Initialized Earth system prediction of the Interdecadal Pacific Oscillation as motivated by Mark Cane

At the Mark Cane Symposium at Lamont in 2014, I presented experimental results from new initialized Earth system predictions with CCSM4 that showed a transition to positive IPO averaged for lead years 3-7 (2015-2019) from the model initialized in 2013. This meant the verification would not occur for another 5 years. After my talk, Mark asked if I had published that prediction. Of course I said no, that it was experimental, and it would not verify until years after any paper could be published. But he strongly encouraged me to submit the paper with the prediction. He said that in the early days of ENSO prediction, he and Steve Zebiak also were making experimental ENSO predictions that would only verify after their papers were published. But they believed they had a physical understanding of the processes they were modeling that were contributing to prediction skill, and it was essential that such predictions appear in the literature. If wrong, we could learn why the prediction was wrong. If right, we could have credible evidence for the physical processes being hypothesized that could produce skill. He noted that since I had ideas as to the physical mechanism for the IPO transition, I was in a similar situation to the one he was in during the late 1980s. So we wrote up the results, and the paper was finally published in 2016 for a prediction that would not verify fully until 2018. After waiting two more years, the prediction did indeed verify, thus providing evidence that the mechanism we were simulating was likely providing skill for the IPO prediction. But then in 2019-2020 the Australian wildfires occurred, and the IPO predictions made for the time periods after 2020 were not skillful. Could the wildfire smoke have triggered a multi-year La Niña that resembled the negative phase of the IPO? There is now evidence that that was the case as derived from initialized hindcasts made with and without the effects of the wildfire smoke.