July 26 2011

Fans of man-made global warming frequently tell us seas are rising, but somehow forget to mention the rise started 200 years ago, long before our coal-fired electricity plants cranked up, and long before anyone had an electric shaver, or a 6 cylinder fossil-fuel-spewing engine. Something else was driving that warming trend.

Here is the data from tide gauges going back 300 years from a paper by Jevrejeva et al 2008 [after the jump].

[Graph based on data from Jevrejura et al located at this site PMSML]

This graph was calculated from 1023 tide gauge records [Jevrejeva et al., 2006] going back to 1850.The 2008 study extended the record further using three of the longest (though discontinuous) tide gauge records available: Amsterdam, since 1700 [Van Veen, 1945], Liverpool, since 1768 [Woodworth, 1999] and Stockholm, since 1774 [Ekman, 1988]. Obviously since there are only three old records, the error bars are a riot.

The Jevrejeva paper is also useful for portraying the 60 year rolling cycle. The regular ups and downs are obvious when the rate of change is plotted (see below).

But wait… there must be a tipping point?

While the graph itself seems like it was made for skeptics (how can anyone say that linear warming trend was started by CO2?) some back-seat critics will say that Jevrejeva et al claim that “it will be worse than the IPCC thinks” – which they do say. But that’s the name of the game isn’t it, to find “acceleration”. Are sea levels are rising faster because of CO2?
Here’s where Jevrejeva et al make the “it’s worse than we thought” statement. Look closely at the reasoning:

“We show that sea level rose by 28 cm during 1700 – 2000; simple extrapolation leads to a 34 cm rise between 1990 and 2090. The lowest temperature rise (1.8°C) IPCC [Meehl et al., 2007] use is for the B1 scenario, which is 3 times larger than the increase in temperature observed during the 20th century. The IPCC sea level projection for the B1 scenario is 0.18– 0.38 m. Our simple extrapolation gives 0.34 m. The mean sea level rise for B1, B2 and A1T is below our estimate. However, oceanic thermal inertia and rising Greenland melt rates imply that even if projected temperatures rise more slowly than the IPCC scenarios suggest, sea level will very likely rise faster than the IPCC projections [Meehl et al., 2007].”

Have I got this right, it appears they predict that:

a/ Based on the acceleration in the last 300 years, they expect seas to rise by 34 cm this century anyway (without man-made global warming).

b/ That the IPCC reckons it will all get much warmer (frying-hot) on top of that trend, thanks to CO2.

If so, this would be double counting, and they can’t have it both ways. The IPCC assumes that all the warming since 1780 is man-made and then extrapolates that wildly. These authors (between the lines) say the sea level rise (a proxy for warming) was natural, and then extrapolate that trend and add it to the IPCC extrapolation. Both extrapolations are based on the same trend — with opposing assumptions, and added together. No No No.

If the warming so far was natural, then CO2 has little effect, so there would be nothing much to add on top of their extrapolation.

Finding curves in short lines

Part of the problem with calculating acceleration with this data is the 60 year cycle of rises and falls. Basically, if we had a nice long record we could figure out the current cycle and see whether it was accelerating. But given that the cycle is 60 years long; we only have good records going back 160 years, and sparse records going back another 150, we really don’t have much at all to work with. Worse, it’s a multivariate system of which we don’t even know all the factors.
Hence I’ve drawn a straight line trend through the top graph. Jevrejura used a polynomial fit to calculate a small acceleration. When we have such short records, who can say which fit is the winner? Wait 100 years and find out.
Since sea levels rose 19cm in the last century and the trend is linear, so we don’t need an intergovernmental panel, $200,000 grant and 5 year study to project a rise for the 21st Century of… 19cm, more or less.
—————————————————–

REFERENCES

Jevrejeva, S., A. Grinsted, J. C. Moore, and S. Holgate (2006), Nonlinear trends and multiyear cycles in sea level records, J. Geophys. Res., 111,

Jevrejeva, S., J. C. Moore, A. Grinsted, and P. L. Woodworth (2008), Recent global sea level acceleration started over 200 years ago?, Geophys. Res. Lett., 35, L08715, doi:10.1029/2008GL033611. [PDF]

—————————————————–

Additional thoughts on the Jevrejeva paper from Lionell Griffith

The one thing that pops out the most is the typical trick of picking convenient dates as starting and ending points for their so called curve fits and using an arbitrary order for the curve. Then they extrapolate that curve beyond all rationality. They would be better off flipping a coin and guessing. At least that way they have a finite chance to be right. The way it is, they are not connected enough to reality to be wrong.

You can fit any order of curve to any set of data as long as you have more data points than orders of your curve. All that does is give a more or less accurate way to interpolate between actual data points used in the curve fit. You can even get high values of goodness of fit but it is all quite meaningless outside of the specific data set. Statistical significance is not always significant in terms of real world validity. Without grasping ALL of the meta data, you can draw no conclusions about reality other than that is what the calculations applied to the numbers produced.

Extrapolation from a random (non causal) curve fit is 100% a dangerous thing to rely on. The error bars explode the further away from the end points you are. Even the ability to estimate the error bars decays to nearly zero at some short distance from the end points. This is a process that should NEVER be relied upon to make judgments about the future PERIOD! Only if you have a causal bases for your fitted curve does extrapolation have any reliability. Even then, the reliability is heavily dependent upon the quality of the input data AND the degree that all causes are included in the curve you are fitting. This alone should be sufficient to discredit anything they conclude. Their statistics are no more valid than those of the Hockey Stick Mann. However, I will give them one point for disclosing as much detail as they did.

Now taking the plot below at face value. The first thing I see is the presentation of two dissimilar data sets (1700 to ca 1860 and ca 1860 to 2000). They may be incommensurate and quite inappropriate to use in ANY kind of curve fitting over the entire time series. From the data itself you cannot determine the cause of the discontinuity at ca 1860. You must have a massive amount of meta data that gives the full context of each time series. Then and only then do you have even a remote chance of blending them into a coherent pattern.

I suggest two things go a long way to explain the discontinuity. The first is that ca 1860 was about the time the little ice age started to resolve itself. The second is that the data set was likely differently instrumented and with greater attention to consistency, frequency, and quality control over the process.

It is quite likely that there is a lot of selection bias hidden behind the graph. There is no way to prove it one way or the other. Check into the exacting work of determining the mass of the electron. The pattern of the results show some interesting things going on even with honest hard working scientists. This even when there was no government financing to stimulate a given end result.

I also find that the second data set shows NO visible response to CO2. It is simply a continuation of whatever the cause of the resolution of the little ice age. There is no visually significant change in the trend line between ca 1860 to ca 1945 and ca 1945 to 2000. You could select starting and ending points such that there were two different trends. This too is a source of selection bias that is invalid. There must be a reason independent of the data itself that is used to choose the starting and ending points.

The null hypothesis (natural process is the cause) is sustained and ANY man produced CO2 causality remains undetectable. You don’t need 100,000 words to say it. You need only a legitimately produced graph and a few supporting words.

The fundamental principle here is one cannot properly go beyond the evidence and call it science. It becomes speculation at best and demagoguery or fraud at worst.