Skip to content

UK Sea Level Changes–A Case Study At North Shields

April 12, 2012

By Paul Homewood




Amidst the latest controversy surrounding Envisat and their unexplained retroactive changes to their satellite sea level database (which, yes you’ve guessed it, have increased sea level rise in the last few years), we should not lose sight of tide gauges, which have been monitoring sea levels for a century or more.

Tide gauges, while not being subject to the calibration issues that satellite measurements face, have one major drawback. Many coastal locations round the world are subject to isostatic changes. Since the end of the Ice Age, land previously covered by glaciers has ben slowly rebounding, while others have been sinking. In the UK the effect can be seen on this map.



                         Rates of Isostatic Rebound 
                         in Great Britain (in mm/yr)


An added complication comes when silting and erosion affect coastal areas. However, although these factors can affect absolute sea level changes, they don’t affect relative sea level changes, at least not over short time scales as their effect is a very long term process.


In this series, I will be looking at a cross section of UK sites and examining tide gauge records to see if there are any trends in the rate of sea level rise. We start at North Shields, which is situated on the Tyne in the north east of England. On the map above, it fits perfectly into the “green” zone, where isostatic change is pretty much zero.

Figure 1 shows the annual mean sea levels there between 1896 and 2009. (There is no data for 2010 or 2011). (Full data is available from PMSML – )





Since 1896, the level has increased from 6793mm to 7007mm, an increase of 214mm or 188mm per century or 7.4 inches. This figure is, of course, pretty much in line with global estimates over the last century, so we seem to have picked a fairly representative site!

But has the rate of increase been increasing in recent decades? The evidence from Figure 1 would suggest no. There is a small blip upwards between 1992 and 2002 following a big drop in 1991, but this has already been reversed and the 10 year average line since 1970 is bang on the long term trend.

But we don’t have to rely on eyeballing this graph. We can go one better than that and look at the year on year changes in Figure 2.





The 10 year average shows quite clearly that the rate of change has not increased since the start of the record. Figure 3 shows the 10 year average line in much closer focus.





Sea level rises over the last 10 years are below the long term mean (red line).

DEFRA are forecasting a sea level rise of 13mm / year by the end of the century. Perhaps someone should tell the North Sea.

  1. Robin permalink
    April 12, 2012 6:43 pm

    Hello, Paul, I look at sea level data too, as you’d expect. Have done so for many years, and have talked to Simon Holgate at Liverpool on a number of occasions. He’s published – naturally enough, being an academic- and /really/\knows what he writes about. I downloaded masses of data from PMSML on my dial-up and looked at some of it! It contains a lot of nonsense data aas well as good stuff. Changes to siting or method (or whatever) are not always recorded, and Simon has I think done a great deal of work trying to verify and homogenise what’s there.

    I like your map of isotactic rise/fall. Hadn’t seen one before, and it puts Newlyn, which is one of the longest records I think, firmly in the sinking zone. I’ll have a look in my files for any that you pick out as worthy of attention and let you know my take on them! Meanwhile will get the North Shields stuff.

    Your first difference plot has a lot to commend it. I often use it on temperature series as a verification tool for my treatment of the primary data. For N.Shields it is really very convincing. “No Acceleration” is absolutely established. I’ve also been looking at the stuff from North Carolina, which has hit the headlines in the past. There, the isotactic or other effects seem to be a mystery.

    I always enjoy your posts.


    • April 12, 2012 8:34 pm

      I gather that the North Carolina is heavily affected by movement of sandbanks and erosion. It would be interesting to see if there is any change in the rate of increase.



  2. Robin permalink
    April 12, 2012 8:19 pm

    Well, I’ve had a good look at the data and of course confirm your first difference plot, which has a slope of zero. Could quote values but they are on the RISC OS side so are not accessible at the moment. Cusum plot of the monthly differences (there is a pronounced systematic effect of month of observation) is as expected a fairly smooth and very strong parabolic form, so I regressed the differences on decimal date, generated the residuals and formed their cusum. This now shows what I take to be regime behaviour with some evidence of abrupt chanage points. It needs quite a bit of work to put all this into my Windows machine (same physical device but different OS), but could do it. A quadratic fit to the original observations has a nonsignificant quadratic term, which is NEGATIVE. A tendency for the rate of increase to decrease!. But not proven of course.


    • April 12, 2012 8:32 pm

      Hi Robin

      There seem to be 2 or 3 outliers in the North Shields data, but they seem to correct themselves straight away. That’s why I have gone for 10 year averages.

      Whether these are right or not, I don’t know, but there is less variability now than around 1910. Without these the trend line would have been even smoother.

  3. April 13, 2012 6:28 am

    Paul, what is the origin of the UK map of isostatic changes and what are the units? Do similar maps or figures exist for other parts of the world?

    My interest is mainly on the relative variation of sea levels over a century or more in order to see if there has been any acceleration or deceleration in recent decades. Just as you show for North Shields, I have found there has been a fairly constant trend in sea level changes worldwide since data was first collected in the early-1800s. My analysis covered 53 stations, worldwide, with records spanning periods from 50 to 202 years.

Comments are closed.

%d bloggers like this: