A multivariate analysis of observed and modeled biophysical variability on the Bering Sea shelf: Multidecadal hindcasts (1970–2009) and forecasts (2010–2040)

Coupled physical/biological models can be used to downscale global climate change to the ecology of subarctic regions, and to explore the bottom-up and top-down effects of that change on the spatial structure of subarctic ecosystems—for example, the relative dominance of large vs. small zooplankton in relation to ice cover. Here we utilize a multivariate statistical approach to extract the emergent properties of a coupled physical/biological hindcast of the Bering Sea for years 1970–2009, which includes multiple episodes of warming and cooling (e.g. the recent cooling of 2005–2009), and a multidecadal regional forecast of the coupled models, driven by an IPCC global model forecast of 2010–2040. Specifically, we employ multivariate empirical orthogonal function (EOF) analysis to derive the spatial covariance among physical and biological timeseries from our simulations. These are compared with EOFs derived from spatially gridded measurements of the region, collected during multiyear field programs. The model replicates observed relationships among temperature and salinity, as well as the observed inverse correlation between temperature and large crustacean zooplankton on the southeastern Bering Sea shelf. Predicted future warming of the shelf is accompanied by a northward shift in both pelagic and benthic biomass.