Thursday 20 February 2014

Met Office - UK's Wettest Winter since 1910

The Met Office have today announced that the winter of 2013/2014 (Dec 2013 to February 2014) is the wettest for the UK since records began (1910). Yet, it's still a week to go before the end of the month, and after a week of relatively settled weather, an unsettled week is forecast to take us into the end of the month.

The Met Office press release lists some statistics:
"With just over a week to go until the end of the season:
  • The UK has now received 486.8mm of rain, narrowly above the previous record of 485.1mm set in 1995.
  • Wales has seen 691.8mm of rain, beating the previous record of 684.1mm in 1995.
  • East Scotland has seen 514.5mm of rain, beating the previous record of 482.2mm in 1915.
  • Southwest England and south Wales has seen 632.5mm of rain beating the previous record of 610.7mm in 1990.
  • Southeast and central southern England has seen 492mm beating the previous record of 437.1mm set in 1915."
These are summed up in a map of percentage deviations from the average.

Despite the overall statistics of a wet winter, the pattern is mixed with some areas being near average, and small areas below average rainfall. In a comment to my previous post, Hautbois raises this point concerning summer rainfall, but using different data to the Met Office data. Such complexity is to be expected when considering issues such as geography, for example mountains shielding regions by forcing rainfall due to orographic uplift, and due to storm tracks.

The top ten wettest years for the UK are as follows (mm of rainfall), note that 2014 isn't even finished yet:

2014 486.8
1995 485.1
1990 470.9
1915 463.1
2007 437.9
1916 433.0
1937 423.8
1994 422.9
2000 417.7
1920 402.3

Here 6 years out of ten are in the 1990s and the 2000s. Put simply, 6/10 of the most extreme wet winters come from the last quarter of the available data.

I've not yet had the time (when I'm not tired from my day job) to get to grips with the hypergeometric formula, so I've resorted once again to the ham-fisted brute-force of a monte-carlo calculation of probability, frankly I know enough to know I don't know enough to spot when I'm screwing up with combinatorial probability. At least with monte carlo I'm less likely to make a balls up.

Anyway, my calculated probability for 6/10 of the most extreme wet winters coming from the last quarter of the available data is 0.0085. This is a low probability, of a similar order to the probability of record wet summers post 2007 as discussed in my last post. However it isn't accompanied by a shift in the atmosphere as remarkable as the summer pattern. That said, as noted in the recent Met Office briefing on the UK 2013/14 winter strorms, PDF, heavy rainfall events may be becoming more frequent.
"However, there is now some emerging evidence that, over the UK, daily heavy rain events may be more frequent (Figure 22). What in the 1960s and 1970s might have been a 1 in 125 day event is now more likely to be a 1 in 85 day event. This supports other evidence that UK rainfall is increasing in intensity."
And as Dr Julia Slingo states:
"There is no evidence to counter the basic premise that a warmer world will lead to more intense daily and hourly rain events."
So perhaps I am being too sceptical of the probability I've calculated and should be viewing this appearance of 6/10 of the most extreme wet winters coming from the last quarter of the available data with more weight. However rainfall in the UK is very variable, graphing the UK rainfall data shows that rainfall was high in the early period as well as the late, the presence of 3 years from the first quarter of the time series (1911 to 2014) in the top ten seems to support this, but with a probability of 0.25 that really isn't so remarkable. One must also be careful not to fall into the muddled thinking of presuming that because there was high rainfall in a  past period that disproves an anthropogenic role in any possible current increase in winter rainfall. Were one to follow such faulty reasoning one might end up presuming that people don't cause forest fires because there are natural forest fires not caused by people, which is of course ridiculous.

My conclusion is that the signal of climate change in winter weather is only marginally arising and is still largely within in the noise of natural variability. But when, with regards to this winter, Dr Julia Slingo states "all the evidence suggests there is a link to climate change," I agree. As an engineer I often follow my nose in investigations, and my hunches have made me hit the off switch on many occasions, which has saved a lot of money in repairs. So as I suspect is the case with Dr Slingo, 'all the evidence' does not mean to me 'every scrap of evidence' but the sort of impression gained by sitting back looking across a broad range of evidence and asking 'what does my gut say'. As I've said previously, while watching what has been unfolding this winter I was thinking 'climate change'.

However, gut feeling aside, as the planet warms in the decades to come, the signal will rise from the background of natural variability, at the very least Clausius Clapeyron tells us that. As with Arctic sea ice, all it takes is patience, we're watching a geological scale event unfold, it's already happening instantaneously from a geological perspective. If the denialists want incontrovertible evidence all they have to do is sit back while real scientists do the leg work, and shut up.

EDIT - It's occurred to me why the 1990s rain in winter was bothering me, in the 1990s the Arctic Oscillation went into a highly positive mode (affecting sea ice), this could account for record winter rain in the 1990s.

No comments: