>> ... correlation of the length of day with ancient
>> sediment layers indicates a non-linear increase over hundreds
>> of millions of years, so that a simple backward extrapolation
>> of present rates yields incorrect results.
>
>Interesting stuff. While I believe SOME non-linear change in the earth-
>day length can be attributed to a decreased tidal effect over time as the
>average earth-moon distance increases, this nonlinear effect must surely
>be minor.
It's been years since I read the referenced material, but off the top
of my head I seem to remember that the non-linearity has been looked
at pretty closely. It seems to correspond with the amount of change
in tidal friction in the ocean, which changes considerably as the
continents drift. When the continents are all squashed together,
there is much less friction than now, when they're all appart and
there's lots of coastline to create extra energy dissipation.
>So I strongly suspect that any appreciable deviation from
>linearity that has been observed, or rather inferred from observations,
>probably reflects observational errors of some sort. These could come in
>many ways, such as in estimating the ancient coral age.
I believe that there has been considerable cross checking with
other sedimentary deposits. The textbook I referenced had a chart
showing the length of day since the Cambrian, and it didn't show
anything like an anomaly.
>How about turning the problem around, and using the "observed" nonlinear
>increase in earth-day length to recalibrate the ancient coral age? How
>old would the ancient coral age need to be to make everything linear?
An interesting idea, but I'm not going to attempt it. :)
Alan