Click images for more details



Recent posts
Recent comments

A few sites I've stumbled across recently....

Powered by Squarespace
« Society rejects action on climate | Main | On food and fearless advisers »

England and Wales rainfall trends

In the wake of the "more rain and more intense rain" story, Doug Keenan sends this graph of England & Wales rainfall records for 1766-2012 (click for larger; data here).

Let's just say the trend towards more rainfall is not obvious. (As indeed is any trend towards less rainfall, which is said to be more likely by the UK Climate Impacts Programme).

[Updated to show England and Wales, rather than UK]

PrintView Printer Friendly Version

Reader Comments (71)

I am intrigued by the way the Met Office has been talking about "the wettest year on record" using records going back 100 years or so, when as the graph shows, their records go back far longer. Is there any innocent explanation for this, or is it just a device to allow them to chant their "Tale of Two Cities" worst of times mantra? Dr Betts can you help?

Jan 5, 2013 at 9:56 AM | Unregistered CommenterDavid S

Well I predict, with high confidence, that there will be a lot of weather this year! [And you can quote me on that!]

Jan 5, 2013 at 9:57 AM | Unregistered CommenterIan E

p.s. I'm a bit puzzled by the graph - is the 2012 point missing: the highest recent point looks to be at about the year 2000 or so.

Jan 5, 2013 at 10:01 AM | Unregistered CommenterIan E

I think this must be the annual data for England and Wales and not the UK. According to the linked site monthly totals are available for England and Wales back to 1766, and for Scotland and NI back to 1931.

I see also that the daily figures are available for England and Wales are available back to 1931. My understanding was that the preliminary MO analysis of intense rainfall events only went back to 1960. This is intriguing as the records are available back to 1931. I think I'll download the data and have a look at what they might reveal.

Jan 5, 2013 at 10:12 AM | Unregistered CommenterPaul Dennis

England Wales, Northern Ireland, Scotland & UK (6 countries)
Rainfall, wind, temperature, sunshine (4 variables)
Max, Min (x2)

= 48 yearly "records". As such we expect a 1 in 100 yearly record to be broken every two years.
Take each month and we have 576 monthly records from which to choose each year so we expect around 6 monthly "all time" records each year.

The only thing which has changed, is not the number of records or the frequency of breaking them, but the gullibility of the public to accept that an "all time record" has any meaning at all.

Now, let's add a few others ... most intense rain, highest gust, pressure, humidity.

Now let's add some other easily computed stats ... most "extreme" years (where max and min deviate). Lowest daytime, highest nighttime. Most extreme day-night. Days with/without rain. Longest run of frost, longest period of unbroken sunshine.

Now, we are probably getting "once in a hundred year" events twice a year.

Now, start reporting by individual towns, and cities "Little wettington had the driest ...", since these are not all independent, we can probably expect an order of magnitude rise leading to several "all time records" every month.

Jan 5, 2013 at 10:15 AM | Unregistered CommenterMikeHaseler

There is an upward trend in the data, of 0.1744 mm per year. Or a little over 1/2 an inch a century.

Its insignificant though, as the standard deviation is nearly 120 mm. I am not a statistician, but I suspect that nearly two orders of magnitude between the standard deviation and the trend, makes the trend no different than zero.

Oddly, the Met seems to have missed the years 1872 and 1768, both of which had more rainfall than 2012.

Jan 5, 2013 at 10:35 AM | Unregistered CommenterLes Johnson

I assume that the y-axis is denominated in mm and not gallons or wellington boots.

Jan 5, 2013 at 10:38 AM | Unregistered CommenterDavid Chappell

Is the graph accurate? I found the following data on the Met Office's site

The graph shows a data point around 2000 with well over 1200mm of rain but I can't see it in the table (1194.8 is stated).

Can anyone confirm my findings, I would hate to give the warmists any ammunition as it is their job to manipulate data.

Jan 5, 2013 at 10:42 AM | Unregistered CommenterSteve Jones

p.s. I'm a bit puzzled by the graph - is the 2012 point missing: the highest recent point looks to be at about the year 2000 or so.
Jan 5, 2013 at 10:01 AM | Unregistered Commenter Ian E

Yes, having looked at the bigger graph, I reckon the last data point is 2011.

The Telegraph put together a graphic based on the Met Offfice's UK totals from 1910, which basically shows 2012 was nothing unusual:


My take on all this is that the alarmists are just getting desperate, spinning any weather and data to suit the CO2 thesis. Remember that the record annual rainfall for England is still less than the average annual rainfall for Scotland, hence if the average track of the jet stream is a little further south than usual then England gets a fair bit more rain. It has nothing to do with the alleged warmer atmosphere having more potential to store H20; if it was why did north-west Scotland have a drought in the spring and early summer? More bollocks from the Met Office. The UK weather and climate is determined by the track of the jet stream (and moderated by the Gulf Stream), and CO2 has feck all to do with it.

Jan 5, 2013 at 10:53 AM | Registered Commenterlapogus

Good job someone has all that copied down somewhere - how long before we find out that rainfall readings before say 1960 (1954 was the 3rd highest in the last 100 years) were all over-readings and have to be adjusted downwards.

Of course nobody would fiddle with rainfall and temperature records in that way, or use just a faction of the available data to drive a trend.

Jan 5, 2013 at 10:56 AM | Unregistered CommenterRetired Dave

Readers interested in UK rainfall may be amused by a CA post of a few years ago by Hendry, a very prominent econometrician: .

Hendry posited a theory of inflation as follows:

A second example will clarify this issue. Hendry’s theory of inflation is that a certain variable (of great interest in this country) is the “real cause” of rising prices. I am certain that the variable (denoted C) is exogenous, that causality is from C to P only and (so far as I am aware) C is outside government control although data are readily available in government publications...

there is a “good fit”, the coefficients are “significant”, but autocorrelation remains and the equation predicts badly. However assuming a first order autoregressive error process at last produces the results I anticipated; the fit is spectacular, the parameters are “highly significant”, there is no obvious residual autocorrelation (on an “eyeball ” test and the predictive test does not reject the model [see the Figure below]

Hendry then explained how he was able to

Hendry then explains how he was able to improve on monetary theory of inflation:

My theory performs decidedly better than the naàƒ⮶e version of the monetary one, but alas the whole exercise is futile as well as deceitful since C is simply cumulative rainfall in the UK. It is meaningless to talk about “confirming theories” when spurious results are so easily obtained.

Sort of like Upside Down Mann.

Jan 5, 2013 at 10:59 AM | Unregistered CommenterSteve McIntyre

Interestingly, when one looks at the seasonal totals in that data, 2012 does not get in the top 10 for Winter, spring or summer. The fall makes it in the top 10, but barely.

In other words, it was not particularly wet in any season, but the total for the year was wet.

Jan 5, 2013 at 11:02 AM | Unregistered CommenterLes Johnson

Is that a 'splatter' chart ?

Jan 5, 2013 at 11:04 AM | Unregistered Commenteresmiff

BUT we were told 2012 MAY have been a record year for rainfall!!!! It was all over the BBC, therefore it must be true!


Jan 5, 2013 at 11:06 AM | Unregistered CommenterMailman

Hi Steve, are you still upside down in New Zealand? ( hope you had a good Xmas and belated wishes for the New Year).

Jan 5, 2013 at 11:06 AM | Registered Commenterlapogus

Happy New Year Steve. Hope you are having a great time although we are badly missing CA.

Jan 5, 2013 at 11:15 AM | Unregistered CommenterDavid S

Steve McIntyre -

I seem to remember a book 50 years ago called the "Use and Abuse of Statistics. It had an example of an almost perfect correlation between winter temperatures in Winnipeg and woollen blankets sold in Bradford UK.

Depending on the wave length of Northern Hemisphere waves the two things are not completely mutually exclusive.

Nevertheless it was an early warning of how easy it is to fool oneself.

Jan 5, 2013 at 11:23 AM | Unregistered CommenterRetired Dave

I noticed this week (presumably in the Tel) that 2012 is now said to have been the wettest since .... 2000.

Jan 5, 2013 at 11:35 AM | Unregistered Commenterdearieme

The Met Office continues to suffer from its recently acquired pretensions about climate. Careless remarks about BBQ summers and snowless winters and droughts in the UK have all been followed by Mother Nature failing to comply with their wishful thinking - the wishful bit being their hope that their faith in the power of CO2 in the system, or at least in computer models giving it a powerful effect, can be relied upon.

Apart these blunders, we have seen seasonal forecasts so vague that scarce a passenger on the Clapham omnibus would not feel resentful about their tax monies being used to pay for them. Yet the pretensions remain. Here for example is Richard Betts falling a little short of the billing that he would be telling us 'what we might expect to see under climate change' with regard to drought: If we could relay that to omnibuses throughout the UK, and then interview the passengers for what they now might 'expect under climate change' with regard to drought, I daresay they might not claim to have a clear vision, or even a clearer vision. Some might even reminisce about the good old days when the Met Office was dedicated to the humble but worthwile task of gathering observations and providing decent weather forecasts for a few days ahead. Others might start grumbling about the wisdom of letting governments have so much in the way of taxes to spend.

Jan 5, 2013 at 11:41 AM | Unregistered CommenterJohn Shade

Retired Dave - indeed, a correlation between severe Danish winters and concurrent mild winters in West Greenland was first noticed by a Danish Missionary in the 18th Century. The periods of jet steam 'blocking' we experienced in three of the previous four winters are nothing new.

I regularly look at this 300mb jet stream forecast rather than trust the Met Office's Unified Model for the next week's weather.

Jan 5, 2013 at 11:45 AM | Registered Commenterlapogus

Les Johnson

The effect you mention is visible in the PR-ed MO graphs as well.

UK Spring/Summer/Autumn/Winter look less suggestive than UK Annual. The seasonal ones seem to have been more in synch with each other the last few years and this is why the annual total has the bump. What this signifies (if anything) I have no idea.

Jan 5, 2013 at 11:47 AM | Registered CommenterPhilip Richens

It's Raining MET! Hallelujah! - It's Raining MET Amen!
I'm gonna go out to run and let myself get
Absolutely soaking wet!
It's Raining MET Hallelujah!
It's Raining MET Every Specimen!
Cats and Dogs says the Team
Rough and tough and strong and mean

Jan 5, 2013 at 11:47 AM | Unregistered CommenterStacey

This post looked specifically at the South West. Conclusion also was no clear trend.

If you look at the totals for 2011 and 2012 as % of average the N-S split flips - last year was drier than normal in in England and wetter than normal in Scotland.

Jan 5, 2013 at 12:04 PM | Unregistered CommenterVerity Jones

Any argument that says there is a recent trend towards more rainfall essentially depends on saying 2000 and 2012 were rather wet while the relatively dry 1921 was a long time ago. This looks thin to me.

Jan 5, 2013 at 12:16 PM | Unregistered CommenterHenry

If I can see anything in that graph at all it is a sinusoidal curve of maximums sitting on one of minimums. Extrapolation of that says we are heading for a period of low maximums and high minimums. As the MO is predicting the opposite and they have a history of being wrong, I am prepared to put a fiver on my prediction: any takers?

Jan 5, 2013 at 12:17 PM | Unregistered Commenterssat

One of the refrains is that rain is getting heavier. A bit of digging (see this comment: found analysis from CRU that purported to show increase in heavy rainfall in Winter in recent years - The caption to Fig 2 reads (bold mine):

"Figure 2: The contribution to each winter's total precipitation made from "heavy" precipitation days. The 1961-1995 average is indicated by the dashed line. The shading represents the uncertainty due to the incomplete spatial coverage with rain gauges. A black smoothing line to highlight decadal variations has been overlaid. The different colours depict periods with different raing gauge coverage; for example the early (purple) period was based on 37 rain gauges, while the most recent (blue-green) period was based on 544 rain gauges. From Maraun et al., 2008."

Now with more gauges isn't it more likely to capture regional variations in intensity and pick up relatively localised downpours? Sheesh!

Jan 5, 2013 at 12:19 PM | Unregistered CommenterVerity Jones

As judged by the flood level marker near the Cathedral steps in Worcester, there was a rainfall event in 1670 in the Severn catchment which has not been surpassed.

Jan 5, 2013 at 12:22 PM | Unregistered CommenterFilbert Cobb

I'll be getting back on the horse in a day or two. It's hard to figure out where to start.

Jan 5, 2013 at 12:22 PM | Unregistered CommenterSteve McIntyre

Hi Steve -
Perhaps ask the journal Psychological Science the exact status of 'Lewandowsky's 'moon' paper...

as he is still refering to it in interviews etc, but as yet not published, bringing the journal into disrepute?

Jan 5, 2013 at 12:27 PM | Unregistered CommenterBarry Woods

Dec 2012 figures are not yet included, so the analysis is only valid up to 2011.

There is my analysis here using the Met Office numbers back to 1910, which I will be updating in due course, when the England & Wales numbers are in. I do expect 1872 to be remain the wettest year though.

Also worth noting that Met/DEFRA models all predict drier summers, and also drier south of England, and wetter north. In reality recent years, not just 2012, have been the opposite.

Jan 5, 2013 at 12:30 PM | Unregistered CommenterPaul Homewood

Phillip: Oddlly, using the reference you gave...

...both sunshine and rainfall have been increasing for about 50 years, on an annual basis.

Jan 5, 2013 at 12:31 PM | Unregistered CommenterLes Johnson

The Met Office graph shows a clear trend (click country, Rainfall, Annual), mainly contributed by Scotland. So after independence, you will be able to claim there is no UK trend!

Jan 5, 2013 at 12:44 PM | Unregistered CommenterBitBucket

Verity Jones - I was going to make this point. The 'historic' rain gauge data is very sparse, and natural / spatial variability makes the early records almost arbitrary at best. Even nowadays in Scotland, I doubt the Met Office rain gauge stations give a representative figure for the whole country (if indeed that is possible given the topography); e.g. Fort William has an average annual rainfall of 2000mm, Ben Nevis summit (less than 5 miles away) is 4000mm, Aviemore (about 50 miles east, but also not far from 4000ft mountains) only 950mm.

Using annual river flow level data would be a much more representative method for calculating regional rainfall averages (abstraction and evapotranspiration rates are not significant in most Scottish river catchments).

Jan 5, 2013 at 12:57 PM | Registered Commenterlapogus

Les: This is what the fuss is about I think, and why Guardian and others similar tend to point out "trends" starting around 1960. Not only these series, but also the sensitive temperatures in the Arctic region have been on the increase since then. They tend not to focus on earlier times because they say the data quality is poor. There is of course another possible reason why they choose to do this.

BTW, Funnily enough, doing a running mean on the data pointed out by DK I can detect an increasing trend in the annual data over the entire 200+ year period amounting to a total of around 5%. Might have made a mistake?

Jan 5, 2013 at 1:03 PM | Registered CommenterPhilip Richens

Phillip: I get an increase of about 5% over the entire period too, comprised of about 0.17 mm/year.

Again, I am not a stats expert, but I don't believe this to be significant.

Jan 5, 2013 at 1:12 PM | Unregistered CommenterLes Johnson

Les: Thanks for checking that. I agree, it's not much. Would be interesting to know the reason for it though.

Jan 5, 2013 at 1:19 PM | Registered CommenterPhilip Richens

And who can remember the great flood of 1771

Jan 5, 2013 at 1:20 PM | Unregistered Commenterjamspid


There is another side to the stats you explain. Maybe you can fill the gaps in my statistical ability for me. Firstly, claimed records by the climate folks invariably turn out to be qualified to within an inch of their lives. They never (to date, to my knowledge) have produced any example of a definitely unprecedented weather or climate event. Second, if there were something happening in climate, the figures you cite for records would be much different. The low level records you cite would be broken on a regular basis far in excess of the probabilities you quote. Isn't that the case? If there was global weirding we'd see real new extreme records pretty much every day?

Jan 5, 2013 at 1:56 PM | Registered Commenterrhoda

Jamspid - I can't remember the Tyne flood 1771, but it was certainly a one in 500 and possibly 1 in a thousand year event; the flow at Hexham was estimated to be 3900 cumecs, from a catchment area 1970km2. This makes all other UK flood events since look fairly minor, though the 'muckle spate' of 1829 on the Findhorn also matched the c. 2500 cumecs of the Tay in 1993. (but from a catchment only 17% of the area of the Tay at Ballathie).

Jan 5, 2013 at 2:02 PM | Registered Commenterlapogus

Les: My apologies, I missed your first comment about this, I see you'd already done the calculation.

Jan 5, 2013 at 2:05 PM | Registered CommenterPhilip Richens

@lapogus Jan 5, 2013 at 12:57 PM
my thoughts too - starting a bit more digging...

Jan 5, 2013 at 2:10 PM | Unregistered CommenterVerity Jones

As judged by the flood level marker near the Cathedral steps in Worcester, there was a rainfall event in 1670 in the Severn catchment which has not been surpassed.

Jan 5, 2013 at 12:22 PM | Filbert Cobb

Oh how you skeptics keep droning on & stating the bleedin obvious. We don't need inconvenient truths splattered all over the papers, it just won't do!!! ;-))

Jan 5, 2013 at 2:19 PM | Unregistered CommenterAlan the Brit

WUWT highlighting some regulars.

Jan 5, 2013 at 2:21 PM | Unregistered CommenterLord Beaverbrook

It is important to remember that the annual rainfall data are spatially aggregated point measurements. To get an average over England and Wales, the Met Office will have had to apply averaging techniques, the simplest of which is the Thiessen Polygon method. This introduces additional error and it amuses me that the average over the region, which is actually only an estimate, not a measurement, is quoted with an implied accuracy of a tenth of a millimetre.

One of the problems with comparing average rainfall over a region over a long period is that the number of rainfall stations over time changes. For example, in the 1700s there would have been far fewer gauges. So identifying a trend is even more questionable given the errors in the estimates of annual regional rainfall with biases introduced by the location of the stations.

Jan 5, 2013 at 2:26 PM | Unregistered Commenterpotentilla

Phillip: My guess is that the dryness of the LIA is the cause of the small increase over two centuries. The LIA in the 18 and early 19th century should have been dryer than the 20th century.

I did a quick look at taking 50 years off, in several steps. I went 1766-present, then1766 to 1950, then 1766-1900 etc. For the end dates in the 18th and most of the 19th century, the trend was very slightly negative. By 1900, the trend had gone flat, then up to 0.1 mm/year by 1950, and finally 0.17 mm/year by 2012.

This was confirmed by taking 50 year increments off, starting in 1766, and coming forward. 1766-present,1800-present, 1850-present, etc. The slope was postive only after about 1800, and increased with each step after leaving the 18th century. This indicates low precipitation early in the record.

It could still be sign of GW, as most of the slope increases were in the latter part of the 20th century.

Interestlingly, from 1980 to present, the slope is highest, at over 3 mm/year. Until I take 2012 off, then it goes flat.

But, it is a very small signal in a lot noise. I would not read too much into this.

As I have seen in other posting, even a perfect sine wave can have a positive or negative trend slope. It depends on the start and stop points. If you start in a trough, and end in a peak, the slope will be positive. If you start in a peak, and finish in a trough, the trend will be negative.

Jan 5, 2013 at 2:33 PM | Unregistered CommenterLes Johnson

I reckon I'll stick to the Hanging Stone On Fence method of weather forecasting..
Stone dry and casting a shadow - sunny.
Stone wet - raining
Stone swinging about - windy.
Can't see stone - foggy.
Works for me....

Jan 5, 2013 at 2:41 PM | Unregistered CommenterDavid

On Twitter, Mark McCarthy points out that the UKCIP points to reduced summer rainfall. Winter is supposed to increase. I don't think this is what we have seen in recent years though.

Jan 5, 2013 at 2:58 PM | Registered CommenterBishop Hill

lapogus - thanks for the link to the 300mb output - very useful. As you might suspect I am quite familiar with blocking, but I had not heard reference to the 18th century experience you mention - thanks.

Filbert Cobb - the 17th and early 18th centuries are quite interesting. I had not heard of the flood event you mention, but of course there was the great storm of 1703, thought by many to be the worst storm recorded in southern UK (who knows?). Also the 17th century had the highest frequency of storm/hurricane events in eastern USA in the last 700 years of sediment records. I know the climate doesn't understand the calendar, but the 17th century was probably one of the coldest of the Holocene Interglacial.

Les Johnson - I don't know about rainfall, but there have been quite a few papers showing how cloud cover has deceased by a few % in that period.

Jan 5, 2013 at 3:10 PM | Unregistered CommenterRetired Dave

I note that the rainfall series after 2006 have been 're-calculated'...'to counter the effect of various station closures over the past several years and to take advantage of some new stations which are now available'.

I also note that one particular individual crops up in every single one of the five references quoted below the Metoffice Hadley UKP data table, a certain PD Jones of CRU-UEA fame. Small world.

Are the records the raw truth, the whole raw truth, and nothing but the raw truth? So help me God?

I guess so, if for no other reason that they still haven't decided which is more alarming- a trend toward drought or a trend toward flood, and are still hedging their Betts. (sorry Richard- no offence intended). That is to say, provided they're not lost, deleted, value added, phantom gridded or anything like that.

Jan 5, 2013 at 3:19 PM | Registered CommenterPharos

Sorry, forgot the links for my earlier comment.

Met have previously predicted drier summers, and drier South of England, along with wetter winters and wetter Scotland and North.

Recent years, as well as 2012, have been the opposite.

Jan 5, 2013 at 3:35 PM | Unregistered CommenterPaul Homewood

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>