(John Sager) writes:

Day length is already increasing due to the transfer of angular momentum
to the moon via tides, and there will be a smaller exchange with the sun.
In principle you should be able to model this, but there are also changes
to the earth's moment of inertia which affect angular velocity, and these
seem to be quite unpredictable - who would have thought before the fact
that we wouldn't need leap seconds for the past 5 years, given the
previous insertion rate?

One of the major sources of this variability comes from long-term cycles
and trends in weather and climate. For example, during those phases of
the biennial and decennial atmospheric oscillations when the jet stream
is more vigorously active, it carries a higher angular momentum, and
therefore the rotation of the Earth must slow down a bit to conserve the
total angular momentum of the Earth-plus-atmosphere; the effect is small,
but the anticorrelation between jet-stream activity and the rotation rate
of the Earth is indeed measurable --- albeit not long-term predictable.
(It is quite likely that changes in circulating ocean currents may also
have a similar effect, albeit on a much longer timescale.)

As for the time variations in the Earth's moment of inertia, one
contributing factor is thought to be changes in the amount of water
locked up as ice in the polar caps versus in liquid form in the ocean's
equatorial bulge; another is wind-driven variations in sea-surface height;
yet a third is changes in the density distribution of the ocean due to
variations in salinity and temperature. Again, all these effects are small,
and measurable, but not predictable over the long term.


-- Gordon D. Pusch

perl -e '$_ = \n"; s/NO\.//; s/SPAM\.//; print;'