swithin’s day at least for England and Wales, and that we can look forward to a few more of them in the next forty days.

PPS
Graphics to back up the story at the usual place:
http://xmetman.wordpress.com/

On Monday, 15 July 2013 20:35:44 UTC+1, exmetman wrote:

I know I’m stretching a point, but the fact is using Met Office data for three out of the last four years when it rained on the 14th of July there was then measurable rainfall for the next forty days - proving St Swithin correct!


Interesting. What about St Swithun's day itself, which is the 15th? There must have been at least 39.

Tudor Hughes, Warlingham, Surrey.

Ooops I thought it was the 14th...
I'll change the dates and rerun it tomorrow!
Thanks for pointing it out
That'll learn me.

On Monday, 15 July 2013 22:02:12 UTC+1, exmetman wrote:
Ooops I thought it was the 14th... I'll change the dates and rerun it tomorrow! Thanks for pointing it out That'll learn me.

What you really should do is run it for every day in the summer (Jun/Jul/Aug) for a number of years and see if St Swithin's day stands out. My guess is it won't. :-)
There may be a different result if it were run for one particular place.. There should be a slight effect because if it rains on St Swithin's Day it is more likely to rain the following day, and so forth, due simply to persistence. But I can't see why 15/7 should be any diffrent from any other day. I might do this myself with my 30 years of daily rainfall data for this place stored on my Acorn A7000, a steam age device but so easy to program, and quite adequate.

Tudor Hughes, Warlingham, Surrey.

On 16/07/2013 02:20, Tudor Hughes wrote:
On Monday, 15 July 2013 22:02:12 UTC+1, exmetman wrote:
Ooops I thought it was the 14th... I'll change the dates and rerun it tomorrow! Thanks for pointing it out That'll learn me.

What you really should do is run it for every day in the summer (Jun/Jul/Aug) for a number of years and see if St Swithin's day stands out. My guess is it won't. :-)
There may be a different result if it were run for one particular place. There should be a slight effect because if it rains on St Swithin's Day it is more likely to rain the following day, and so forth, due simply to persistence. But I can't see why 15/7 should be any diffrent from any other day. I might do this myself with my 30 years of daily rainfall data for this place stored on my Acorn A7000, a steam age device but so easy to program, and quite adequate.

Tudor Hughes, Warlingham, Surrey.


I've still got my Sinclair Spectrum 48K in the loft. I was hoping
that'll be valuable in my dotage.

On Monday, July 15, 2013 8:35:44 PM UTC+1, exmetman wrote:
Hi



A bit late on this one today, I didn't realise it was St Swithin (or St Swithun’s) day till late this afternoon. I knocked up some codes to analyse the daily UK precipitation data set made available by the Met Office on their HADOBS site. The series only extends back to 1931 but I didn't think that I was going to find a past event in that time but how wrong I was!



It turns out that if you choose the England and Wales series there have been three St Swithin events in the last four years – 2009, 2010 & 2011, I know that sounds incredible but its true. The fact is it all depends on how you classify a wet day. Because the daily UK Precipitation for each of the 11 regions are a composite, dry days are not as common as you would think, and there are many days with less than 0.05 mm of rain. But if you’re messing with St Swithin you have to play it by his rules, and as far as I see it a day with even 0.01 mm of rain should be classified as a wet day compared with a day with 0.0 mm of precipitation – a completely dry day.



I know I’m stretching a point, but the fact is using Met Office data for three out of the last four years when it rained on the 14th of July there was then measurable rainfall for the next forty days - proving St Swithin correct!



I do have some radio button tucked away in a pop-up menu so that I can select a rainy day threshold of 0.0, 0.2 or 1.0 mm and Obviously the number of St Swithin events then drops to zero if I use anything other than 0.0 for wet, but interestingly if you swap to the wettest region – Northern Scotland – St Swithin events have still happen twice in the last 80 years or so for days 0.2 mm.



Bruce.



PS I think we can safely say that 2013 will have 0.0 mm of precipitation for St Swithin’s day at least for England and Wales, and that we can look forward to a few more of them in the next forty days.



PPS

Graphics to back up the story at the usual place:

http://xmetman.wordpress.com/

Hi Bruce, I found some interesting facts - just for my particular area in Wanstead.

Since 1961 Wanstead has enjoyed 29 dry St Swithun’s Days (56%). From these the longest dry spell that followed was just 18 days, in 2000.

On average, if St Swithun’s is dry, the weather stays fair for five days, with rain arriving on the 20th.
Another interesting observation is that the 40-day period that followed a dry St Swithun’s is on average 13% WETTER than if it rained on July 15th.

Putting superstition and singularities aside current weather models back this up, also suggesting an extended period of dry weather lasting well into July. However, the latest jet stream forecast suggests a period of unsettled weather, ie rain, on… 27th. We shall see.