I regularly get comments claiming that ocean cycles are the cause of global warming. They couldn’t be more wrong.
The oscillation du jour is AMO, the Atlantic Multidecadal Oscillation. In part this is because of a post by Bob Tisdale making a number of ridiculous claims. As the latest such comment I received says,
He concludes, that multiple regression isn’t a way removing the signal. The signal is still clearly present like in Fig. 2 of this post. If you would just plot MEI on on top of that you would see it doesnt do a very good job. I can see it just by looking at it. The “eye balling” method seems to be a much more robust method in removing the signal, as Bob has shown us.
His analysis is much more comprehensive and takes AMO in account, for example. Only roughly 27% of the trend remains in GISS 60N-60S LOTI data after all of the artifacts which are propably due to inner heat transfer processes, have been removed.
Yes, folks. Bob Tisdale actually believes that “eyeballing” the correct lag and scale factor for fitting time series is better than multiple regression. The WUWT crowd laps it up. It’s near impossible to have an intelligent conversation with people who believe such things.
I won’t dwell on the folly of substituting “eyeballing” for analysis. But by all means let’s take a look at the AMO. Wikipedia describes it thus:
The AMO signal is usually defined from the patterns of SST variability in the North Atlantic once any linear trend has been removed. This detrending is intended to remove the influence of greenhouse gas-induced global warming from the analysis. However, if the global warming signal is significantly non-linear in time (i.e. not just a smooth increase), variations in the forced signal will leak into the AMO definition. Consequently, correlations with the AMO index may alias effects of global warming.
Could it be that the global warming signal is significantly nonlinear? Here’s global temperature from GISS:
Here’s a linear fit:
It sure seems that the global warming signal is nonlinear. In fact, here’s a much better characterization of that signal, from a lowess smooth:
Variations in the forced signal do leak into the AMO definition. Correlations with the AMO index do alias effects of global warming.
The way to compute AMO, according to the ERSL AMO page, is as follows:
So the AMO is actually detrended (but only linearly detrended) North Atlantic Sea Surface Temperature (N.Atl SST). We won’t bother with any smoothing.
Conveniently, the ERSL AMO page gives a direct link to the SST data averaged over the N. Atlantic. It looks like this:
Evidently there’s another step we have to perform: we need to translate N.Atl SST to N.Atl SST anomaly. That looks like this:
I computed these anomalies using the entire data set as a “baseline” period. We can linearly detrend this data, then compare it to the given data for the AMO itself:
You can’t really see the separate curves, because they’re so close. But we can look at the difference between the two:
If you look closely (click on the graph for a larger, clearer view) you can see that the only difference is a residual annual cycle — and a very small one at that (it never goes beyond plus or minus 0.035 deg.C). Either they’ve used a different baseline period, or they computed anomalies in some other way (sometimes anomalies are computed by fitting a Fourier series, whereas I simply computed monthly averages and defined anomaly as the deviation from the monthly average). But as I say, the difference is tiny and doesn’t amount to a hill of beans. No doubt about it, the AMO is linearly detrended North Atlantic Sea Surface Temperature.
What could be the cause of changes in N.Atl SST anomaly? Let’s compare the N.Atl SST anomaly to GISS temperature:
Obviously they’re strongly correlated. Bob Tisdale (and others) simply can’t wrap their brains around the fact that global warming is the cause, not the effect, of much of the changes in N.Atl SST anomaly. Therefore global warming is the cause, not the effect, of much of the variation in the AMO.
If we’re really interested in how the north Atlantic is oscillating, apart from the global warming forcing, we can at least get a rough idea by simply taking the difference between N.Atl SST anomaly and GISS temperature:
Of course there are better ways to do this, but it still illustrates the point: that when you look at AMO correctly, the possibility that it’s much of the cause of global warming vanishes. It’s no more sensible than the notion that “eyeballing” is a better choice than multiple regression, an idea which is — how shall we say? — absurd.
There’s a host of other serious problems with Tisdale’s “analysis” (is that even the right word?). But I won’t bother with all of them. After all, I only have one lifetime.