I went back to 1979, and found that out of 850mb wind, 200mb wind, sea-level pressure, OLR, SSTs, I made a custom index for subsurface temps at -150m.. I found that it was >150% correlation than all those things listed above, for the N. Pacific PNA pattern in now-time.

Because the oceanic kelvin waves that produce said SSH anomalies are generally preceded well in advance by a restructuring of the IPWP, which is coupled with atmospheric circulation/convection. The IPWP vacillations are even lower frequency than ENSO (multiyear to intradecadal for IPWP vs seasonal to multiyear for ENSO).

I've also found that sea-level height, or sea-level height anomalies has a very high correlation with the H5 pattern +time. Can you explain that one Phil?

Y’all need to stop posting CFS trash on here, that s**t might as well be straight out of the Y2K era. There’s a reason that joke of a seasonal model is being decommissioned before decade’s end.

Z500 verification scores demonstrate that you are wrong. The EPS and EPS AIFS have been the highest performing model suites.