The Xie paper in #19, although in need of editing, covers the same territory that I am trying to do -- infer both the role of QBO and Solar cycle TSI in ENSO.

> "An 11-year period is found in the PC1 of EOF analysis of tropical TCO variations (figure 1(a)), which is deemed related
to the solar cycle (Shindell et al1999). .... For another, as El Niño Modoki has an 11-year cycle since the 1980s (Ashok et al 2007) and the variations of EMI lead the tropical TCO changes, whether the 11-year cycle of topical TCO also relatives to El Niño Modoki activity which also deserve further investigation by simulation"

Clearly the solar cycle on the decadal range is correlating with ENSO.


The Xie paper brings up another issue for me to ponder.

I am rationalizing why the SOIM wave equation model provides such a good fit to ENSO while no one (to my knowledge) has reported on this approach. Although Clarke and others have described this wave equation formulation, it appears to be comatose in practice. The apparent success that I have been seeing probably comes down to a few reinforcing factors that have never been combined effectively in past studies:

1. Combining the periodic QBO, TSI, and wobble terms as an aggregate forcing.
2. Adding a Hill/Mathieu modulation factor to the wave-equation, scaled to the TSI factor.
3. Filtering the forcing terms so they are smooth but modulate in sync with each of the source frequencies
4. Applying machine learning via symbolic regression to efficiently push the solution in the right direction.

I could see how researchers using Fourier analysis and linear DiffEq approaches might quickly give up, as including a modulated forcing and Hill factor is not conducive to traditional methods of solving the basic DiffEq.

The old adage is always at play to prevent from fooling oneself -- "if it's such a good idea, why hasn't anyone thought of it before?" The answer is that the playing field is complicated. The climate science community has bitten the big bullet and committed themselves to working on immense general circulation models (GCMs) . To consider a radical simplification of a GCM -- to the point that it maps to a simplified wave equation as I am evaluating would be a huge step backward. Yet, that happens over and over again in physics, that a primary effect provides a first-order model to the overall behavior.