Columbia River preseason adult spring Chinook run size and arrival timing predictions, 2008

Preseason Adult Spring Chinook Run Size and Timing Forecasts
Preseason Forecasts Data Dates
Method Source URL
to river mouth
Upriver Spring Chinook and Snake R. summer Chinook. Sum of Bonneville Dam counts plus fish of upriver origin landed in lower river fisheries. Forecast for 4 and 5-year olds. January 1-June 15
Various historical years
Age-specific linear regressions of cohort returns in previous run years. TAC,
U.S. v Oregon
1, 2, 3
Columbia River Fisheries, Fish Division, ODFW
Salmon Preseason Reports, Pacific Fishery Management Council
to Bonneville Dam
Cumulative Bonneville Dam counts March 15-June 15 1982-2007 Date-based linear regression last year's jacks CBR Escapement Forecaster Adult Forecasts, Columbia Basin Research, UW
to Bonneville Dam
Modeled cumulative Bonneville Dam counts 1982-2007 Gaussian function fit-based linear regression CBR Adult Forecasts, Columbia Basin Research, UW
April 22
Peak Arrival Day to Bonneville Dam
Previous year jack timing at Bonneville Dam, River flows, Coastal downwelling 1978-2007 Linear regression model CBR
Genetics and Environment Timing Model
Adult Forecasts, Columbia Basin Research, UW
  1. Joint Columbia River Management Staff. 2008. 2008 Joint Staff Report: Stock Status and Fisheries for Spring Chinook, Summer Chinook, Sockeye, Steelhead, and Other Species, and Miscellaneous Regulations. Online:
  2. Pacific Fishery Management Council. 2008. Preseason Report I: Stock Abundance Analysis for 2008 Ocean Salmon Fisheries. Online:
  3. Columbia River Fisheries. 2008. 2007 Adult Returns and 2008 Expectations, Columbia River. Online:
  4. Anderson and Beer. In review Ecological Applications. "Oceanic, Riverine, and Genetic Influences on spring Chinook salmon migratory timing."

The Columbia Basin Research (CBR) Columbia River adult spring Chinook salmon run predictions now include mean run timing as well as run size ( The preseason prediction for 2008 is that the run size will be above average but the run timing will be average, with the peak run occurring about April 22.  Figure 1 shows the 2008 predicted size and timing plotted with historical size/timing information using the two CBR methods.

  • Method A is date-based and defines the spring Chinook run as between March 15 and June 15. This method is used by the Escapement Forecaster to produce the preseason run size prediction for 2008 of 307,000 fish. The June 15 cutoff date has been used since 2005 to best match the estimate and run dates used by the Technical Advisory Committee.
  • Method B is fit-based and defines the run size with a normal arrival distribution and no cutoff dates. The 2008 preseason run size prediction is 197,500 spring Chinook. The results of this method are used in conjunction with the Genetics and Environment Timing model to produce peak arrival day predictions.

Figure 1. (A) Historic median arrival day and cumulative counts during counting period Mar 15 and June 15. (B) Peak Arrival Day and run size determined by fitting the spring run to a normal distribution.  Points are in center of text. Blue-circled 2008 represents the estimated timing and run size based on the two methods. Axes are the same in both graphs.

Brief details of the preseason predictions and methods are in the following sections.

Arrival Distribution and Run Peak

The duration of the spring Chinook run is defined by two methods. The Escapement Forecaster, which uses the same cutoff date as used for management of in-river harvest, defines the run as the total number of adult spring Chinook salmon passing Bonneville Dam between March 15 and June 15. The Genetics and Environment Timing model, a better ecological characterization of the spring run, simultaneously describes the distributions of the spring, summer and fall runs by fitting three superimposed normal distributions to the combined run (Figure 2). The mean arrival date of each normal distribution characterizes the arrival peak independent of the method used to determine run size. Please refer to Run timing of adult Chinook salmon passing Bonneville dam on the Columbia River for further details.

Figure 2. Passage counts of adult and jack Chinook salmon at Bonneville Dam in 2006 showing the tri-modal pattern. The thin line depicts the observed counts. The thick lines depict fitted Gaussian functions for spring, summer and fall runs. The fixed spring run cutoff date is depicted with a vertical red line.

Genetics and Environment Timing model Preseason Run Timing Prediction

The spring Chinook peak run timing prediction is based on indices of the run's genetically based run timing plus indices of the ocean and river currents fish encounter on their homeward migration. The genetic run timing basis is inferred from the timing of the previous year's jack run.  The effects of ocean currents encountered by both jacks and adults fish approaching the Columbia River are inferred from the January coastal downwelling indices in the years of their respective returns.  The river velocities jack and adults encounter are inferred from river flows in the month prior to their arrivals at Bonneville Dam. The model, calibrated with data between 1978 and 2007, has an r2 = 0.78 (Figure 3). A manuscript detailing the run timing model is in review (Anderson and Beer, submitted to Ecological Applications).

Figure 3. Spring Chinook predicted arrival date vs. the mean arrival date (1978-2007) from the Gaussian function fit to the "observed" counts.

Escapement Forecaster Preseason Run Size Prediction

The Escapement Forecaster, the date-based method, preseason adult run size is estimated using a linear regression of the previous year's jack run size against adult run size using adult passage counts at Bonneville Dam since 1982. Figure 4 illustrates the regression obtained with run size determined by cutoff dates: Adult.count = 25667 + 13.9 *Jack.count, r2 = 0.79.

Figure 4. Regression of adult counts on previous year's jack counts (spring run only). Large green dot is the 2008 preseason run size prediction based on the number of jacks observed in 2007 (20239).

The run size prediction using the fit-based linear regression model gives r2 = 0.72.