Here are slides from a talk (9 October 1997) I gave along with several other UW people exploring the different hypotheses as to what caused the run "failure".

I have prepared a couple of figures to explore what role SST might have played in the small Bristol Bay return this year (1997).  Shown here are Sea Surface Temperature Anomaly maps for the months April, May, June and July for the years 1982, 1983, 1996, and 1997. What is most striking is that 1982 does not resemble 1997 in any way.  The Bering Sea was relatively cool in the summer prior to the 1983 ENSO teleconnected impacts.  Conversely, 1997 shows a massive warming much earlier than can be attributed to ENSO.

I have also made some animated maps of SST anomalies.  First are the weekly anomalies from April 6th to July 20th, the time when the sockeye should return.  It can be seen that large positive temperature anomalies arrived suddenly in late May.

Many of the fish returning in 1997 would have migrated seaward in the spring of 1995,  Here's a looping movie of 1995 SST anomalies that spring.

Here's a look at Precipitation Anomalies since 1995 (was there a regime shift?)