Jacox, M.G., M. A. Alexander, S. Siedlecki, K. Chen, Y.-O. Kwon, S. Brodie, I. Ortiz, D. Tommasi, M. J. Widlansky, D. Barrie, A. Capotondi, W. Cheng, E. Di Lorenzo, C. Edwards, J. Fiechter, P. Fratantoni, E. L. Hazen, A. J. Hermann, A. Kumar, A. J. Miller, D. Pirhalla, M. Pozo Buil, S. Ray, S. C. Sheridan, A. Subramanian, P. Thompson, L. Thorne, H. Annamalai, S. J. Bograd, R. B. Griffis, H. Kim, A. Mariotti, M. Merrifield and R. Rykaczewski, 2019:

Seasonal-to-interannual prediction of U.S. coastal marine ecosystems: Forecast methods, mechanisms of predictability, and priority developments.


Progress in Oceanography, sub judice.

Abstract. Marine ecosystem forecasting is an area of active research and rapid development. Promise has been shown for skillful prediction of physical, biogeochemical, and ecological variables on a range of timescales, suggesting potential for forecasts to aid in the management of living marine resources and coastal communities. However, the mechanisms underlying forecast skill in marine ecosystems are often poorly understood, and many forecasts, especially for biological variables, rely on empirical statistical relationships developed from historical observations. Here, we review statistical and dynamical marine ecosystem forecasting methods and highlight examples of their application along U.S. coastlines for seasonal-to-interannual (1-24 month) prediction of properties ranging from coastal sea level to marine top predator distributions. We then describe known mechanisms governing marine ecosystem predictability in these regions and how they have been used in forecasts to date. These mechanisms include physical atmospheric and oceanic processes, biogeochemical and ecological responses to physical forcing, and intrinsic characteristics of species themselves. In reviewing the state of the knowledge on forecasting techniques and mechanisms underlying predictability in U.S. marine ecosystems, we aim to facilitate forecast development and uptake by (i) identifying methods and processes that can be exploited for development of skillful regional forecasts, (ii) informing priorities for forecast development and validation, and (iii) improving understanding of conditional forecast skill (i.e., a priori knowledge of whether a forecast is likely to be skillful).

Preprint (pdf)