Date: Mon, 14 Apr 2008 14:46:47 -0700
From: "Gary G. Ford" [1946-2013]
Subject: ???? - [Fwd: Re: Researcher: Basic Greenhouse
Equations "Totally Wrong"]
To: "Astronomer Bill" <****@wwheaton.com>
CC: "Jentje Goslinga" <***********@shaw.ca>,
"Mark Thornally" <**********@yahoo.com>,
"Daniel Bowen" <************@gmail.com>,
"Arjan,Stam" <**********@hotmail.com>,
"Gary G. Ford" <1946-2013>
--- "Gary G. Ford" [1946-2013] wrote:
Astronomer Bill,
Daniel sent a controversial news article
on a claim by a couple of experienced
Hungarian Climate Simulationists, one of
whom resigned in protest from NASA for
NASA refusal to allow his "discovery" to
be revealed (see attached).
Arjan also given an elaborate commentation,
and his quoting of Milankovitch Theory details
(I am wondering however about one of the sentences
...) precedes the actual item which Daniel forwarded.
Now, Dr. Goslinga is included because he a very
experienced and insightful Fluids Events Applied
Mathematician, Numericist, Algorithmist, and past-
-master high precision numerical methods/solution
designer/programmer, with ample C/C++ and Intel
CPU's FPU ( 80-bit [extra-double-long] registers)
Assembly Code, as well as expert in traditional,
somewhat more limited, Fortran variants (including
modern) experience.
Dr. Goslinga has made many practical, numerical,
often very complete and mathematically clean,
ingenious, solutions for flow or physical process
simulation situations, modeling quite well many
real world events, such as in Gas/Oil Pipeline
Networks, Petroleum Reservoir, Production
Simulations, and other process, phenomena or
physical problems of science, engineering, and
industry, including cavitation bubble collapse,
as well as formation of residual stress in cooled,
extruded thermoplastic pipes.
Mark Thornally is a Witness for The Future.
Now to the title, of the letter, which quotes,
"Totally WRONG" ...
An example of an often "not quite right"
habit of hydrodynamicists comes to my mind,
accordingly:
Now, I don't know about 'totally' but I do know
what Birkhoff said in his book "HYDRODYNAMICS:
A study in fact, theory and similitude" which
Dover reprinted some years ago:
(Paraphrase follows, as next paragraph:)
The Kelvin "Conservation of Vorticity" Theorem
WOULD apply to Real Navier-Stokes Flow, TOO,
if the velocity field could be fully expanded
analytically [GGF Note: I presume in time and
space, but I no longer remember which], to an
arbitrary number of terms, but should not, as
Navier Stokes Solutions in General cannot even
be guaranteed everywhere having two orders of
continuous derivatives ... And in fact, real
flows, thought to be well modeled by the Navier
Stokes Equations, both theoretically and in
in practice, often fail to conserve vorticity.
(end of paraphrase/approximate quote)
[GGF: I know spatial derivatives can be so limited,
as I studied this matter while researching/writing
my M.Sc. thesis in the Mathematical Theory of
Viscous, Incompressible, Navier-Stokes Flow,
which last I considered in application to flow
in pipes with a bulge or a constriction over a
bounded length. Also took a two-semester graduate
level Mathematical Hydrodynamics course, several
graduate level engineering fluid mechanics courses,
did a detailed Wind Tunnel course project, in which
I measured turbulence in wake of a dual, rather than
the traditional single, cylinder. While I did not
submit my Mech Eng thesis, dropping to work in the
Calgary, Alberta "Oil Patch" and support my wife
and young children (now working adults), instead,
I was earlier awarded my Mathematics M.Sc. for the
Pipe Flow Mathematical Study. I knew SOMETHING
about such matters at one time]
In the context now of the "Not Generally for Navier
Stokes Flow" Series Expansions, I am aware Applied
and Numerical Mathematicians, even Engineers, often
ASSUME and use truncated series expansion approaches
which might not be fully "hygienic" from a purely
mathematical standpoint, for flows supposedly modeled
well by the Navier Stokes Equations, usually the
Incompressible Case, in various projects, applied or
of research.
Here's the awful truth (also discussed by Birkhoff):
Real, VISCOUS, and Turbulent, Flows are renowned
for Dissipating (NOT CONSERVING) vorticity, and
viscous flows for CREATING vorticity in conditions
of "velocity shear" near a surface. The latter are
often subsumed, even dismissed, as "boundary layer
effects".
So to "tweak the truth" and "damn the consequences"
I have found to be common practice in the fluid flow
arena of Science and Engineering.
For decades after Kelvin, Horace Lamb, PEOPLE also
ASSUMED that "non-penetration" boundary conditions
(great for 2-D - "planar" [section] flows of "ideal"
[frictionless] fluids) were sufficient to handle
real world - say aerodynamic, in which case all wind
and drag forces were presumed to be explicable from
"pressure surfaces" or vortex "pressure centers" -
flows (the ideal, planar cases being often handled
by holomorphic conformal complex variable maps of
harmonic functions of two variables), but it's not
so good an assumption for many real world situations.
For flow of real world (non-liquid Helium, ie) fluids
in pipes and many other real world flow cases, where
viscous friction, and/or turbulence, is/are dominant,
"non-penetration - so called "Neumann" boundary
conditions - are NOT appropriate or realistic and
must be replaces by No-Slip (zero tangential contact
velocity), Dirichlet, Boundary Conditions, or mixed
types and more/other.
So I do not find it impossible that many numerical
climate simulations have "over looked something"
given what I know about presumed "simpler" and
"more scientifically studied" fluid flow theory
and computations.
Read now the attachment below (after the ps)?
Gary, 1946-2013
PS: For the sake of Aerofoil Numerical Simulations,
abundant wind tunnel results generally are, luckily,
or have been, available to check and correct methods,
so that Aircraft Lift/Drag/Performance Simulations -
including for Supersonic Flow Equations, which last
I was once been taught - are generally pretty good,
although new and adventuresome designs often need
extensive Wind Tunnel Testing pre-Simulation, EVEN
IN THE PRELIMINARY STAGE, as Numerical Simulation
of New Approaches cannot be guaranteed 'correct'.
>>==>>
I doubt however similar is true for "Climate Simulations".
>>==>>
*In Petroleum Reservoir Simulation, I have seen ways in
which people can/do cheat, often unconsciously, even
ignorantly, to skew results toward what 'experts' may
appear to expect. I doubt that, Politically Charged,
Climate Simulation is necessarily more 'Scientific'
than are such somewhat "warped" and "assisted" [to get
the presumed "right answer/conclusion"] industrial flow
simulations.*
>>==>>
NOW ... the controversarium follows ...
[THE ABOVE ARE LARGELY MY OWN BIGOTED OPINIONS.]