>>The paper is essentially Curry whining that climate science is too difficult, instead of getting to work and figuring out the physics and math, like the rest of us try to do.
> Where is she whining ? I mean she is critizising certain features of GCM's, like in particular that
>> The solution of Navier–Stokes equations is one of the most vexing problems in all of mathematics: the Clay Mathematics Institute has declared this to be one of the top seven problems in all of mathematics and is offering a **$1 million prize** for its solution.
Obviously they are unsolvable in general because the equations are largely under-determined. Boundary and initial conditions alone do not provide enough constraints to solve the set of equations in 3-dimensions. No one complains that two equations with three unknowns is unsolvable.
Compare that to Maxwell's equations which tend to be more determined because of the interactions between the B and E fields and how boundary conditions are applied.
Here's an example of what I am talking about. Consider my solution to the QBO. This is essentially solving Navier-Stokes for atmospheric flow. When I worked out the primitive equations, I knew that I would have to be ruthless in reducing the dimensionality from the start. The first simplification was working at the equator, which eliminated a few of the terms arising from the Coriolis effect. QBO is also a stratified system, so the vertical cross-terms are inconsequential. Of course the time-space part of Navier-Stokes was separated by noticing the standing-wave nature of the phenomenon has a wavenumber of zero, which obviously helped quite a bit.
I eventually ran into the last remaining under-determined constraint dealing with a transverse spatial term; I eliminated this by cleverly associating the latitudinal equatorial flow line with a nodal lunisolar forcing. This provide a running boundary-slash-initial condition and thus reduced the initially "unsolvable" Navier-Stokes equations to a Sturm-Liouville formulation -- which fortunately had a remarkable closed-form analytical solution.
So not only did I solve a variant of Navier-Stokes -- i.e. the primitive equations of Laplace's tidal equations -- but it didn't even require an iterative numeric solution. I do have to admit that I used many of the ideas from understanding how to solve the 2-dimensional Hall effect via Maxwell's equations in coming up with the answer.
The other example is ENSO. This is also a standing wave equation which has a rather straightforward solution but needs a numerical computation assist to iterate the solution, unless one can leave it as a convolution of a Mathieu function with sinusoidal forcing terms. Of course I partially adapted this idea from sloshing research in the hydrodynamics literature.
> Where is she whining ?
Perhaps I should be the one whining to Professor Curry ...
**WHERE IS MY $1,000,00.00 PRIZE AWARD !!!!!** :-B :)