Ocean biogeochemical (BGC) models are, in addition to measurements, the primary tools for investigating ocean biogeochemistry, marine ecosystem functioning, and the global carbon cycle. These models contain a large number of not precisely known parameters and are highly uncertain regarding those parametrizations. The values of these parameters depend on the physical and biogeochemical context, but in practice values derived from limited field measurements or laboratory experiments are used in the model keeping them constant in space and time. This study aims to estimate spatially and temporally varying parameters in a global ocean BGC model and to assess the effect of those estimated parameters on model fields and dynamics. Utilizing the BGC model Regulated Ecosystem Model 2 (REcoM2), we estimate ten selected BGC parameters with heterogeneity in parameter values both across space and over time using an ensemble data assimilation technique. We assimilate satellite ocean color and BGC-ARGO data using an ensemble Kalman filter provided by the Parallel Data Assimilation Framework (PDAF) to simultaneously estimate the BGC model states and parameters. We assess the improvement in the model predictions with space and time-dependent parameters in reference to the simulation with globally constant parameters against both assimilative and independent data. We quantify the spatiotemporal uncertainties regarding the parameter estimation and the prediction uncertainties induced by those parameters. We study the effect of estimated parameters on the biogeochemical fields and dynamics to get deeper insights into modeling processes and discuss insights from spatially and temporally varying parameters beyond parameter values.
