Warmist Trenberth on attempts by climate models to predict behavior of ENSO, NAO, PDO, etc: "That is very much experimental science"

Q&A On Two Climate Science Questions – A Contribution By Kevin Trenberth | Climate Science: Roger Pielke Sr.

[Trenberth] Global climate models deal with all time scales. ENSO is predictable for up to a year or so.  The NAO is largely a natural atmospheric mode and has limited predictability of a few weeks. The PDO is not independent of ENSO and it depends on what is meant by “PDO”.  If it is truly “decadal” then it has persistence predictability for 5 years or so. Climate models have varying degrees of success in simulating these and prediction requires initialization of the ocean and other climate components.  That is very much experimental science but there is considerable promise, as outlined in a recent article in BAMS by Jim Hurrell and colleagues (which you wrote a comment on that failed to impress me).  The ability of models to deal with modes and regional climate is part of chapter 14 of the next IPCC report.  There is promise, but models need to improve.