Future prospects - Ice Ages and Interglacials: Measurements, Interpretation, and Models - page 344

temperature remains a mystery. One thing is certain: not all proxies are equally
credible. M&W were concerned with the statistical processing of data, assuming
the data were adequate. More likely, the data were poor in quality, such that even
the best statistical analysis would result in garbage in-garbage out (GIGO).
As Steve McIntyre (climateaudit.org) pointed out:

''The fundamental problem in paleoclimate is not the need for some novel
multivariate method, but better proxies and reconciliation of discordant existing
'proxies' . Team reconstructions use highly stereotyped proxies over and over

again in different guises

...

... .''

The details of the statistical analysis in M&W (and the 13 commentaries
that followed it) are quite complex and are only intelligible to specialists. One
possibility raised by M&W was:

too weakly connected to global
annual temperature to offer a substantially predictive (as well as reconstructive)
model over the majority of the instrumental period. This is not to suggest that
proxies are unable to detect large variations in global annual temperatures (such
as the differences that distinguish our current climate from an ice age). Rather,
we suggest it is possible that natural proxies cannot reliably detect the small and
largely unpredictable changes in annual temperature that have been observed
over the majority of the instrumental period.''

'' ... it is possible that the proxies are ...

This appears to be the truth of the matter. Proxies seem able to detect very
large excursions in temperature, such as occur in transitions between ice ages and
interglacials, but probably are not able to resolve small temperature changes
within an interglacial. The problem of backcasting historical temperatures from
proxy measurements calibrated during a limited period of overlap between tem-
perature measurements and proxy measurements is very complex and requires
very sophisticated statistical analysis which might be beyond the capability of
climate scientists. M&W attempted to bring such a sophisticated statistical analysis
to bear on the problem:

we conclude unequivocally that the evidence for a 'long-handled' hockey
stick (where the shaft of the hockey stick extends to the year 1000 ad ) is lacking in
the data. The fundamental problem is that there is a limited amount of proxy
data which dates back to 1000 ad ; what is available is weakly predictive of global
annual temperature. Our back-casting methods, which track quite closely the
methods applied most recently in Mann et al. (2008) to the same data, are unable
to catch the sharp run up in temperatures recorded in the 1990s, even in-sample
...

'' ...

. Consequently, the long flat handle of the hockey stick is best understood to
be a feature of regression and less a reflection of our knowledge of the truth.