For Chinook and coho, state and tribal staff spend the winter putting the relevant numbers together to reconstruct the run of the previous year. Usually this reconstruction is a terminal run of catch and escapement. This dataset is used to make the upcoming forecast in a wide variety of methods. For coho, it is often an estimate of smolt outmigration applied to a marine survival rate (common in Puget Sound). It can also be the current year jacks to the history of adult returns per jacks the previous year such as for Columbia River coho. For Chinook is it often broodyear sibling regressions. It also can be simply a recent year average return (occurs when no reliable predictor has been found). Descriptions of the forecast methods for Chinook and coho stocks are in the PFMC Preseason I reports. The explanations can be a bit cryptic for most but you shouldn't have any trouble. The 2022 version of this report should be out in a couple of weeks.

https://www.pcouncil.org/documents/2021/03/2021-preseason-report-i.pdf/