Skip to content

Impact Of Sea Level Rise On The Dawlish Line

January 3, 2016

By Paul Homewood 




Last month I reported on a study which investigated the potential effects of sea level rise on the Dawlish rail line in Devon. For those who missed it, my post is here.

One of the authors was Roland Gehrels, and I must thank him for taking the time and trouble to comment and explain much of the detail from the study. 

There was one point which I took particular issue with at the time. This was the claim that during the lifetime of the Great Western Railway, which was completed in 1846, there has been a 20cm rise is sea levels in the English Channel. However, almost half of that occurred in the last 40 years.

My view was that this gave the misleading impression that recent sea level rise had been much greater than in the past. This certainly does not tally with NOAA tide gauge records, which show the current rate of rise less then it was up to the 1960’s. 







Roland correctly states that the rise in the last forty years has been about 100mm, compared to 176mm in the last century. We have no tide gauge data at Newlyn, or anywhere else nearby, before 1916, so we don’t know what sea level rise was beforehand. (The study uses data from Brest for this earlier period). Nevertheless, it would come as little surprise to learn that sea level rise was minimal until late in the 19thC.

But why use this particular 40-year period? As the above graph shows, sea levels were falling during the 1960’s and 70’s, a period of falling global temperatures, partly tied into the cold phase of the AMO. We would not measure temperature trends from winter to summer, so why try to measure sea level trends during only half of the AMO?





As the AMO lasts about 60 years, let’s look at the trend at Newlyn between 1950 and 2009, so we can see the true trend across the whole cycle. (The study itself uses data up to 2009).

This is not cherry picking on my part. It is widely accepted that trends need to be measured over at least 60 years for them to be meaningful, as, for instance, leading oceanographer Bruce Douglas maintained. Also the paper by Chambers, Merrifield and Nerem, Is there a 60-year oscillation in global mean sea level?” found that there is a significant oscillation with a period around 60-years in the majority of the tide gauges examined during the 20th Century, and that it appears in every ocean basin.




We find that the rate of rise is 1.78mm/year, almost identical to the trend of 1.76mm since 1916. Put simply, there has been no long term acceleration in sea level rise. 


This is not just some academic point scoring. As I commented before, there may be all sorts of good reasons to reroute the Dawlish line. And as Roland correctly points out, sea level rise is cumulative, meaning that gradually it will impact more and more on the railway, a problem that is compounded by the fact that the land there is sinking by about 0.5mm/year.

But it is important to base such decisions on realistic and robust data. Which brings me onto the real reason for this post!

The paper uses sea level predictions from DEFRA’s UK Climate Projections on which to base estimates of line disruption and financial impact.





Even on the lower emissions scenario, the assumption is a rise of 4.7mm/year up to 2020, and 5.2mm from 2020 to 2040. This clearly is not supported by real world evidence of what has actually been happening.

The analysis contained in the study may be used to assess the economic viability of rerouting the Dawlish line. However, no decision of this magnitude should be made on the basis of such wholly unrealistic, airy fairy data.


I have asked Roland to contribute to this discussion.

The original paper, “Sea-level rise impacts on transport infrastructure: The notorious case of the coastal railway line at Dawlish, England” is available here.

35 Comments leave one →
  1. John Moore. permalink
    January 3, 2016 7:13 pm

    I live within a few miles inland from Dawlish and have known that line and the beach alongside it since the 1930s when a small boy. I was in touch with the Ordnance Survey some four years ago and they told me that Newlyn Harbour (adjoining Penzance in West Cornwall), which has always been the datum point for contours and spot heights on all their maps, has risen some 7.5 inches since 1916 and that they still use the 1921 reading for their current maps. I would suggest that compared with the variation in heights of high tides depending on Spring Tides (when the sun and the moon are on the same side of the Earth) and on the wind direction and strength, that the few inches calculated as the mean would be insignificant by comparison. A strong Easterly gale would pile up the waves considerably. I suggest these were the conditions that brought the damage last year.

    • AndyG55 permalink
      January 4, 2016 5:12 am

      I love those youtube vids of the preserved steam locos running along the Dawlish foreshore 🙂 🙂

  2. Paul2 permalink
    January 3, 2016 7:23 pm

    We’ve had a similar problem this neck of the woods. A very old wall and the waves on that beach are extremely powerful so there’s no surprise when these sort of things happen:

  3. The Old Bloke permalink
    January 3, 2016 7:24 pm

    By a strange coincidence I have been studying the tide gauges also. One aspect of all this data is that on the Isles of Scilly, sea levels have been falling over the same period that they have been increasing at Newlyn.?? Also, in the Maldives there are three reporting stations. Only one shows an increasing sea level, the other two show decreasing. Further examination of Nordic countries shows a dramatic sea level fall. We are being told that because of Global Warming sea levels will rise, and in view of the fact that we have been told that last year was the “hottest ever”, if that is the case, then why are so many stations been recording sea level falls and over many years?

    • Adam Gallon permalink
      January 3, 2016 9:54 pm

      Nordic countries (certainly Sweden) suffer (If that’s the right word) from glacial isostatic rebound.

  4. J2NH permalink
    January 3, 2016 7:41 pm

    Excellent! I can’t thank you enough for the work you do.

  5. January 3, 2016 8:29 pm

    We return to the issue that the people issuing “scientific” reports for decision makers are not accountable for their articles. This very persuasive tabulation of sea rise looks like “proven” science while it is at best biased opinion.

    This sort of misinformation could lead to taxpayers money being squandered: the authors should have to justify their claims in open debate. This is a problem in UK law: expertise is a fairly arbitrary accreditation and needs challenging but is not.

  6. john cooknell permalink
    January 3, 2016 9:11 pm

    The problem is Brunel was forced to build the Railway on the beach. He did not want to do this for obvious reasons!

    The sandstone cliffs (ancient prehistoric desert dunes) could no longer erode into the sea, erosion used to provide more sand for the beach, the sand protecting the cliffs from storms, the beach has got steeper and steeper and eventually the sea will win!

    • johnmarshall permalink
      January 4, 2016 10:44 am

      Brunel instigated the installation of groynes on the beach to trap sand and build the beach up forming a storm berm that protected the line. Maintenance of the groynes ceased years ago, they no longer carry out the function that was intended and the protecting storm berm has washed away. Result a vulnerable rail line. Reinstating the groynes will build the beach/storm berm and the line will be saved.

      • Gabbro permalink
        January 4, 2016 4:45 pm

        It amazes me just how much of what used to be seen as vital maintenance work to the coasts, river systems and inland waterways has been ignored and for so long. Floods, sea encroachment, crumbling cliffs and shifting sands is the inevitable result.

  7. January 3, 2016 9:16 pm

    There’s little new under the sun. The Great Storm of 1859 washed away all but the west wall of the 12th century church of St Brynach in Pembrokeshire.

  8. January 3, 2016 9:56 pm

    I would have thought that tides, waves, wind and low air pressure storm effects would dominate the estimation of how often the train service would have to be suspended, and that to a good approximation the “background” (unperturbed) sea-level could be assumed to be constant.

    Also, a low cost alternative to moving the whole line might be to simply raise the rails over the short section that floods, or to put some wave dampening mechanism in the water.

    • johnmarshall permalink
      January 4, 2016 10:45 am

      Yes, groynes will do that job.

  9. January 3, 2016 10:17 pm

    “… so why try to measure sea level trends during only half of the AMO?”

    I’m forced to assume that was a rhetorical question.

  10. John Peter permalink
    January 3, 2016 10:21 pm

    “I have asked Roland to contribute to this discussion.” What about justifying his projections?
    I guess you will never hear from him. Send Network Rail a link to this article and warn them about using Roland’s “projections”.

  11. January 3, 2016 11:54 pm

    We keep seeing these 21-year (4 X 5) cycles of the Sun. Wow.

    The impact of solar processes should be background information in the MSM. But it is not. I guess “natural processes” have to be ignored when the anthropocentric are not as obvious or authenticated as the narrative claims.

  12. John F. Hultquist permalink
    January 4, 2016 2:43 am

    In the table there are Low, Medium, High, & H++ emissions.
    Are these estimates of GHG emissions?
    I wonder what happened to the years in the left-most column for the H++ scenario?
    It looks like there could be nearly a 2 m rise in sea level in the next 84 years.
    Does anyone believe that?

    • January 4, 2016 10:50 am

      Yes, they are GHG emissions, John

      I guess the H++ ones are “what ifs” with no specific time frame in mind

  13. John Peter permalink
    January 4, 2016 8:03 am

    “It looks like there could be nearly a 2 m rise in sea level in the next 84 years.
    Does anyone believe that?” Yes – Dr. James Edward Hansen of NASA/GISS fame.

  14. January 4, 2016 8:48 am

    Will the Express & Echo print a modified piece?

    Those climate projections (UKCP09 are being used by The Environment Agency to justify the scale and scope of works on the TE2100 “flood” project in the Thames Estuary. Problem is ….the EA as far as I can see (and I have looked) will not release (as in have deliberately withheld) the engineering/ technical details – preferring to hype fantastical “climate change” rather than expose the innards of their arithmetic.

    It’s all not good enough – at all…..

  15. January 4, 2016 8:57 am

    I had to laugh the other week when the BBC said due to Arctic melting sea ice, sea levels were rising. Every schoolboy who was taught science in the 50’s like me knows that is impossible, Archimedes proved over 2,000 years ago that a simple experiment by filling a jug with ice then topping it up with water to the brim, when the ice melts there is no overflow. Try it yourself, as we all did in class 60 years ago.

    • Le Gin permalink
      January 4, 2016 12:33 pm

      And I made a complaint to the BBC, citing Archimedes Principle. Guess what? Archimedes was wrong, apparently!

      • John189 permalink
        January 4, 2016 1:00 pm

        It would be interesting to see the BBC’s reasoning as to why “Archimedes was wrong”. Melting of the Arctic Ocean is the melting of sea ice. There may be a difference between salt water and fresh water, but I would think that fresh-water ice melting into northern seas would lower their salinity, so any change in sea levels would, if any, be minute. Melting of land ice-cap, e.g. Greenland, is of course a different matter.

      • Le Gin permalink
        January 4, 2016 7:24 pm

        Here you go, the full reply!

        I haven’t had time to put together a response yet, but have checked out the links in the reply, and others I found myself.
        I am not a physicist, so am trying to understand about mass/density differences between sea ice and fresh ice. Tips and hints welcome🙂

        “…Many thanks for getting in touch and apologies for the delay in replying.
        I was sorry to learn you felt our online article on climate change contained inaccurate information.
        We consider we obtained our research from reliable and reputable sources. An article from the National Snow and Ice Data Centre (where the source comes from) states that the ‘melting of floating ice will raise sea level’:
        This article from Yale Climate Connections states that it is ‘land ice’ and not ‘sea ice’ which makes sea levels rise:
        However this article also from the National Snow and Ice Data Center says that ‘ice sheets’ (which is the term in the article) from the Antarctic would make sea levels rise about 6 metres:
        Please be assured that we always try to ensure that anything we publish is reliable and thoroughly researched.
        We of course welcome your feedback and I’ve already made sure your comments have gone to the right people here at the BBC, including the news editors and senior management.
        Thank you again for sharing your concerns.
        Yours sincerely,
        Paul Carson
        BBC Complaints

      • John189 permalink
        January 6, 2016 2:33 pm

        Thank you for the BBC reply. The problem is sloppy use of language by both the BBC and the NSIDC Newsroom (your first link). What they are referring to is ice from land which enters the sea and melts, which will indeed contribute to sea-level rise. The word “Arctic” without specifying ocean or merely the region at the top of the world is, in the context of melting ice, highly misleading,

  16. Kelvin Vaughan permalink
    January 4, 2016 3:37 pm

    This is what they need.

  17. January 4, 2016 4:55 pm

    Paul, I am happy to contribute a bit more. You are of course correct to point out that longer records are better for determining statistically robust trends. This is an important and valid point to make. Tide-gauge records are often gappy and short, but Newlyn doesn’t suffer from this and is one of the best records we have. After all, it’s where our Ordnance Datum was defined. The fact that sea level at Newlyn has risen by about 10 cm in the last four decades was the original issue of dispute if I understood you correctly and this has indeed occurred if you calculate the trend since the early 1970s. I am glad you no longer see this as a ‘fraudulent’ claim. But let’s not get side-tracked as it is not the critical point of the paper – the cumulative sea-level rise is what is doing the damage at Dawlish.

    You now throw the AMO into the mix, one climate index that has some correlation with steric (water density) changes, mostly driven by ocean temperature. If we are interested in climate signals reflected in sea level it is best to look at the global sea-level curve, rather than individual tide-gauge records (I explained this in my previous postings). And, it turns out that global sea level is rising faster now than it has been before. The global sea-level observations can be seen here and the current rate is about 3.3 mm/yr, much faster than the 20th century global mean rate, and certainly faster than sea-level rise in the 19th century when the rate was close to zero (as far as we can establish from available records and proxy reconstructions). See Fig. 13.3 in the IPCC report It is 95% certain that globally the 20th century has seen higher rates of sea-level rise than any previous century in the last 2600 years (a paper on this will be published in the coming weeks – you heard it here first!). So, whether you take a decadal, a centennial or a millennial view, sea levels are on the up at an increasing rate.

    Is 4.7 mm/yr sea-level rise on the south coast of Devon by 2020 under a low scenario unrealistic? I wish it was, but predictions based on available information tell us otherwise. Just take 3 mm/yr and add a little coastal subsidence and you can account for most of the rise – the little extra comes from slightly higher rates of sea-level rise that are predicted to occur in the next 10 years. But there is some decadal variability in sea level, as can be observed from existing observations, and this is not taken into account. The sea-level predictions we use in our paper are taken from the UKCP09 climate predictions, which in turn use the IPCC predictions. The predictions are ‘smooth’ curves, which is a limitation that may be resolved when the next set of predictions becomes available (I think in 2018). So the predictions are not perfect, but it is the best we can do for now.

    The whole point of our integrated approach is that it is also firmly founded on historical data, on ‘experience’, and is therefore less “airy-fairy” than you might think. The methodology is transparent and allows people to come up with different predictions should more, or better, information become available. The only part that is modelling based (and perhaps a bit of a black box for some people) are the UK sea-level predictions from UKCP09. Interestingly, our predictions in the table are conservative compared to engineering predictions (see Fig. 8 in our paper).

    For information how the UKCP09 predictions are put together I refer to this website The predictions are based on our best understanding of how the world works. Being skeptical is always good, but by using words like “airy-fairy” you do a disservice to the hundreds of excellent scientists who have worked on these predictions. To give some credibility to model-based sea-level predictions, it is worthwhile to point out that, since IPCC predictions started in the early 1990s, the observations in the last 20 years have followed the IPCC ‘high’ scenarios. A sobering thought. This can be seen in this open access paper Globally we can expect a rise of about 1 m by the end of this century if we follow the trajectory we are on at the moment. Even if we stop emissions, sea-level rise is likely to continue for centuries, because some processes are irreversible in the short term. Our coastlines are not immune and we don’t gain anything by ignoring the warning signals or by misinforming the public that all is fine.

    Our paper, by the way, is based on the PhD study by Dr David Dawson, now at Leeds, and was funded by the Devon and Cornwall county councils, and also by Network Rail (who, to their credit, gave us complete freedom), and all are fully aware of its findings. I would think that most people in the Southwest would agree that maintenance of the Dawlish railway section is unsustainable in the coming decades. I personally favor the opening of the line via Okehampton and Tavistock. The sooner this is realized, the more money can be saved. We hope our study will make a small contribution to solving this important problem.

    This is the first time I have taken part in a blog discussion in my 25 year career. I must admit that I have no idea who you are, Paul, and I can’t find anything about you on your blog or on Google. Perhaps you can add something about your background? I was possibly seduced because of my love for the Southwest and its beautiful coastline. I lived and worked there for many years and I miss the area. It’s fair to say that it is common for scientists to steer well clear of blogs, especially when a lot of ignorance is on show, and sadly the tone of discussion is often shouty and uncivilized which is not to everyone’s liking. But I have enjoyed the opportunity to provide some clarifications of our paper and I thank you for making some interesting and valid points. In case your readers are interested, an excellent and highly informative climate blog is Worth a look. It’s run by some of the best scientists in the business.

    Now it’s time to go back to my day job. Let me end by wishing you and your readers all the best for 2016. The previous year ended with the Paris climate conference which gives us hope that climate change is finally going to tackled by our governments. May we all remain skeptical and inquisitive, whilst not ignoring the scientific evidence that stares us in the face. Let’s also do the best we can to make our beautiful planet a better place for future generations. It’s the only home we’ve got and we need to take care of it.

    • January 4, 2016 7:07 pm

      Thanks Roland

      Many scientists, such as HH Lamb, believed that sea levels were higher than now in the Middle Ages, and that changes up and down were actually significant before and after. What is your view?

      • January 4, 2016 7:32 pm

        Paul, the 20th century saw the highest rates of sea-level rise (prob. = 95%). Lamb was wrong. All best, Roland

      • January 4, 2016 7:40 pm

        It’s again about distinguishing between global and regional sea levels. Our study is the first global analysis. Lamb, I believe, discussed regional sea levels around the North Sea. He couldn’t have done a global analysis when he wrote his book in 1982 – the data simply weren’t available.

      • January 4, 2016 8:47 pm

        Some of the areas Lamb looked at are actually sinking, which rather strengthens Lamb’s argument that absolute sea levels were higher in the MWP.

        I presume you are aware of the abundant evidence that glaciers worldwide advanced massively during the LIA, thereby contributing to fall in sea levels. How much of the 20thC rise is merely a natural rebound from that?

    • Ktm permalink
      January 5, 2016 7:13 am

      I find it more than puzzling to ignore a perfectly good ride gauge record that is right on site to go use the satellite estimates from Colorado. First of all, those satellite estimates include a glacial isostatic adjustment that is completely irrelevant to discussions about what the sea level will be at particular locations on earth in the future. The sea level gauges are there, any glacial rebound that might affect sea level in the area is already captured by them. Estimates based on the satellite data are not fit for purpose, which suggests an ulterior motive in choosing to highlight that irrelevant data.

      The satellite data itself suffers from the same problem you acknowledged was a fair criticism, since the satellites have only been operational for a relatively short period and can’t capture longer cyclical changes. It is a hybrid record, patched together from sea level gauges in the past then splicing a different measuring system on top of it in recent decades. This again seems like a very poor way to conduct solid science, when long uniform sea level records are available.

      • January 5, 2016 6:37 pm

        The geodetic reference frames and the carnival of adjustments have been abused quite a bit by folk who in some cases should know better.

        What concerns me as a surveyor is that these are rather small measurements taken in the presence of a considerable amount of noise and a whole host of tweaks have to be applied.

        I’ve looked at quite a few sea level analysis papers and one thing is usually missing in those that tend to the “bigger” end… = an honest and comprehensive error budget analysis both for the geodetics and ocean / water properties. I have a bit of an issue with kludging all the seas of the world into an arbitrary overall number too….

  18. January 4, 2016 6:34 pm

    See a Dawlish intercity train get a soaking from storm Frank (last week).

  19. john cooknell permalink
    January 4, 2016 8:13 pm

    I would agree with Roland that the railway line built on the beach has no future. Brunel thought this when he built it!

    Coastal erosion is the problem not sea level rise, Brunel knew about sea level rise and built the line at a level that would be safe for centuries.

    However, he also knew about coastal erosion and built groynes. There were extensive timber groynes at Dawlish, and a massive stone groyne at Langstone Rock, these retained the sand which protected the line retaining wall from storms.

    Longshore drift is from west to east , and the massive stone Groyne at Langstone Rock at the east end of this section of beach retained most of the sand. This stone Groyne was washed away in a storm the previous year, and was not repaired.

    If you wish to read voluminous detail of the Geology and the problems identified then this is as good as anything.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: