Roger Andrews: Sea level reconstruction from tide gauge data and a match with SST

Posted: September 26, 2011 by tchannon in climate, Ocean dynamics

Roger Andrews has computed an annual sea level time series which is mentioned in the previous thread. I’ve plotted the data so that it can be shown.


Figure 1

“Last year I constructed a relative global sea level rise series between 1900 and 2010 from scratch using 328 unadjusted tide gauge records from the PSMSL data base. When I compared it with the unadjusted global ICOADS SST series I got a remarkably close match (R^2 = 0.92 for annual means, with a 1C rise in SST corresponding to a 100mm rise in relative SLR).”

Thread here.

Roger deleted the WW2 SST data as known unreliable. Data has been normalised and is plotted, Figure 1, I confirm r2=0.92

The question of whether SST leads sea level or not needs answering given I have suggested in a previous post that going on satellite measured tropsphere temperature and sea level that it does.

Given the data here I completely agree it doesn’t.


Figure 2

Figure 2 shows the spectra and the great similarity long term but the faster detail is bad.

Using first differences does produce a peak at 15 years but looking at the data suggests this is nothing important.

I’ll leave comment at that for the moment but can edit later.


  1. Roger Andrews says:

    I think the important issue here isn’t whether SST leads or lags sea level rise, but why we see such a close correlation between these two independent and unadjusted series. It’s hard to see how it could be an artifact of the data, and if it isn’t we have to consider the possibility that SST (or more likely ocean heat content) is the dominant control on SLR. And as noted in previous posts, SST for most of the 20th century seems to have been controlled by changes in solar activity, so maybe SLR was too.

  2. laterite says:

    Hi Tallbloke. I have looked at this lag too. The lags seem to be explained by looking at at the relationship between the quantities compared. Here, SST and sea level are measurements of effectively the same thing, so no lag.

    By comparison, there is a 3month, or a 2.75 year shift between insolation and temperature because the comparison is between a forcing and its integration (or accumulation) as temperature. When you integrate a sine wave you get a -cosine – a shift of one quarter of the period with is one year or 11 years depending on the scale and where its measured. See

  3. Ray Tomes says:

    Tim, For me this is much more convincing than the previous plot. It shows similarities in the shorter and medium term fluctuations that suggest real causality. In particular the 50-60 year cycle can be seen if the trend is removed (highs ~1900, ~1950, ~2010 and lows ~1920, ~1980 relative to trend).

  4. P.G. Sharrow says:

    It would appear to me that Sea Surface Temperature and Sea Surface Level are both measurements of the same thing, energy levels of the thermal active part of the oceans. Warmed waters expand but weigh the same and so have a higher “sea level” in the warmed area. Air temperature represents energy moving from the land and sea towards the cold of space. pg

  5. tchannon says:

    That seems reasonable PGS although my brain kicks in.

    Gauge data will differ in scale from far open ocean.

    If the ocean temperature profile can be deduced, do the figures for expansion equate?

    Lot of ifs and I think something like that has been published. There again we must not forget Morner and so on.

    As a reminder. Can’t find the doc I was searching for.

    Click to access MornerEtAl2004.pdf

    Seems he is still active, not seen these PDF before.

    Click to access NilsAxelMornerinterview.pdf

    Click to access Morner.pdf

  6. P.G. Sharrow says:

    Very right Tim. Ocean surface levels vary, due to many factors. Gravity variance can be as much as 600 feet lower over deep trenches relative to shore line levels. Small changes in local gravity as well as magnetic fields can swamp any millimeter measurements and we know the earths magnetic field is undergoing major changes. Continental movements change the volume size of the ocean basins and earth movements have increased of late. A lot of moving parts to consider. 😦

    I don’t think there is a simple answer to ” sea level” changes. pg

  7. Roger Andrews says:

    According to conventional wisdom only about half of the sea level rise since 1910 was caused by thermal expansion of sea water. The rest was caused by ice melt, mostly from retreating glaciers. However, it’s not unreasonable to suppose that the rate of glacier retreat may also be related to solar activity, particularly if the ice is liberally sprinkled with soot.

    A more fundamental question is how much sea level has actually risen over the last 100 years. My series shows about 100mm of SLR since 1910, much less than the 170-180mm shown by the “official” series. A recent version of this series (note the propaganda, incidentally) can be viewed at

    [current CSIRO imaged added by Tim]

    The 70-80mm difference is caused by the fact that I measure sea level rise relative to the coastline while the official series measures – or attempts to measure – absolute SLR, or how much sea level has risen relative to the center of the earth. The idea here is to come up with an estimate of how much mass has been added to the oceans, which is a scientifically a valid approach, but it suffers from two defects:

    1. To measure absolute sea level rise the tide gauge records must be corrected for vertical land movement. VLM corrections are obtained from crustal deformation models that simulate the effects of glacial unloading since the end of the last Ice Age, and these models indeed do a good job in areas in and around the old NH ice sheets. However, they do a lousy job everywhere else because they don’t allow for tectonic uplift, sediment compaction, gravity-induced subsidence etc., and as a result the corrections applied in many areas are just flat wrong.

    2. We don’t care how much sea levels are rising relative to the center of the earth. What we care about is how much they are rising relative to the coastlines along which people live. (And local conditions, not absolute sea level rise, dictate the level of risk here. The residents of Skagway, Alaska have nothing to worry about because tectonic uplift is causing sea levels there to fall by about 20mm/year. On the other hand, Bangkok is subsiding by about 20mm/year because of excessive groundwater pumping, and if this rate of subsidence continues the city will be largely inundated by 2050.)

    So how much has sea level risen over the last 100 years? I think 100mm is the best estimate we are going to get, and if this estimate is correct we needn’t worry too much about what rising sea levels might do to our “grandkidz” over the next 100 years.

  8. P.G. Sharrow says:

    Maybe the answer is……………………. It is not enough to get worried about, or, there is nothing that can or should be done about it any way! 🙂 pg

  9. Just rebuild back from the beach as the houses give way to storm damage, most of the population in Asia that live along the shore line are in boats anyway.

  10. Roger Andrews says:

    If sea level tracks SST, and if SST is controlled by solar activity, then it’s likely that sea levels will fall rather than rise in the future. See

  11. Doug Proctor says:

    Trenberth et al consider that only a portion of the SLR is attributable to SST rise, perhaps, he suggested somewhere, 40%. If SLR were only a portion, with land-based ice melt being the dominant part, would not the SLR have a faster rise than temperatures?

    I realise that this assumes that there is a temperature lag between new ice-water entering the system and stabilization of the oceanic temperatures. Which seems likely, as the new water entering the system is colder than the average. In fact, one of you smart math guys should be able to model it – you know the amount of water that is purported to be entering the system, you know the temperature of that water and the world SST, and therefore the proportion impact the new water should have.

    Scenarios are easiest: 100% as 2 or 3C water vs average SST, down to 0%. At the least it should tell you the times for equilibrium. And a calculation of the warming energy will give you a W/m2 figure to compare with the 0.85 W/m2 of “missing” heat.

    Complicated thinking for ten in the morning.

  12. steven mosher says:

    see steveFs post at Lucia’s. Also see GRACE

  13. tchannon says:

    The only rational match I can find is something confused in it’s identity, rankexploits/The Blackboard/ some seem to call it Lucia’s. Briefly seen the site in the past.

    That site uses a Google ECMA script based tool for local search and it is broken.

    Assuming this is the site Steven meant can someone else please try to find a post by steveF there and post a direct link if the item looks interesting.

  14. Tenuc says:

    The link could be this one perhaps…

    My take is that the post seems to confirm the issues with the data and how it is obtained, also that sea level rises are likely to be small over the next 100y based on assumed temperature increases.

  15. […] Roger Andrews: Sea level reconstruction from tide gauge data and a match with SST […]

  16. Dallas says:

    I have seen you on the other blogs much. Here is a little something you may find interesting;

    Where Did Climate Science Go Wrong?

    Climate Science, specifically the Theory of Global Warming, is based on assumptions. If those assumptions are valid, the supporting work will be valid, if not?

    The first assumption is that our world without greenhouse gases would be 33 degrees cooler. That assumption is based on the assumption that the Earth would reflect just as much sunlight as it does today. A valid assumption provided the surface absorbed all the sunlight or if the top of the atmosphere temperature is used consistently to determine warming. If the top of the atmosphere is used, then the assumption can be made that the atmosphere was so low, that the surface and atmosphere could be at the same temperature.

    That’s what, three assumptions supporting one main assumption?
    There is more to heat transfer than photons. If the Earth had nitrogen and oxygen in the same quantities as today, conduction and convection assure that the Earth would have an atmosphere even without greenhouse gases. With Oxygen in the atmosphere and ultraviolet energy emitted by the sun, the Earth would be assured of having a Tropopause, the temperature inversion in the lower atmosphere. That would mean the surface would be at least 2 degrees cooler than the top of the atmosphere if the Troposphere is assumed to be the top.

    The Earth also has a wealth of water, water, temperature, equatorial sunlight and atmosphere ensure there would be latent cooling as water evaporated and is carried aloft by convection. That could produce nearly 10 degrees cooler temperatures at the surface than at the top of the atmosphere. That could result in a 30% error in the Greenhouse Effect portion of the overall atmospheric effect.

    Assumption two, the greenhouse effect will uniformly warm the surface with most impact at the poles and upper troposphere. Unfortunately, this is also an incorrect assumption.

    In the tropics, additional Carbon Dioxide has minimal effect on the surface due to the high concentration of water vapor. The Carbon and water vapor radiant effects are more strongly felt above this water vapor where its downward impact warms the water and clouds below it. This increases upper level convection promoting cooling which partially offsets the warming potential of the Greenhouse Effect. In the Northern polar region, were the air is less saturated with water vapor, the additional Carbon Dioxide enhances warming as predicted. In the Southern polar region where temperatures are much lower, water vapor does not increased significantly with the radiant forcing of additional CO2 reducing the impact of the Greenhouse Effect.

    Trying to match observations with these faulty assumptions has led Climate Science down a path of attempting to “Force” data to match expectations instead of matching theory to observations. Not sound scientific practice.

    Is the Earth warming due to Anthropgenic Greenhouse Gases? Yes, but it will also cool because of them. The additional greenhouse gases amplify natural variability in some regions and not others. More, proper, research needs to be done.