the film
forum
library
tutorial
contact
Ecology and salmon related articles

NOAA's Spring Chinook Forecast more
Optimistic than Harvest Managers'

by Bill Rudolph
NW Fishletter, February 12, 2013

Using a collection of ocean indicators they have developed, NOAA Fisheries scientists have estimated that 2013's upriver spring Chinook run in the Columbia River will come in around 200,000 fish, counted at Bonneville Dam.

This is more optimistic than the 141,000-fish forecast from the technical advisory committee [TAC] made up of basin harvest managers from the U.S. v. Oregon process. When expected harvest impacts below the dam are included, the feds' prediction adds up to 221,000 fish--56 percent higher than TAC's, and nearly identical to the 2011 return, which was the second largest in the past decade.

The feds also expect 440,000 fall Chinook to pass Bonneville this year. TAC hasn't released its preseason prediction for this run, but said in December it is expected to be better than last year's return, which was somewhat less than the 466,500-fish prediction (to river mouth).

The feds' have based their prediction on an analysis that ranks their ocean biological and climatic indicators over the past 15 years. They say that 2011 overall ranked nearly 6th highest. With 2012 conditions slightly better, they have estimated that the 2014 spring Chinook return will be a little better, 215,000 fish, and the fall run around 460,000.

Their graphs of ecosystem indicator ranking compared to adult returns two years later has an R-squared of 0.79 for spring Chinook, which means the correlation explains nearly 80 percent of the variability between the two factors. The R-squared for fall Chinook is 0.64, and only 0.50 for coho, which means a coin flip is just as good at explaining the variability between the indicators and coastal coho returns. But the NOAA scientists estimate about a 3-percent return of Oregon coastal coho this year, which would put it in the top four returns of the past 14 years.

The NOAA group refined its analysis by weighing principal indicators more heavily, which has improved the R-squared for spring Chinook to 0.82, but hasn't really made the fall Chinook and coho analyses more robust.

NOAA scientist Brian Burke performed even more work on spring Chinook data using the maximum covariance analysis method to predict 2013 returns with 95-percent confidence intervals. This has yielded a 221,000-fish estimate for Bonneville Dam counts plus harvest, with 97,000 returning to Ice Harbor (lowest dam on the Snake) and 19,500 to Priest Rapids.

The feds said 2012 ocean indicators sent a mixed message, similar to other recent years, with some pointing to the "potential" for above-average returns. That included still strong La Nina conditions, negative PDO, positive copepod indicators from May-September, and high catches of spring Chinook smolts in June trawl surveys.

But there were some negative indicators as well, including a late start to the upwelling season, and almost two months before upwelling became strong, along with "very warm" sea surface temperatures in June and July. The upwelling season was one of the shortest in recent years, only 161 days long, compared to more than 200 days in 1999, 2002, and 2009.

"Our best guess is to expect average to above-average returns of coho in 2013 and Chinook in 2014, but similar to the statement we made last year, the mixed signals add greater uncertainty to our predictions," said the latest report.

Burke told NW Fishletter that the various biological and climatic indicators are bound to change in importance in the future, so there is a fair amount of uncertainty attached to the modeling effort. But the feds' predictions have been pretty close to observed returns the past couple of years. For the 2012 spring Chinook return, the NOAA researchers' estimate was 179,000 fish, close to the observed return at Bonneville of 186,000 fish plus an estimated lower-river harvest of 16,000 fish, which added up to 202,000 fish--a prediction that was off by only 12 percent. The 2011 prediction, based on data collected through 2010, matched the observed return of over 221,000 fish nearly perfectly; it was off by only 6 fish.

TAC's prediction for the 2012 upriver spring Chinook run overestimated the actual return by 35 percent, but its 2011 preseason estimate was only 11 percent under. For the three years prior to that, TAC over-predicted the upriver spring run by at least 33 percent.


Bill Rudolph
NOAA's Spring Chinook Forecast more Optimistic than Harvest Managers'
NW Fishletter, February 12, 2013

See what you can learn

learn more on topics covered in the film
see the video
read the script
learn the songs
discussion forum
salmon animation