“Are Extreme Events ‘Teachable Moments?'”
a talk by Prof. Michael Wallace, Atmospheric Sciences
May 9th, 2014. This review by hh.
Mike is a much honored meteorologist and climate scientist, a member
of our National Academy of Sciences who is noted in part for
discovering and documenting ‘teleconnections’ between weather events
separated spatially by large fractions of the earth’s circumference.
That, and he is one of the nicest people I know.
His talk yesterday was driven by his concern that discussions in the
‘Media’ of extreme events attributable to ‘global warming’ ..
Hurricane Sandy for a recent example .. are exceeding both good
science and good sense.
He cited recent work .. his and others .. to assert that detecting
trends in new weather extremes will not significantly reflect
increasing _variance_ in the data, but a small shift in their mean
intensities. A consequence of this, for example, is that if the near
future holds new droughts, their _frequencies_ would most probably be
indistinguishable from the present, with small and nearly impossible-
to-detect extra _intensity_ from ‘global warming’. Thus attributing
near-future extreme weather events as increasingly ’caused’ by warming
is incorrect, and claims of this flavor damage our credibility.
Mike is an environmentalist with the best of us, but suggests our
better concerns should be more broadly distributed. We should not
casually try to attribute exceptional weather to global warming, but
direct our attention more broadly to trends in pollution, disease,
extinctions, and the many pressures upon the health of species.
There is much to recommend this.
I agree, sort of, but ..
Slight upward shifts of average temperatures, when convolved with
episodes of crop failures, for example, would asymmetrically distort
the probability distribution of costs [*]. In this case, extremes of
costs might more confidently be attributable to ‘global warming’, than
would extremes of weather. [But for gosh sakes, let’s continue to
study the latter, especially in the tropics.]
Quibble Quibble? Oh, well, that’s what professors are paid to do.
Meanwhile, let’s hear it for a carbon tax.
hh
[*] Asymmetric Probability Distribution Functions [PDFs] are
referred to as ‘skewed’. Positively skewed PDFs have ‘high-
end tails’ [or ‘black swan’ events] that may significantly
affect the means, but have too few events to permit confident
detection of accelerating trends.
> o__
> _,>/’_ o__,
> (*) \(*) /\/\