HardwareBanter - View Single Post

**Eric Stevens**

On Wed, 17 Oct 2018 10:58:39 -0400, Wolf K
wrote:

On 2018-10-17 00:14, Eric Stevens wrote:
[...]
Are you referring to which side of the current to sail on according to
direction of travel? If you are, he is right, as any sailing
directions will confirm. If you are not referring to that, I don't
understand what you are getting at.
[...]

It's about heat transported by ocean currents.

The Atlantic Conveyor moves warm water from the (sub-)tropics in the
northern/northeastern Atlantic. Since it floats on top of the colder
water there, that cold water subsides, and flows south (more or less)
well below the surface. The Conveyor is part of the worldwide
circulation/transport of heat by ocean currents. Here's a link that both
explains the system, and presents recent attempts to understand the
system better:

https://www.carbonbrief.org/the-atla...-rapid-project

Background as I have distilled it from many decades of reading science
journals and magazines:

As you know, water has a high specific heat, so even slight changes in
this system of warm and cold ocean currents can have large effects on
the circulation of air above the oceans, ie, the weather. See El Nino
and El Nina. If the Conveyor changes more than X (where X is at best a
rough estimate at this time), the climate of the northern Atlantic will
change. I.e., the climate from Greenland to Norway will change.

The ocean currents are obviously one of the factors driving the annual
weather cycles ("the climate"). The climate as a whole is a network of
feedback loops. Such networks are "chaotic systems". They cycle around a
sequence of state changes (eg, the seasonal changes of weather in your
locality) with some variability in each cycle. If some factor in the
system changes beyond some limit, the whole system tips into a new cycle
of state changes.

The unknowns are the triggering factors and their roles in the feedback
loops, and thus the rate of change into a new cycle of changes. The
"tipping point" could be on the order of a few seconds to many thousands
of years. The earliest climate models (1970s) suggested that climate
could change as quickly as about 100 years, depending on which factors
changed and by how much. Since these models did a good job of
"retrodicting" (matching known climate changes), these results created a
puzzle. That drove the creation of more powerful models, which have
merely refined these results: it is in fact possible for the climate to
change very rapidly. Since then, minor climate changes (such as the
Little Ice Age of the late Middle Ages) have shown that climate can
change very quickly indeed. Finer grained data from sediments and rocks
suggest that climate has occasionally tipped quite rapidly in the past,
probably on the order of a thousand years or so.

Statistics is not the best tool for analysing and understanding chaotic
systems like the weather and climate. That's why even eminent
statisticians are poor guides to understanding weather and climate. NB
that before the advent of supercomputers, weather prediction was
statistical, and notoriously unreliable beyond a short time frame, which
in Great Britain was approximately 1/2 a day (as I recall only too well
from my childhood there). Supercomputers enable the modelling of
multiple feedback loops one state-change at a time: the current state is
the input for calculating the next state. This has improved weather
prediction so that it's reliable for up to two or three days here, and
pretty good for up to a week or so. Even so, every so often the
prediction is badly off: some factor exceeds some limit, and instead of
a shower we get a thunderstorm.

Basically, any system of feedbacks between three or more entities is
chaotic. See the Three Body Problem for a very old example.

BTW, life itself is a driver of weather, and in the long run of climate.
Eg, ground cover affects the rate of water loss in the soils, and so
affects the hydrologic cycle that we call "rain."

Best,
--

Regards,

Eric Stevens