Description: Phytoplankton species composition and the associated community size structure are expected to change with a warming and freshening of the Arctic Ocean. Cell size controls many physiological (bottom-up) processes, such as nutrient uptake, photosynthesis and growth, thereby affecting the functioning of the planktonic ecosystem as a whole. Furthermore, predator-prey interaction (top-down control) is highly size dependent. The size structure of the phytoplankton community in the Fram Strait has been analysed, based on observations of cell abundance and size. Non- parametric size spectra are obtained from microscopic observations, using a statistical approach that also provides respective confidence intervals. A bootstrap approach is applied, with cell counts and size measurements being resampled respectively. Kernel density estimates (KDE) are derived for all resampled data sets. The collection of KDEs yield robust continuous descriptions of cell density versus cell size together with their confidence limits. With this approach we resolve detailed changes in community size structure that shall be used to improve and constrain results of a size-based plankton ecosystem model. Size dependencies of bottom-up and top-down effects on biogeochemical mass flux will be investigated. The calibrated model can then be applied for deriving reliable projections of how the planktonic ecosystem in the Arctic may be affected by climate change.

Description: The overall goal of this work is to investigate the performance of ecosystemmodels and to relate their results to existing observations in the NorthAtlantic. Therefore different data assimilation methods are applied. Avariational adjoint technique and a micro-generic algorithm ($\mu$GA) areutilized to estimate model parameters, such that the misfit between modelresults and observations is minimised. Data assimilation experiments areperformed with nitrogen based ecosystem models, comprising three and fourstate variables (NPZ- and NPZD models): dissolved inorganic nitrogen (N),phytoplankton (P), herbivorous zooplankton (Z) and detritus (D). TheNPZ-model simulates mean concentrations of the different variables withinthe upper mixed layer, while the NPZD-model has a vertically resolved grid.Physical boundary conditions are obtained from three-dimensional simulationsof the ocean's circulation in the North Atlantic, with daily mean atmosphericforcing from ECMWF-reanalysis data.First, data assimilation experiments are conducted with observations fromthe Bermuda Atlantic Time-series Study (BATS) in order to optimise theNPZ-model. While applying the adjoint method different optimal parametersets are obtained when starting from different initial parameter sets. It isshown that for parameter optimisation of an ecosystem model, theapplication of the $\mu$GA is superior to the performance of the adjointmethod.Second, simultaneous assimilation experiment are performed with theNPZD-model using observational data from three locations in the NorthAtlantic: BATS, the site of the North Atlantic Bloom Experiment (NABE) andthe Ocean Weather Ship-India (OWS-INDIA). The simultaneous optimisationyields a best parameter set, which can be utilized for basin wide simulationsin coupled physical-biological (general circulation) models of the NorthAtlantic.The parameter set retrieved from the simultaneous optimisations producessubstantial differences in the biogeochemical fluxes when compared withmodel results using previously published parameters. In contrast to earliermodels the rapid cycling of organic matter for sustaining primary productionis emphasized. Furthermore, systematic discrepancies between$^{14}$C-fixation rates and modelled primary production are identified.It is suggested that carbon based primary productivity may not beadequately represented by ecosystem models when a constant nitrogento carbon conversion factor is assumed.The chosen physical boundary conditions are adequate to simulate thebiogeochemical fluxes at the BATS and NABE sites. At high latitudes(OWS-INDIA), however, the physical-biological interactions in themodel cannot represent the observed chlorophyll distribution in spring.It is suggested that during this period short-termed alterations ofstratification, rapid biological response and deep mixing of phytoplanktonare necessary in order to reproduce chlorophyllconcentrations at depths of 150-200m.

Description: Assimilation experiments with data from the Bermuda Atlantic Time-series Study (BATS, 1989¯1993) were performed with a simple mixed-layer ecosystem model of dissolvedinorganic nitrogen (N), phytoplankton (P) and herbivorous zooplankton (H). Our aim is to optimize the biological model parameters, such that the misfits between model results andobservations are minimized. The utilized assimilation method is the variational adjoint technique, starting from a wide range of first-parameter guesses. A twin experiment displayedtwo kinds of solutions, when Gaussian noise was added to the model-generated data. The expected solution refers to the global minimum of the misfit model-data function, whereasthe other solution is biologically implausible and is associated with a local minimum. Experiments with real data showed either bottom-up or top-down controlled ecosystemdynamics, depending on the deep nutrient availability. To confine the solutions, an additional constraint on zooplankton biomass was added to the optimization procedure. Thisinclusion did not produce optimal model results that were consistent with observations. The modelled zooplankton biomass still exceeded the observations. From the model-datadiscrepancies systematic model errors could be determined, in particular when the chlorophyll concentration started to decline before primary production reached its maximum. Adirect comparision of measured 14C-production data with modelled phytoplankton production rates is inadequate at BATS, at least when a constant carbon to nitrogen C : N ratio isassumed for data assimilation.

Description: An optimization experiment is performed with avertically resolved, nitrogen based ecosystem model, comprising fourstate variables (NPZD-model): dissolved inorganic nitrogen (N),phytoplankton (P), herbivorous zooplankton (Z) and detritus (D).Parameter values of the NPZD-model are optimized by assimilatingobservations at three locations in the North Atlantic simultaneously,namely at the station of the Bermuda Atlantic Time-series Study(BATS; 32N 64W), at the site of the North AtlanticBloom Experiment (NABE; 47N 20W) and at the position of theOcean Weather Ship-India (OWS-INDIA; 59N 19W).A method is described for simultaneous optimization which effectivelymerges different types of observational data at distinct sites in the ocean. Amicro-genetic algorithm is applied for the minimization of a weighted leastsquare misfit function. The simultaneous optimization yields a bestparameter set which can be adopted for basin wide simulations incoupled physical-biological for basin wide simulations in coupledphysical-biological (general circulation) models of the North Atlantic.The optimal parameter estimates are shown to representa compromise among local parameter estimates obtained fromsingle-site optimizations at the individual locations.The optimized parameter set is compared with a set of traditionallypublished values. Drastic changes due to the optimization canbe attributed to the estimates of the initial slope parameter ofthe light limited growth function ($\alpha$) and to the recyclingof organic nitrogen. High estimates of $\alpha$ arewell constrained by chlorophyll observations at the BATS andOWS-INDIA sites, in order to match initial bloom phases.The optimization points towards rapid remineralization processes,expressed in high optimal estimates for the phytoplanktonmortality/excretion rate, which is necessary forthe model to achieve a better agreement with primaryproduction measurements.

Description: This study relates the performance of an optimized one-dimensionalecosystem model to observations at three sites in the North Atlantic Ocean:the Bermuda Atlantic Time series Study (BATS, 32N 64W),the location of the North Atlantic Bloom Experiment (NABE, 47N 20W),and Ocean Weather Ship INDIA (OWS-INDIA, 59N 19W).The ecosystem model resolves dissolved inorganic nitrogen (N),phytoplankton (P), zooplankton (Z) and detritus (D), therefore calledNPZD-model. Physical forcing, such as temperature, eddy diffusivitiesand surface radiation are taken from an eddy-permitting generalcirculation model of the North Atlantic Ocean, covering a period from1989 through 1993. When an optimized parameter set is applied, theNPZD-model produces substantial differences in the biogeochemicalfluxes with respect to model results using previously published``traditional'' parameter values.Annual nitrogen fluxes are determined forthe upper 126 meters together with standard deviations which areassociated with uncertainties in the optimal parameter estimates. Incontrast to earlier models the rapid cycling of organic matter forsustaining primary production is significantly enhanced. An analysis ofprimary production reveals systematic discrepancies between$^{14}$C-fixation rates and modelled primary production under nutrientdepleted conditions.The chosen physical boundary conditions are adequate to simulate thebiogeochemical fluxes at the BATS and NABE sites. At high latitudes(OWS-INDIA), however, the physical-biological interactions in themodel cannot represent the observed chlorophyll distribution in spring.During this period the short-termed alterations of stratification andmixing together with a rapidly following biological responseare insufficiently resolved in order to reproduce chlorophyll concentrationsfound at depths of 150-200m.