Language selection

Search

Terms of Reference

Run Timing and Diversion Rate Models for Fraser River Sockeye

Regional Peer Review Process - Pacific Region

October 27-28, 2015
Nanaimo, BC

Chairperson: Peter Chandler

Context

Pre-season forecasts of adult Fraser Sockeye Salmon (Oncorhynchus nerka) marine run timing and diversion rate (the proportion of fish migrating through Johnstone Strait versus Juan de Fuca and Johnstone Straits combined) are essential for planning fisheries and are a Canadian responsibility under the Pacific Salmon Treaty (PST).  Fisheries and Oceans Canada (DFO) Fisheries Management annually requests Science Branch to provide pre-season forecasts for marine run timing and diversion rate of Fraser River Sockeye Salmon stocks.

The performance of both statistical models currently used to generate the run timing and diversion rate forecasts has degraded during the last decade.  Additionally, the oceanic variables used in the current timing model are themselves derived from an oceanographic model that is not domestically maintained and there is a risk that these data may not be available in the future.  Recently, regionally developed and supported oceanographic models have been incorporated into new run timing and diversion rate models in attempts to improve performance of these models.

The objective of this review is to assess the performance of newly developed run timing and diversion rate models, including those that utilize near real-time oceanographic data.  DFO Science will utilize advice arising from this Canadian Science Advisory Secretariat (CSAS) Regional Peer Review Process to provide Fisheries Management, Pacific Salmon Commission and the Canada-US bilateral Fraser River Panel with annual forecasts of Fraser River Sockeye Salmon run timing and diversion rate.

Objectives

The following working paper will be reviewed and provide the basis for discussion and advice on the specific objectives outlined below.

Michael Folkes, Richard Thomson, Roy Hourston. Evaluating Models to Forecast Fraser Sockeye Return Timing and Diversion Rate. CSAP Working Paper 2013SAL07.

The specific objectives of this review are to:

  1. Review statistical models developed to forecast  the run timing of Early Stuart R. and Chilko R. Sockeye Salmon (four models considering six variables), and the diversion rate of the combined return Fraser Sockeye Salmon stocks (four models considering seven variables).

    Models include naive time series models, single variable regression, generalized linear models, and generalized additive models.

    The input oceanographic variables include sea surface temperature, salinity, and sea currents. The sea current data were independently derived from the NEPSTAR (Thomson et al. 2013), OSCAR (Bonjean and Lagerloef 2002), and OSCURS (Ingraham and Miyahara 1988) models.

    The influence of a geomagnetic variable will also be evaluated for the diversion forecasts.

  2. Describe the data inputs and characteristics (e.g. source and date available) of the different models.
  3. Present relevant performance metrics related to bias and precision using both jackknife and retrospective approaches for each model.
  4. Synthesize and compare performance between models across evaluation approaches and metrics.

Expected Publications

Expected Participation

References Cited

Bonjean, F., and G.S.E. Lagerloef, 2002: Diagnostic Model and Analysis of the Surface Currents in the Tropical Pacific Ocean, Journal of Physical Oceanography, Vol. 32, No. 10, pages 2938-2954.

Ingraham, J., and M.K. Miyahara. 1988. Ocean surface current simulations in the north Pacific ocean and Bering sea (OSCURS-numerical model). U.S. Dep. Commer. NOAA Tech. Memo. NMFS F/NWC-130: 155 p.

Thomson, R., R. Hourston, and S. Tinis. 2013. OSCURS for the 21st Century: Northeast Pacific Salmon Tracking and Research (NEPSTAR) Project, Year 3 Interim Report. Annual report submitted to the Pacific Salmon Commission. 37p. (request copy from csap@dfo-mpo.gc.ca)

Notice

Participation to CSAS peer review meetings is by invitation only.

Date modified: