Assimilation of Ocean‐Color Plankton Functional Types to Improve Marine Ecosystem Simulations

Abstract

We assimilated phytoplankton functional types (PFTs) derived from ocean color into a marine ecosystem model, to improve the simulation of biogeochemical indicators and emerging properties in a shelf sea. Error‐characterized chlorophyll concentrations of four PFTs (diatoms, dinoflagellates, nanoplankton, and picoplankton), as well as total chlorophyll for comparison, were assimilated into a physical‐biogeochemical model of the North East Atlantic, applying a localized Ensemble Kalman filter. The reanalysis simulations spanned the years 1998–2003. The skill of the reference and reanalysis simulations in estimating ocean color and in situ biogeochemical data were compared by using robust statistics. The reanalysis outperformed both the reference and the assimilation of total chlorophyll in estimating the ocean‐color PFTs (except nanoplankton), as well as the not‐assimilated total chlorophyll, leading the model to simulate better the plankton community structure. Crucially, the reanalysis improved the estimates of not‐assimilated in situ data of PFTs, as well as of phosphate and pCO2, impacting the simulation of the air‐sea carbon flux. However, the reanalysis increased further the model overestimation of nitrate, in spite of increases in plankton nitrate uptake. The method proposed here is easily adaptable for use with other ecosystem models that simulate PFTs, for, e.g., reanalysis of carbon fluxes in the global ocean and for operational forecasts of biogeochemical indicators in shelf‐sea ecosystems.

The Assimilation of Phytoplankton Functional Types for Operational Forecasting in the Northwest European Shelf

“…Second, these innovations were used by NEMOVAR to create a set of surface total (chlorophyll) increments, similarly to the DA of sea ice concentration described by Waters et al (). The diagonal elements of the background error covariance matrix were a monthly climatology of log‐transformed error variances obtained from the 100‐member EnKF POLCOMS‐ERSEM reanalysis of Ciavatta et al (). These variances were regularized and smoothed using the moving averages algorithm and rescaled to the range 0.02–1.5 (mg/m 3 ), so that the average ratio of background error to obervation error was similar to that calculated in the region when assimilating OC‐CCI data into NEMO‐HadOCC (Ford & Barciela, ).…”

“…This issue can be avoided by directly assimilating PFTs chlorophyll when the PFTs chlorophyll a data are available. Such an approach was taken in an early 1‐D study by Xiao and Friedrichs () and recently by Ciavatta et al () in a 3‐D model configuration of the NWE Shelf.…”

Ecoregions in the Mediterranean Sea Through the Reanalysis of Phytoplankton Functional Types and Carbon Fluxes

“…We used the EnKF with an ensemble size of N = 100 members. We included in the analyzed state vector 35 out of the 51 ERSEM pelagic state variables, to keep the analysis computationally affordable, in analogy with Ciavatta et al (). Coherently with the paper's objective to reanalyze the plankton community structure, we included in the state vector all the 27 variables related to the phytoplankton, zooplankton, and bacteria functional types, as well as eight variables representing DIC, labile, and refractory dissolved organic matter and small particulate matter.…”

“…Daily values of the variables in the reanalysis data set ( y ) were matched up point to point in space and time with the data ( o ). In particular, we computed and then showed in robust skill diagrams (Butenschön et al, ; Ciavatta et al, ) (a) the bias calculated as the median value of the reanalysis‐to‐observation mismatch, bias* = median( y − o ), normalized by the interquartile range of the data (IQR o ); (b) the unbiased median absolute error, MAE′ = median{abs[ y − o − bias*]}, normalized by IQR o , and taken with the algebraic sign of the differences between the interquartile range of the output and the data, sign (IQR − IQR o ); and (c) the correlation coefficient. In these diagrams, the closer the point is to the center, the lower the error of the simulation (i.e., the median bias and the unbiased absolute error).…”

“…State‐of‐the‐art ecosystem models can describe PFTs and the related biogeochemical and trophic fluxes, but the skill of model simulations is often poor, mainly due to the difficulty of formulating and parameterizing PFT processes (Shimoda & Arhonditsis, ). Merging PFT ocean color and ecosystem models through data assimilation was proposed as a novel approach to estimate better the phytoplankton community structure and related fluxes in marine ecosystems (Ciavatta et al, ). Better simulations of diatoms, dinoflagellates, NNP, and PIC biomass, as well as estimates of related biogeochemical fluxes, were delivered in both a reanalysis simulation (Ciavatta et al, ) and (pre)operational predictions (Skákala et al, ) in the North West European Shelf Sea ecosystem, through the assimilation of a regional ocean color PFT product (Brewin et al, ).…”