Impact of assimilating physical oceanographic data on modeled ecosystem dynamics in the California Current System

Abstract

A method to model ocean ecosystems using data-constrained physical circulation estimates is investigated. Physical oceanographic data is assimilated into a Regional Ocean Modeling System implementation of the California Current System using an incremental 4-Dimensional Variational method. The resulting state estimate drives a complex, self-assembling ocean ecosystem model for the year 2003, and results are evaluated against SeaWiFS surface and CalCOFI subsurface observations and with ecosystem model output driven by an unconstrained physical model. While physical data assimilation improves correlation with observations, this method also drives elevated levels of phytoplankton standing stock, leading to a large bias particularly in regions of low mean concentration. We identify two causes for this increase: biological rectification of fluctuating vertical nutrient transport due to gravity wave generation at assimilation cycle initialization and increased nutrient variance on density surfaces. We investigate one and propose other possible remedies for these deleterious side-effects of this data assimilation method.