UW Aquatic & Fishery Sciences Quantitative Seminar

Nicholas Bond

UW Joint Institute for the Study of the Atmosphere and Ocean (JISAO)

Anticipating Future Ocean Conditions for North Pacific Fisheries

Abstract

Sets of global climate simulations have been carried out for the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report.  A total of 23 different coupled atmosphere-ocean general circulation models were employed under common emission scenarios. A protocol has been designed for using these simulations towards the projection of environmental factors known or suspected to be important to fisheries.  The method relies on critical evaluation of the models’ 20th century hindcast simulations.  The first step has been to determine the degree to which each available model was able to replicate the spatial pattern, temporal scale and magnitude of variance associated with the leading mode of variability in North Pacific SST, i.e., the Pacific Decadal Oscillation (PDO).  The subset of 12 models successful at replicating the PDO were then examined further using a technique representing an adaptation of Bayesian Model Averaging (BMA).  This technique provides weighted ensemble means and estimates of uncertainties in the models’ predictions for individual parameters in specific regions. It has been applied to the transport of larval flatfish in the Bering Sea, and feeding conditions for juvenile salmon along the Pacific Northwest coast, among other examples. As long as the physical environmental controls for a specific population or region are known, and can be forecast with some reliability, the present protocol represents a reasonable way to achieve an early indication of the likely trends in selected populations.  It should be considered as complementary to direct simulations, in which climate scenarios are used to force regional ocean numerical models, which in turn are linked to biological models, i.e., dynamical downscaling.