We use a mixture of physical and statistical models to forecast ENSO, including physical models of varying degrees of complexity. A long-standing challenge has

We use a mixture of physical and statistical models to forecast ENSO, including physical models of varying degrees of complexity. A long-standing challenge has been for these models to exhibit reliable skill at lead-times beyond ~6-12 months, especially when forecasts are initialized in the early Spring. Data-driven models ("AI" or "machine learning") seemed to break through this barrier a bit, but what's notable about this paper is that a physical model with a new underlying mechanism seems to catch up a bit.