The Centre for Science Advice Pacific (CSAP) Regional Advisory Process (RAP) meeting was held on March 9, 2010 at the Pacific Biological Station, Nanaimo, B.C. One working paper was reviewed.
A draft research document outlining the methods and results used to generate the 2010 Fraser River Sockeye forecast was presented to Regional Advisory Process participants on 9 March 2010. The document represents an extension of the Scientific Advisory Report (SAR) tabled in November in 2009, in which 2010 Fraser Sockeye forecasts were also presented. The intent of the research document was to allow for the review and approval of forecasting models not included in the 2006 Pacific Scientific Advice Review Committee (PSARC) approved forecasting methodology. The models used in the 2010 forecast for Fraser Sockeye were developed to incorporate recent productivity trends into the forecasting process. Three separate forecasts for all Fraser Sockeye forecasted stocks (found in three forecast tables) were generated and evaluated for the 2010 Fraser River Sockeye Salmon forecast:
There were two reviewers of the working paper, one internal and one external. Reviewers concluded that the authors conducted a thorough analysis for forecasting the 2010 adult returns of Fraser River sockeye salmon. The resulting wide probability distributions of the 2010 Fraser Sockeye forecasts and uncertainty regarding future survival represented by three different forecast tables reflect the reality on the west coast of North America generally; namely that forecasts are very uncertain due to many poorly understood processes affecting survival rates. There was recognition and appreciation that the authors applied some new methods to their forecasting work (e.g., Bayesian inference for estimating uncertainties in forecasts and a Kalman filter to estimate time-varying productivity). Most importantly, in some of the 2010 forecast analyses, the author’s recognized the substantial decrease in productivity (adult recruits-per-effective-female spawner, or R/EFS) of most of the Fraser River sockeye stocks over the last 20 to 50 years. To reflect this time trend, they developed some new non-parametric models and applied the Kalman filter version of a Ricker model (with a time-varying 'a' parameter). The author’s also conducted appropriate sensitivity analyses for various assumptions about changes in productivity.
Discussion focused on the issues of recent productivity trends and the effects of model selection and evaluation. Considerable discussion was given to the issues of uncertainty in forecasts and potential approaches to reduce uncertainty in the Fraser Sockeye forecast. Apart from recommendations for future work to explore alternate environmental indicators and sibling (‘jack’) models to be used as informative priors to the forecasts it was concluded that significant improvements are unlikely to reduce uncertainty in the forecasts.
The group commended the authors on the high quality of the work and accepted the paper subject to minor revisions.
View complete PDF document
(15 pages; 92K)
This document is available in PDF format. If the following document is not accessible to you, please contact the Secretariat to obtain another appropriate format, such as regular print, large print, Braille or audio version.
This report uses scientific and technical terms and is published in the official language of the working group or scientific expert that produced the document. If this document is not accessible to you in the official language of your choice, please contact the Secretariat.