Roger Andrews: The Sea Level House of Cards: Part 1 – The Science

Posted: December 29, 2012 by tallbloke in Analysis, Natural Variation, Ocean dynamics

Guest post by Roger Andrews on the IPCC’s sea level rise attribution:

by Roger Andrews : Dec 2012

According to the recently-released draft of the IPCC AR5 global sea levels rose by about 0.17m in the 20th century and are projected to rise by 0.29 to 0.96m more in the 21st. This projection is based on the assumption that the warming generated by man-made greenhouse gases causes ice to melt and sea water to expand, thereby raising sea levels. Here we will review the supporting data the IPCC presents in the AR5 draft to see whether this assumption stands up to scientific scrutiny.

If the assumption is correct then there has to be a time-dependent relationship between sea level rise and AGW, so the first question to be addressed is; when did man-made greenhouse gases begin to have a significant impact, or in other words, when did AGW start? According to the following statements in the AR5 draft Summary for Policymakers it didn’t start with the Industrial Revolution. In fact it didn’t start until quite recently (the boldface is mine):

It is very likely that more than half of the ocean warming observed since the 1970s is caused by external forcing, mainly due to a combination of both anthropogenic forcing and volcanic eruptions (see Figure SPM.4). It is extremely likely that this warming has resulted in global mean sea level rise due to thermal expansion during this period. 

It is extremely likely that human activities have caused more than half of the observed increase in global average surface temperature since the 1950s. There is high confidence that this has caused large-scale changes in … sea level in the second half of the 20th century. 

Climate model simulations that includeonly natural forcings (volcanic eruptions and solar variations) can explain a substantial part of the preindustrial inter-decadal temperature variability since 1400 but fail to explain more recent warming since 1950 (see Figure SPM.4).

Two of these statements cite Figure SPM.4, so let’s see what this Figure shows. The upper graph in Figure 1 below reproduces the comparison of observed and modeled global land surface temperatures shown in Figure SPM.4 and the lower graph plots the difference between temperatures modeled using natural plus anthropogenic forcings and natural forcings only, revealing the anthropogenic contribution. According to the lower graph roughly 80% of the total anthropogenic land surface warming since 1860 occurred after 1960.


FIGURE 1 (From AR5 Figure SPM.4)


Figure 2 reproduces the Figure SPM.4 model-observed comparison for ocean heat content. It doesn’t go back far enough to provide a complete picture, but the data that are available indicate that all of the anthropogenic increase in ocean heat content occurred after 1970.


FIGURE 2 (From AR5 Figure SPM.4)



Based on these results we can reasonably pick 1970 as the year in which man-made emissions first began to have a significant warming impact, or in other words the year in which AGW began.

Now that we’ve established this we can look at some long-term observational data. First glacier ice loss, which is one of the two major contributors to sea level rise and which is assumed to be controlled by air temperature. AR5 Figure 4.12 plots various estimates of cumulative global glacier mass loss since 1800, and Figure 3 is a cleaned-up version of it:


FIGURE 3 (From AR5 Figure 4.12)

If AGW is the only thing causing glaciers to retreat we should see a hockey stick shape that breaks upwards around 1970, and if an AGW component is superimposed on a natural component we should see an increase in the rate of glacier mass loss after 1970. But we see neither. The data show glacier retreat beginning after the Little Ice Age, which is generally agreed to have ended around 1850, picking up speed in the mid-1900s and if anything tending to slow down rather than speed up after the onset of AGW. There’s nothing here to suggest that the current phase of glacier retreat is anything other than a natural event that started well over a hundred years before the onset of AGW and which continued unaffected by man-made greenhouse gas emissions after AGW began. And in Section of the AR5 draft the IPCC admits that at least the first 50 years – and depending how we interpret the wording maybe the first 120 years – of glacier retreat indeed probably was a natural event:

Overall, the combined records suggest that the net decline of global glacier volume began in the 19th century before significant anthropogenic radiative forcing had started, and was probably the result of natural climatic variability

The other major contributor to sea level rise is thermal expansion of sea water, which is governed by ocean heat content. Unfortunately there is OHC record before 1954 (and arguably no reliable OHC record before the ARGO data became available in 2003) so we will move on directly to sea levels.

Figure 4 shows four reconstructions of mean global sea levels based on tide gauge records, two of which (Jevrejeva et al. and Church & White) are presented in the AR5 draft. Like the ice mass loss reconstructions they show sea levels rising at a roughly constant rate since the end of the Little Ice Age and no visible acceleration in the rate of rise after the onset of AGW in 1970:



The level of correlation between sea level rise and AGW can also be evaluated statistically by comparing annual changes in sea level against anthropogenic forcings, and Figure 4 is an XY plot of annual sea level change since 1880 using the Jeverejeva data) against the GISS annual anthropogenic forcing estimates, which are substantially the same as those used in the IPCC’s climate models. The correlation coefficient (R2 = 0.000) is about as low as it’s possible to get.



 So how does the IPCC conclude that man-made greenhouse gases are causing sea levels to rise? It cites two basic lines of evidence. First is the postulated acceleration in the rate of rise during the 20th century – presumably related to AGW, although the IPCC doesn’t explicitly say so – cited  in AR5 Section 13.2.2:

(T)he rate of sea level rise has increased since the late 19th century and has continued during the 20th century  …. the estimated acceleration in GMSL since 1900 ranges from0.000 mm yr–2  in the Ray and Douglas (2011) record, 0.013 mm yr–2 in the Jevrejeva et al. (2008) record, and 0.012 mm yr–2  in  the Church and White (2011) record.

Two positive estimates and one zero estimate hardly make a compelling case for the existence of an acceleration component, and if the acceleration began in 1900 it predated AGW by 70 years and can’t have been caused by it anyway. There are also doubts as to whether the 20th century sea level record even contains an acceleration component. The sea level data on which Church and White based their acceleration estimates, which are presented in Figure 5 of their 2011 paper (here) and reproduced in Figure 6 below, show no change in the rate of sea level rise since 1930:



The second line of evidence comes from “semi-empirical” climate models which “do not explicitly simulate the underlying processes, (but) assume that sea level rise is caused primarily by changes in global ice volume and global ocean heat content in response to changes in global temperature or radiative forcing” (AR5 Section 13.1.5). Translated this means that the models use observational data as well as forcings to simulate the ice melt and ocean heat content contributions to sea level rise, which should make it a lot easier to match observations. But as shown in Figure 7 (a cleaned-up version of AR5 Figure 13.4 e) they still can’t hindcast 20th century sea levels:



But the fact that the models can’t hindcast 20th century sea levels doesn’t stop the IPCC using them to project 21st century sea levels, and Figure 7 shows the results:


With that our review of the IPCC’s basic data ends because there are no more basic data to review. What’s the conclusion? That there is no evidence to support the assumption that man-made greenhouse gas emissions are causing sea levels to rise. Zero, zip, zilch, nada. The science behind human-induced sea level rise isn’t just flawed, it’s nonexistent.

And now it gets worse, because the science used to develop the 20th century sea level rise estimates is flawed too. Details in Part II.

  1. Arfur Bryant says:

    There is one major logical argument to any suggestion of an AGW signal proposed by anyone…

    If anyone states that the AGW signal started after 1945, then they have to explain why the global temperature warming of 1910-1945 is of almost exactly the same size as the warming period of 1975-1998 (appx 0.65C).

    If anyone states that the AGW signal started before 1900 (eg 1850), then they have to explain why there have been observed cooling periods in: 1878-1910, 1945-1975 and a flattening from 1998 to today.


    For sea level, there simply is NO acceleration.



    The data shown above pre-1970 from tide gauges are useful in showing the big picture.

    As for OHC, this data is as good as you’ll get:


    (Note the graphs will automatically scroll through…)
    For interest, click on the left-side links for the ‘Heat Content Basin Time Fields’ and note that, at the 700m level, only the Indian Ocean has shown any warming at all in the last six years.

    The IPCC hypesters are living in a world of their own imagination.

  2. Paul Vaughan says:

    8000 years of AMO?
    Tamino writes: “Nature shows us tentative results more often than definitive ones, and if we don’t heed such revelations, then we won’t learn as much about her workings as we could.”

    I agree that we are wise to respect & appreciate the power & beauty of nature.

    Tamino recently looked at Sea Level Rates:

    New York – The Battery:


    In the related discussion Aaron Lewis (September 29, 2012 at 7:16 pm) made some incisive comments on nonlinearity that are sure to irk many:

    Nonlinearity promises to be a provocative topic. Greg Goodman (a.k.a. P. Solar) might appreciate an exchange in which Vaughan Pratt appears to have convinced Tamino of the applicability of Heisenberg uncertainty to the disentangling of cycles & trends.
    (Greg might dislike Pratt’s advocacy of boxcars. Tamino likes LOESS. Greg likes Gaussian.)

    Greg (if you’re around), you may realize by visual inspection that the green multidecadal sea level wave I show ( ) is a composite. I explored a number of sea level metrics. Straight sea level rate gives a phase match with SST, but not an amplitude match. To get an amplitude match the multidecadal component has to be powered up. There are a number of ways to do this. I chose second order central differencing since its intuitively simple & easy to implement. Wavelets can just as easily do the job.

    (There’s more to say — a bit too far OT, so this is enough for now…)

  3. Doug Proctor says:

    From 1900 to 2010, the GMSL rose about 175 mm in Figures 4, 6 and 7. In Figure 8, the GMSL rose about 200 mm over the same interval. Am I reading the graphs wrong?

    Perhaps the GIA is 50mm. Does this mean that when the IPCC/Gore says the GMSL will rise 1000mm by 2100, they actually mean INCLUDING 333mm of GIA?

    Numbers float like continents of the mantle.

    BTW: the continents are rising from GIA. The Law of Inertia or Momentum says that the continent, once in motion, will tend to stay in motion until acted upon by a restraining force. The continents will “bob”, rising above their position of equilibrium and then sinking below their new position of equilibrium until frictional forces reduce bouyancy forces below a threshold minimum.

    As a geologist I am always dealing with cycles of shallowing and deepening of the continental oceans of North America, Devonian to Tertiary. I wonder if, knowing density differences of crust and mantle, ice masses and/or initial rates of GIA from the Hudson Bay area, we could figure out how much “bobbing” we are likely to get ….

  4. Paul Vaughan says:

    @Doug Proctor (December 30, 2012 at 3:35 am)

    The spatial pattern of GIA isn’t as simple as “continents are rising”.

  5. Ed Andrews says:

    Looking at the above historical data, which doesn’t seem to have changed recently, it is clear that sea levels really aren’t rising much at all in the historic context…. We are crying about them coming up 8 inches in 200 years; during several melting events of the last 20,000 years they rose as much as 80-100 FEET in intervals that long (see Meltwater Pulse 1A)

    So what I have to wonder is when they are going to give up on sea level rise due to AGW as a world-changing emergency; we should be looking at the natural events that contribute here….

  6. Paul Vaughan says:

    Further to Paul Vaughan (December 30, 2012 at 3:53 am) addressing Doug Proctor (December 30, 2012 at 3:35 am) …

    PGR = post-glacial rebound (another term for GIA = glacial isostatic adjustment)

    Map of theoretical PGR:

    From WUWT:

    Bill Illis (November 30, 2012 at 3:30 am) wrote:
    “We should also note that the Land is rising at an average of 0.3 mms/year (due to rebound from the ice ages). That means the actual sea level rise over the last 20 years relative to the Land from this glacial melt is less than one-quarter of an inch.”

    I asked Bill:
    “Bill, what is the conventional wisdom on the spatiotemporal variability of this estimate?”

    His response:
    Bill Illis (November 30, 2012 at 3:55 pm) wrote:
    “Peltier is the godfather on this.

    It varies a lot by location, large positve numbers in those areas which were depressed by the ice age glaciers, and smaller negative numbers in those areas which were uplifted (by near-by depressed areas and by the loss of ocean weight on near-by continental shelf areas as sea levels were lower in the ice ages).

    Temporally, there is little change over the last 250 years or 250 years into the future.”

    (caution: the comment on temporal variability is based on very crude modeling estimates)

  7. oldbrew says:

    ‘According to the lower graph roughly 80% of the total anthropogenic land surface warming since 1860 occurred after 1960’

    What formula was used for attribution?

  8. oldbrew says:

    A recent GWPF article notes:

    ‘a new report by the US National Oceanic and Atmospheric Administration shows that global sea levels rose by only 1.1 – 1.3 mm/year from 2005 – 2012, which is less than half of the rate claimed by the IPCC. This is less than 5 inches per century.’

  9. Joe's World {Progressive Evolution} says:


    Two points.
    Not a single mention or consideration of water loss to space.

    The second is my research in past water.
    Pressure of water from 4.5 billions years and the water height would indicate massive weight exerted.
    The glacial melt?
    Considering water changes directions at the 48 degree latitude, so melt is going in two directions in a warming period from an ice age.
    Erosion of rock is worldwide even underwater. But conventional wisdom THINKS it was from glaciation(even though ice ages have not past a certain area of latitude).
    Mountains…taking water away would generated higher and higher altitude.

  10. Roger Andrews says:

    Doug Proctor: “From 1900 to 2010, the GMSL rose about 175 mm in Figures 4, 6 and 7. In Figure 8, the GMSL rose about 200 mm over the same interval. Am I reading the graphs wrong?” The Figure 8 data are the same as the Figure 4 data, i.e. 150-200mm of SLR depending on whose reconstruction you use.

    Paul Vaughan: “The spatial pattern of GIA isn’t as simple as “continents are rising””. Correct. GIA removes water from continents and adds it to the oceans, thereby inducing continents to rise and ocean basins to sink, but most tide gauges are located in the gradient area along shorelines, which can make it difficult to estimate an exact VLM component. More importantly, GIA doesn’t allow for tectonic activity, volcanism, sediment compaction etc. I plan to get into this and some of the other points you bring up in Part 2.

    Oldbrew: “What formula was used for attribution?” It was, as always: climate model simulations = correct result. 😉

    On the road for the next few hours. Back later.

  11. Joe's World {Progressive Evolution} says:


    You are doing the EXACT same as climate scientists are and are cherry picking for a minor period of the planets history.
    I know it is the only DATA you have but…absolutely NO consideration is given to our planets history.
    Look at Mars…

  12. oldbrew says:

    ‘Look at Mars’

    Sounds like the slippery slope fallacy…

  13. Brian H says:

    The warmist strategy: pile one contra-factual assertion on top of another. When one is disproven, ignore and utter another one. Repeat. Repeat. ….

  14. Greg Goodman says:
    (Greg might dislike Pratt’s advocacy of boxcars. Tamino likes LOESS. Greg likes Gaussian.)

    As Pratt says they are brilliant for removing a perfectly sine wave from a signal as long as you don’t mind screwing up what remains.

    Since we rarely have a perfect sine to be removed and even if we did we presumably ARE interesting in what remains and don’t what it screwed up and distorted, runny means seem of little use.

    The one case where they can be of use is if you stack them with suitably chosen frequencies as someone did for him in designing the F3 filter for his AGU poster fiasco.

    In fact that was about the only good thing about that presentation and even then he botched the implementation and ruined what was otherwise a well constructed filter.

    Though I do generally favour gaussian, the tripple runny mean he _meant_ to use would probalby have been better than a gaussian for removing a specific frequency. But there again Hale cycle is neither fixed nor sinusoidal so any advantage his filter may have had was not applicable anyway.

  15. Greg Goodman says:

    Brian H says:
    December 30, 2012 at 7:12 pm

    The warmist strategy: pile one contra-factual assertion on top of another. When one is disproven, ignore and utter another one. Repeat. Repeat. ….

    OH, I thought Pratt was just being a …. , it’s actually supposed to be a strategy. Right.

  16. Joe's World {Progressive Evolution} says:


    Water displacement is an extremely tricky slope to find ALL the contributors and measure the exact amount being displaced.
    Freezing water expands which also changes the weight displacement.
    Space debris into the oceans.
    Biggie is volcanic activity and the added materials of gases and solids.
    Every pebble, peer or ship added displaces water in oceans.
    Pressure changes of the atmosphere(which we measure the atmosphere by water pressure).
    The moon’s orbit.

    Water study is absolutely fascinating as well as very deceptive.
    How much evaporation/precipitation changes the levels as well draining from landmasses or is being retained in certain areas in the way of snow and ice.

    A glass of water:

    Every mm has a different pressure also a slightly different velocity when you measure the circumference of the planet around back to it.
    Every mm is of a slightly different distance.

  17. Paul Vaughan says:

    Greg, can you concisely describe Pratt’s “F3” filter and the rationale behind the design?

  18. Paul Vaughan says:

    “The existence of these systematic amphidromes determines that the phase pattern of SLP is basically rotational rather than zonal. Therefore, the latitudinally banded description […] reveals only an incomplete and perhaps distorted picture of the global SLP phase variation”

    Teaser from an important new paper:

    Chen, G.; Qian, C.; Zhang, C. (2012). New insights into annual and semiannual cycles of sea level pressure. Monthly Weather Review 140, 1347-1355.

    Click to access 2012ChenZhang_MWR.pdf

    Refreshing exploration.

    With adequate budget (for computing resources, staff, & software development) this study could be taken several layers further using multi-extent quaternion wavelets, probably cracking problems that are currently considered intractable.

  19. Roger Andrews says:

    Re the earlier comments on OHC, and to enlarge on my statement that we probably don’t have any good ocean heat content data before ARGO, here are some OHC reconstructions:

    The first two are longer-term reconstructions from NODC (National Oceanographic Data Center) and SODA (Simple Ocean Data Reanalysis) from KNMI Climate Explorer. Third is Figure 2 from Lyman et al. 2101 (here) showing the results of applying different corrections to the same XBT data after 1993.

  20. Wayne Job says:

    Way back when Brittannia ruled the waves the Admiralty took on a task as big as a moon landing.
    They found and engraved rocks they termed dry rocks at low tide all over the world. Oddly the rocks around Australia are still dry at low tide. This can only mean one of two things, either the ocean has not risen much in the last two centuries or the gravity of the moon is getting stronger.
    I would tend to go with the ocean is not rising.

    The British Admiralty spent the wealth of Britain to mark these rocks, they are the corner stone of any movement in ocean height, if they are ignored it is like saying the moon landings did not happen. Compared to the money spent on other AGW crap a survey of all the rocks would give a definitive answer. No proxies or guess work needed.

  21. tchannon says:

    Any idea where the documents are kept?

  22. Roger Andrews says:

    I don’t know where the documents are kept, but John Daly touched on the question here:

    As to how much the ocean has risen, according to the IPCC it’s about 180mm since 1900 and according to me it’s about half that. Reasons for the difference to be discussed in part II if anyone is interested.

  23. tchannon says:

    Looks a major unearthing task. I wonder though how many have been seen by sailors in recent years but never thought of writing it up?

  24. […] Figure 3: Sea level over the last 300 years. Source: Roger Andrews. […]

  25. Chaeremon says:

    @tchannon: you may want to find vocabulary used by navigators: imperial, chart datum, Admiralty charts, like here. All these charts should be accessible to navigators of the UK.

    Navigators have all this information on charts for use before a voyage begins, i.e. while still in the port of registry. This information And Any Discrepancy was constantly updated, with actual observations, upon return, reported by licensed navigators to the authorities.