Met Office sea level forecast, update

Posted: February 11, 2014 by tchannon in Analysis, Natural Variation, waves, weather

Tim writes

According to Richard Betts, a Met Office go between, an original publication omitted fundamental information in relation to forecast. It has now been amended by adding a base date of 1990.

The projected SLR of 11-16cm by 2030 for the English Channel comes from the UKCP09 projections, see this report, Table 2, columns for “High” (16.0) and “Low” (11.4) for London.

But crucially, these numbers are relative to 1990 (the UKCP09 baseline), not 2014. This was not stated. Clearly there’s been some sea level rise since 1990, so the numbers between 2014 and 2030 would be smaller.

The omission led to the impression this was a new forecast when actually it is historic and I suggest misleading. UK09 is a report using a near 20 year old baseline. Betts says this is for London! London is on sedimentary strata and an artesian well, plus ground conditions are unstable.
Is that for a gauge or a model guess?

Why choose a London projection then play switch to English Channel? Seems odd to me.

Unimportant, the sea level change over a wide area is around 1.7x mm/year where London will be the same.


My figure of 1.74, came from a few seconds of software run here on annual data. What follows shows 1.77 is closer, of no importance.

Readers only interested in the Met Office matter can skip the rest. Boring Tim chit chat.

A general look a the Newlyn data

On looking at the monthly data it was immediately obvious the annual as I had used for a quick casual work was as I said corrupted by Nyquist violation. This is a common problem.

The original quick analysis included this pair of periodicies 3 and 6.01. Three doesn’t happen, at least that close. When plotted the two form a classic sign of bother.

With monthly data which will still be poor but probably less so, the 3 vanishes.


Local sea level data does often have a  clear annual cycle.


Based on monthly there is a good shape to the annual cycle and the explanation will be atmospheric pressure change, with the highest mean levels late autumn with storms and the lowest during summer.


Long time readers will know what I have done here. A non-descrete Fourier model of the dataset has been approximated, result looks adequate based on experience.

The construction contains an extremely slight rising trend (not talking linear trend here), assume ky. The residual does not show a rational deviation from flat so I am stating there is no sensible rising trend.


Lunar factors are there including a circa 18y wobble which could mislead someone into assume a rising trend at the end.

I told the plot software to compute and display the linear trend. (see follow section, Other information)

Signs there of Nyquist having been broken, not a surprise given the underlying data is very fast moving and sampling is inadequate, plus probably improper dataset decimation in time.

That will do, not worth putting more time into this.

Other information

Isabel B. Araújo and David T. Pugh (2008) Sea Levels at Newlyn 1915–2005: Analysis of Trends for Future Flooding Risks. Journal of Coastal Research: Volume 24, Issue 4A: pp. 203 – 212.

Which finds 1.77 ± 0.12 mm y-1 (up to 2005)

See figure above, agrees.

Sites have history, so it is for Newlyn.

Sonel have a coincident site. Photo of cabins.

First try ever at embedding a map. Not quite as intended, too late to fiddle with it. Station is at the end of the south harbour breakwater. Can switch to aerial image.

[update1, additional material]

History and photos

The UK Fundamental Benchmark

The Tidal Observatory was established to determine the mean sea level that is the starting point for levelling in the UK. This brass bolt is the benchmark for the whole of the United Kingdom, that is, all heights are referenced to this point.

The height of the benchmark was established over a six year period from 1915 to 1921, when visual observations of water level on a tide staff were made every 15 minutes, 24 hours a day, 365 days a year.

From the data collected over this period, mean sea level was found and this vertical level transferred to the head of the bolt.

History of selection as UK reference

To take all the readings required to calculate Mean Sea Level, you needed a tide gauge. Examples were set up Felixstowe (1913), Newlyn in Cornwall (1915) and Dunbar (1917). Subsequently it was decided to solely rely on the tide gauge at Newlyn, which is the one we’ve got here in Southampton, stored out of use.

It was chosen over the others because it was situated in an area of stable granite rock and because the gauge was perched on the end of a stone pier at the harbour entrance it was exposed to the open Atlantic.

DEFRA and EA from 2007, technical report
Move along there, we ain’t got nuffin

Paper on GPS monitoring of UK references


Posted by Tim

  1. […] Update 4: The Met Office have ‘clarified’ their statement, which apparently relates to a 1990 baseline. See Richard Betts’ comment below. Thanks, Richard, for your efforts on this. See also Tim Channon’s new graph here: […]

  2. tallbloke says:

    Nice work Tim. Looks to me like my eyeball estimate was near enough for coarse fishing. At 30-75mm too high. The fact we’re looking at an inch or three difference between long trend and MET-o’s worry-mongering tells us sea level rise in the C21st is a non-issue.

  3. p.g.sharrow says:

    Actually it appears to me that the trend is down for the last 10 years. WUWT, did they quit pumping the aquifer? How long has the barrage been in place? That would cause water table lift. pg

  4. Stephen Richards says:

    When your Met Off publish these figures (wrong data) do they 1) not check their clarity before publishing. 2) deliberately mislead in the hope and expectation of fooling the sheeples. 3) Are they just plain incompetent.

  5. tallbloke says:

    STephen: Definitley number 2. Even after the update to the document it left the same impression to less techy readers of fast sea level rise. And left the techy reader to do calcs like Tim’s to be able to see the true picture (less than alarming). One to three inches above long term trend doesn’t really set the alarm bells ringing does it?

  6. tallbloke says:

    Blimey, an almost sensible and informative article on weather and climate from the graun:

  7. oldbrew says:

    TB: after highlighting the role of the jet stream, they still slip the propaganda in at the end.

    ‘”As the air warms it can hold more water.”

    Warming is static for over 17 years. Fail.

  8. Me_Again says:

    Is the sea level rise as a result of human activity or does it sit in the interglacial expectations cubby hole?
    Has it accelerated, remained steady or diminished?
    Is Britain in a stable tectonic region whereby the plates are not rising or slipping?

    Must be really difficult to determine whether any change is as a result water level rising or land slipping.

  9. michael hart says:

    I expect Dame Julia Slingo will be issuing some general thanks for those who help out the Met by reviewing their publicly released documents for them.

  10. Keith W. says:

    Note that for the South of England much of the apparent sea level rise is actualy due to land sinking at a rate of around 1 mm per year due to post glacial isostatic rebound. At the same time land in Scotland is rising with the pivot point being pretty much along the line of Hadrians Wall


  11. […] Keith W. on Met Office sea level forecast,… […]

  12. tchannon says:

    Tim writes,
    Seems that was legible enough but I forgot various things.

    Sonnel is part of an altimetry network, an increasing number of major tide gauges are GPS monitored for height change. Difficult site, hence the damaged aerial they mention.

    I’ve not looked to see whether the site is on Cornish granite; geology of tide gauge sites is often fascinating but in this case I doubt there are pumped aquifers. A case I investigated but have not written on so called accelerating sea level rise turned out to be deep aquifier pumping (dropped level 100ft then stopped pumping); claimed bedrock turned out to the a scam claim, thin layer of broken limestone over deep alluvial with complex aquifiers, oil and gas work, lots of stuff. End result a straight line similar to Newlyn.

    I forgot to give links to the prior articles at the Talkshop.

    What else?

  13. tchannon says:

    “Chaeremon says:
    February 11, 2014 at 8:44 am

    @Tim, the paper you mention Sea Levels at Newlyn contains a treasure trove in the references, have a look at the full abstracts of (Australians):”

    Ah yes, I forget to mention is paywalled. I’ve not read it.

    If as implied by the abstract there is long term hourly data an analysis might show some detail of lunar effects and more about the effects of the around-the-corner location.

    Just remembered something I forgot to add the Newlyn plot of probably the storm surge with did for the wall at Dawlish but I have found evidence the Dawlish matter is about relatively recent defective work, looks like head rolling stuff. Won’t happen, cover up.

    When I was looking at the Dawlish matter I came across this newspaper photo of same storm damage at Newlyn.

  14. tallbloke says:

    No problems Tim.

    Nice little sketch here shows what we’re up against:

  15. Does anyone know why are there recent gaps in the Newlyn tide gauge data at the PSMSL (Permanent Service for Mean Sea Level) site? Why data for last year are missing?

  16. tchannon says:

    Not details. Did come across some vague words.
    The bubble was moved 2010, might be a gap and lead to rejection of readings under quality control rules for that period.

    I’ve added at update to the article giving some additional material.

  17. Roger Andrews says:

    Peter B: The ten month gap in the Newlyn record in 2010 was probably caused by relocation of the bubbler gauge. Don’t know what caused the four-month gap in 2007.

    PSMSL hasn’t posted any tide gauge data for 2013 yet. I guess these things take time.

    According to the best data I can find the Channel coast is sinking isostatically by about 1mm/year, and if so then only about 70m of the total 170m sea level rise recorded at Newlyn over the last ~100 years was caused by rising seas. Hard to get excited about that.

  18. Steve Richards says:

    Why would anyone ‘re-release’ an old paper on sea level rise and mention London, just now?

    With all of the floods in Somerset and now London, the warmists have not missed out on a chance to get out one of there dodgy claims again.

    The news in the UK is full of ‘there is evidence that may indicate dangerous changes to….’

    Getting the word ‘dangerous’ or ‘high’ or ‘bad’ in every sentence, but softly speaking the word ‘may’.

    It seems such a shame that the media have no one to challenge this nonsense when these standard phrases are voiced……

  19. tchannon says:

    I live where there are two rivers, perhaps many more sheds. (under cover operations)

    Had a look at the water, seen this level before.

    Mentioned to the missus I was going to dig out the books, she said two were in the corner of x bookcase, need a map hereabouts. Nope, there are 6 on history in this small town.

    Won’t go into the saga far back, nothing new. Most fun is a photograph showing flooding in very identifiable location, must be above where it is at the moment.

    No photos but a report on extensive flooding, far above at present I think 1885, quite a few dates. I have a lifetime of local knowledge anyway.

    Something I had forgotten was a photo of one of the snow episodes, “18 inches” of snow in April and this extended over a wide area, the photo suggests less but still more than I recall actually seeing. (I was away 1963)

    And so on. Are these agony aunts on the same planet?

    Floods are relatively well recorded in history.

    Not entirely but it is my belief there are patterns. This fits with non Gaussian (so don’t use Gaussian assuming statistics which is why there are claims of 1000 year events and so on).

  20. Doug Proctor says:

    The only reason you would go for the Met Office high or low is if you have bought into the CAGW warming/IPCC melting narrative. The 1954 – 1975 could easily be the 1991 – 2012 period.

    The last two years have been ignored: that is also because the IPCC narrative drives the interpretation. What the skeptics call “data”, the Met Office is calling “anomalies”, or “outliers”.

    How many years of sea-level data will it take the Met Office to change their sea-level assessment? However many years it takes to drop the IPCC assessment: without dropping the temp rise, you can’t drop the sea-level rise.

    The sea-level people are in a bind. A big one.

  21. Brian H says:

    The linkage between air temps and sea level seems to be almost entirely a matter of IPCC and MET assertion. Is there any contemporary correlation data?

  22. tchannon says:

    Not so far as I know. With too much science these days correlation is causation.

  23. Me_Again says:

    This lady has her head screwed on the correct way round.

    She is a bit ‘sarky’ but it ticks all the boxes for a ‘non met office-non-climate scientist’ like me.

  24. tallbloke says:

    Guardian falls for it

    “The joint report, from the Met Office and the Centre for Ecology and Hydrology, entitled The Recent Storms and Floods in the UK, points out that the 12cm (4.7in)rise in sea level over the 20th century has exacerbated coastal flooding. It says a further rise of between 11cm and 16cm is expected by 2030, two-thirds of which is attributable to the effects of climate change.”

  25. tallbloke says:

    The whole issue of sea level rise has nothing to do with storms anyway. It does affect flood levels, marginally on the timescale of human lifetimes, but even then, the phase of the moon and the air pressure level has a hundred times more effect on surge level than sea level rise. And given that sea level rise rate has been pretty much constant since looong before co2 started rising, besdies the 1993-2003 decade, the Grauniad argument fails in it’s larger analysis, as well as falling into the MET-O deliberate obfuscation trap.