Dave asked:
> "What is your model's Tidal-based ENSO Prediction for the next few cycles? It should reliably predict when short or long periods will occur."

One step better -- here is an extrapolation for two centuries from now

![](https://imagizer.imageshack.com/img923/1623/ZBWOJr.png)

There's still much structural uncertainty remaining in the model. As I said, we have yet to reach the repeat cycle of the Mm,Mf+annual combination which is a basic calibration interval for discriminating & training of the forcing amplitudes.

The rule of thumb is that one day for conventional tidal analysis training is equivalent to at least a year's worth of long-period tidal training analysis. And the nonlinear modulation makes it progressively harder.