ALBERT

Description: Mesocosm experiments on phytoplankton dynamics under high CO2 concentrations mimic the response of marine primary producers to future ocean acidification. However, potential acidification effects can be hindered by the high standard deviation typically found in the distribution of the replicates exposed to the same treatment. In experiments with multiple unresolved factors and a suboptimal number of replicates, post-processing statistical inference tools may fail to detect an effect. In such cases, model-based data analyses are suitable tools to unearth potential responses to the treatment and to identify which uncertainties may give rise to the observed divergences. As test cases, we use data showing high variability from two independent mesocosm experiments, where, according to statistical inference tools, biomass appeared insensitive to changing CO2 conditions. Our simulations, in stead, show earlier and more intense phytoplankton blooms in modeled replicates at high CO2 concentrations and suggest that uncertainties in average cell size, phytoplankton biomass losses and initial nutrient concentration potentially outweigh acidification effects by triggering strong variability during the bloom phase. We also estimate the thresholds below which uncertainties do not escalate into high variability. This information may help to interpret controversial results about acidification and to design future mesocosm experiments.

Description: A series of studies were conducted during the last two decades to investigate effects of ocean acidification (OA) on phytoplankton physiology, plankton ecology, and biogeochemical dynamics of marine ecosystems. Among those studies are experiments with tanks or bags called mesocosms, with some enclosed water volume that typically comprised a natural plankton community found in the surrounding environment. The Pelagic Ecosystem CO2 – Enrichment Study PeECE-I experiment is one such study, where mesocosms were perturbed and exposed to different carbon dioxide (CO2) concentrations to determine responses in growth dynamics of the coccolithophorid Emiliania huxleyi, a marine calcifying algae. The data from replicate mesocosms of PeECE-I show some natural variability and significant differences were revealed in the accumulation of particulate inorganic carbon (PIC) between mesocosms of similar CO2 treatments. In our study we reanalyse PeECE-I data and apply an optimality-based model approach to understand most of the variability observed, with major focus on total alkalinity (TA) and calcification. We explore how much of the observed variability in data can be explained by variations of initial conditions and by the effect of CO2 perturbations. According to our model approach, changes in cellular calcite formation are resolved at the organism-level in response to variations in CO2. With a data assimilation (DA) method we obtain three distinctive ensembles of model solutions, with low, medium and high calcification rates. Optimised values of initial conditions turned out to be correlated with estimates physiological model parameters. The spread of ensemble model solutions captures most of the observed variability, corresponding to the combinations of parameter estimates. Optimised model solutions of the high CO2 treatment are shown to systematically overestimate observed PIC production. Thus, the simulated CO2 effect on calcification is likely too weak. At the same time our model results yield large differences in optimal mass flux estimates of carbon and of nitrogen even between mesocosms exposed to similar CO2 conditions.

Description: To describe the underlying processes involved in oceanic plankton dynamics is crucial for the determination of energy and mass flux through an ecosystem and for the estimation of biogeochemical element cycling. Many planktonic ecosystem models were developed to resolve major processes so that flux estimates can be derived from numerical simulations. These results depend on the type and number of parameterisations incorporated as model equations. Furthermore, the values assigned to respective parameters specify a model's solution. Representative model results are those that can explain data, therefore data assimilation methods are utilised to yield optimal estimates of parameter values while fitting model results to match data. Central difficulties are 1) planktonic ecosystem models are imperfect and 2) data are often too sparse to constrain all model parameters. In this review we explore how problems in parameter identification are approached in marine planktonic ecosystem modelling. We provide background information about model uncertainties and estimation methods, and how these are considered for assessing misfits between observations and model results. We explain differences in evaluating uncertainties in parameter estimation, thereby also addressing issues of parameter identifiability. Aspects of model complexity will be covered and we describe how results from cross-validation studies provide much insight in this respect. Moreover, we elucidate inferences made in studies that allowed for variations in space and time of parameter values. The usage of dynamical and statistical emulator approaches will be briefly explained, discussing their advantage for parameter optimisations of large-scale biogeochemical models. Our survey extends to studies that approached parameter identification in global biogeochemical modelling. Parameter estimation results will exemplify some of the advantages and remaining problems in optimising global biogeochemical models. Our review discloses many facets of parameter identification, as we found many commonalities between the objectives of different approaches, but scientific insight differed between studies. To learn more from results of planktonic ecosystem models we recommend finding a good balance in the level of sophistication between mechanistic modelling and statistical data assimilation treatment for parameter estimation.