A few years ago Tim Palmer wrote a nice paper showing that the Lorenz
chaos ideas probably do not apply to climate.
Sorry - but I can't recall the reference. Somewhere obscure.
Tom.
+++++++++++++++++++
On 2/27/2012 10:22 AM, Mike MacCracken wrote:
Hi Jim—I’ll certainly agree climate can be path dependent, so some
locations in climate space that are like attractors. The classic
suggestion, however, that the energy balance models gave that had a
snow-covered Earth as an additional possibility I believe was flawed
due, at least in part, to their failure to treat the diurnal cycle
(there is, in my view, just no way that high albedo snow survived under
the noonday tropical sunlight).
I’d also comment regarding climate sensitivity that the concept itself
has problems—a good example being that orbital element cycles top of the
atmosphere have virtually no change in annual, global integral, but the
result is ice ages, so a very large climate change from a redistribution
of incoming solar by season and latitude. IPCC and others account for
this by counting the ice sheet albedo as an external forcing, but that
is only plausible because (most—Berger et al. do) computer models can’t
run long enough to simulate ice sheets as an internal feedback.
Given both issues, I guess the admonition we’d both agree on is be
careful and thoughtful.
Mike
On 2/27/12 9:09 AM, "Jim Fleming" <jflem...@colby.edu> wrote:
Dear Mike,
Fine, but those quotes are from papers by Lorenz on climate, not
weather. They are not essentially about the propagation of
transient signals or random events, or even "stability," but, as he
says, indicate that the solutions to any boundary constraints plus
dynamic forcings lie along attractors. That is, there are multiple
solutions and we can't treat forcings as linear, at least in the
"climate of the models" he was exploring.
Jim
On Sun, Feb 26, 2012 at 8:50 PM, Mike MacCracken
<mmacc...@comcast.net> wrote:
Hi Jim--I think you need to be careful in applying the Lorenz
butterfly
effect to climate--the equations are for the weather and how it
varies
within a system that is bounded by external conditions.
For climate, so the average of the weather when there are a
particular set
of external boundary conditions. Now, the climate can certainly
have some
path dependent phenomena--so hysteresis of various types (e.g.,
Greenland
ice sheet, carbon cycle given the various ways it gets stored
and released,
etc.), but I just don't think that it is random in the sense
that Lorenz was
portraying the situation for the weather.
Certainly there is some variability given a certain set of
external boundary
conditions and so it can seem that the climate sensitivity
varies somewhat,
and it is certainly possible that one can get to points where
there may be a
jump between different possibilities created by hysteresis (like
we can
reach a warmth were Greenland ice sheet melts away, and this
might happen
pretty rapidly), but I would not call this a random sort of
event in the
butterfly sense.
As to the paleoclimatic jumps, it may well be that there are causal
mechanisms for those--bursting of ice dams and outflow of
freshwater, etc.
Again, that does not really seem random to me, but a process
that can have
jumps--and it may be hard to predict precisely, but is it really
a butterfly
event, etc. It seems to me were the Earth system's climate, we'd
have not
had so many millions of years between major glacial periods,
we'd not have
had the stable Holocene, we'd not be finding the Milankovitch
cycles, etc.
Best, Mike MacCracken
On 2/26/12 1:16 PM, "Jim Fleming" <jflem...@colby.edu> wrote:
> Before you go too far down the path of CO2-climate
sensitivity, please
> consider the work of Ed Lorenz who was interested in models that
> emulate sudden jumps from one equilibrium state to another, in
> patterns similar to the paleo-climatic record.
>
> In his paper "Climate is what you expect," Lorenz wrote:
> A climatologist encountering [fluctuations in real data] could
not, on
> the basis of the data alone, say whether the dominant cause of the
> changes was external or internal.
>
> Elsewhere (climate of the model), for climate variations over
a 100-
> year interval, as given by a simple one equation model designed to
> support internally produced climatic changes:
>
> "Two or more distinct climates may well be compatible with the
same
> external conditions."
>
> Finally:
> "If the climate system is treated as a dynamical system--a system
> whose evolution is governed by precise laws, or, more
frequently, by
> equations that represent such laws, the climate then becomes
> identifiable with the attractor of the dynamical system‹but the
> dynamical system may have more than one attractor!"
>
> Can we rely on single variable linear or logarithmic (non-chaotic)
> forcing to predict the future state of the climate system?
>
> Jim
--
You received this message because you are subscribed to the Google
Groups "geoengineering" group.
To post to this group, send email to geoengineering@googlegroups.com.
To unsubscribe from this group, send email to
geoengineering+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/geoengineering?hl=en.
--
You received this message because you are subscribed to the Google Groups
"geoengineering" group.
To post to this group, send email to geoengineering@googlegroups.com.
To unsubscribe from this group, send email to
geoengineering+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/geoengineering?hl=en.
