Assuming that the official forecasts are the best that can be done at the time, I attempt to update them according to ocean conditions in the months later than those that were presumably available to the official forecasters. I use the summer and fall oceanic indexes to conduct reverse stepwise multiple regressions to find a best updated model. Assuming that my updates are overfit due to a short time series, I apply an ad-hoc down-weighting to them.

The reverse stepwise multiple regressions were conducted and are self-documented in Excel spreadsheets.