Given a 30 percent drop in the Atlantic conveyor belt flow rate in 12
years:
http://www.timesonline.co.uk/article/0,,2-1898493,00.html
it may be that there is a powerful but not yet understood biological
feedback occurring in the ocean due to a reduced nutrient flow from
the depths caused by the conveyor belt shutdown. Polar fresh water
runoff is suggested to be the cause of the conveyor belt slowdown.
The conveyor belt is driven by salinity density changes, and
injection of fresh water in the polar regions disrupts the cycle.
The salinity changes, nutrient changes, and thermal changes brought
about by the slowdown may be affecting both the location of and
population of carbon sequestering organisms, and the carbon
sequestration rate in general. A sudden population change would
occur more like a step function than a logarithmic function.
"The results, published today in Nature, show that the outward flow
of the Gulf Stream has not changed, but the strength of the cold
water returning from the Arctic has fallen by 30 per cent since 1992.
Over the same period, the flow of warm water branching off towards
Africa has increased by 30 per cent. This suggests that the warm
waters are being diverted away from Europe."
A sudden population change could cause sudden onset long term linear
or non-linear effects. Given the large number of feedback cycles
involved a non-linear and even exponential parameter change seems
likely but we probably will have to wait until it is too late to find
out.
Maybe you have a specific underlying model in mind to explain the
regime change Robin? Attempting curve fitting for extrapolation into
the future is sometimes a lot more meaningful if there is an
underlying physical model to justify the curve family. However, in
this case it looks like there has been an abrupt regime change of
some kind, and we don't even know what family of curve to apply.
Horace Heffner
