Andreas Moll

End-To-End Models for the Analysis of Marine Ecosystems: Challenges, Issues, and Next Steps

Abstract There is growing interest in models of marine ecosystems that deal with the effects of climate change through the higher trophic levels. Such end-to-end models combine physicochemical oceanographic descriptors and organisms ranging from microbes to higher-trophic-level (HTL) organisms, including humans, in a single modeling framework. The demand for such approaches arises from the need for quantitative tools for ecosystem-based management, particularly models that can deal with bottom-up and top-down controls that operate simultaneously and vary in time and space and that are capable of handling the multiple impacts expected under climate change. End-to-end models are now feasible because of improvements in the component submodels and the availability of sufficient computing power. We discuss nine issues related to the development of end-to-end models. These issues relate to formulation of the zooplankton submodel, melding of multiple temporal and spatial scales, acclimation and adaptation, behavioral movement, software and technology, model coupling, skill assessment, and interdisciplinary challenges. We urge restraint in using end-to-end models in a true forecasting mode until we know more about their performance. End-to-end models will challenge the available data and our ability to analyze and interpret complicated models that generate complex behavior. End-to-end modeling is in its early developmental stages and thus presents an opportunity to establish an open-access, community-based approach supported by a suite of true interdisciplinary efforts.

Estimation of the variability of production by simulating annual cycles of phytoplankton in the central North Sea

Abstract A physical and a biological one-dimensional upper layer model for the stimualtion of the annual cycles of both the physical and the phytoplankton dynamics, are used to estimate the annual primary production in the central North Sea. The simulations are driven with actual 3-hourly meteorological standard observations and estimated radiation data for the 25 years 1962 to 1986. The high variability of the forcing generates a considerable variability in the physical and biological oceanic mixed layer dynamics. As an example, the model results from two years with contrasting meteorological conditions, 1963 and 1967, are discussed in detail. The mixing regimes generated are very different which result in different annual phytoplankton cycles. During 1963 when conditions were warm and windless, the early establishment of a calm upper layer water mass enabled a strong spring plankton bloom; whereas in 1967, which was stormy and cold, convective overturning continued until April, suppressing an early spring bloom and prolonging the blooming into summer. For the meteorological conditions observed in 1962 to 1986, the simulations yield an integrated annual water column gross production of 83.5–99.0 gC m−2a−1 and an integrated annual water column net production ranging between 43.0 and 64.2 gC m−2a−1 for the central North Sea. Grazing by the prescribed copepod population ranges from 24.5 to 40.0 gC m−2a−1. The production events are described irregularly over the different years, total gross production varies only about 17%, and total net production by about 21%. The nutrient taken up by the algae is 2.6 to 3.2 times the winter concentration of that layer which in summer is situated above the seasonal thermocline. The additional nutrient is provided by local regeneration and by turbulent entrainment from below the thermocline. Local regeneration in the upper layer provides about 2.4 and 0.3 times the entrained amount of phosphate during spring and summer, respectively. In the 25 years 16 late summer or early fall storm events entrained more than 1.2mmol P m−2d−1 into the depleted upper layer, potentially initiating new production events. The simulated annual cycles can be validated with the available data only in the sense that the variability, but not single events, can be compared to measurements. Such comparisons between simulated and field data show that the simulation reproduces the general features of annual phytoplankton cycles. This establishes confidence in those calculated estimates, for which field data are not directly comparable. It is concluded that weather-induced variability can explain most of the observed variability in phytoplankton in annual cycles. A typical annual cycle of phytoplankton biomass dynamics is presented. Ratios of daily process contributions show that the balances between the different processes change during the annual cycle. Diagrams of the mean and seasonal phosphorus flow are derived from the simulations. Two thirds of the primary production are channelled through the copepods, and one third is lost by other processes. Organic matter corresponding to more than the initial amount of nutrients in the mixed layer is sedimenting out of the upper layer, and about the same amount is regenerated at the bottom and mixed into the water column at the end of the year. The critical points in the model: grazing, recycling of nutrients and mixing in the bottom boundary layer, are discussed. The model still needs to be refined with respect to these processes in order to achieve the delicate balances required to generate fall blooms. A series problem is the appropriateness of primary production measurements for a comparison with simulated quantities. Attempts should be made to establish a one-to-one correspondence between model-derived production quantities and measurements. Single events are important, so both sampling strategies and the estimation of fluxes from data should take account of the possible occurrence of such events, which may have been missed in the observations, by presenting ranges covering the realistic variance rather than mean values.

Regional distribution of primary production in the North Sea simulated by a three-dimensional model

Abstract A biological one-dimensional water column model for the simulation of the annual cycles of the phytoplankton dynamics and a physical transport model are coupled into a three-dimensional primary production model to estimate the annual primary production of the North Sea and its regional differences. The simulations are driven with actual forcing, taking into account monthly river loads from 14 rivers, daily velocities and diffusivities from a baroclinic North Sea model, and solar radiation calculated every 30 min. The high variability of the forcing generates considerable variability in the physical and biological dynamics. The simulated annual cycles of chlorophyll, phosphate and daily net primary production are validated with available data from different years in the sense that the simulated single year can be compared to the measured variability. Such comparisons show that the simulation reproduces the general features of annual phytoplankton cycles in terms of chlorophyll and the triggering nutrient phosphate. This establishes confidence in the calculated annual primary production of the North Sea area. The simulation of 1986 yields an integrated annual water column net production ranging between 92 and 345 g C m−2 yr−1 for different regions. The simulated annual production agrees quite well with the general quantitative knowledge of the observed yearly production, except for the British coast, where the production is overestimated due to lacking inorganic suspended matter attenuation. The mean net primary production of the North Sea is 145 g C m−2 yr−1.

Interannual variability of the North Sea primary production: comparison from two model studies

Abstract The North Sea is known to be a very productive area. Several models and in situ measurements have been used to quantify the primary production and its spatial variability, but large uncertainties still exist. Except for general statements about the level of production, very little is known about the interannual variability of it. In this paper two state of the art ecological models have been run with realistic forcing for 10 different years to investigate this interannual variability. The focus has been on differences due to changing wind fields, and differences due to changes in river discharges. Separated into ERSEM boxes, both spatial and temporal variability is discussed. The models suggest that the interannual variability in the North Sea primary production is around 15%, and that the variability locally is higher than this, thus an increase in one area is often compensated by a decrease somewhere else. The impact of the river nutrient inputs is estimated to be less than 10% of the total production. The modelled variation in primary production, can in both models be related to, and explained from, actual variability found in the modeled physics. The large variability in primary production found due to these changes in the underlying physical fields, suggests that a realistic three-dimensional circulation model is essential for this kind of modelling.

Combined analysis of field and model data: A case study of the phosphate dynamics in the German Bight in summer 1994

The intention of this paper is to analyse a specific phenomenon observed during the KUSTOS campaigns in order to demonstrate the general capability of the KUSTOS and TRANSWATT approach, i.e. the combination of field and modelling activities in an interdisciplinary framework. The selected phenomenon is the increase in phosphate concentrations off the peninsula of Eiderstedt on the North Frisian coast sampled during four subsequent station grids of the KUSTOS summer campaign in 1994. First of all, a characterisation of the observed summer situation is given. The phosphate increase is described in detail in relation to the dynamics of other nutrients. In a second step, a first-order estimate of the dispersion of phosphate is discussed. The estimate is based on the box model approach and will focus on the effects of the river Elbe and Wadden Sea inputs on phosphate dynamics. Thirdly, a fully three-dimensional model system is presented, which was implemented in order to analyse the phosphate development. The model system is discussed briefly, with emphasis on phosphorus-related processes. The reliability of one of the model components, i.e. the hydrodynamical model, is demonstrated by means of a comparison of model results with observed current data. Thereafter, results of the German Bight seston model are employed to interpret the observed phosphate increase. From this combined analysis, it was possible to conclude that the phosphate increase during the first three surveys was due to internal transformation processes within the phosphorus cycle. On the other hand, the higher phosphate concentrations measured in the last station grid survey were caused by a horizontal transport of phosphate being remobilised in the Wadden Sea.

Modelling Water Column Processes in the North Sea [and Discussion]

In the North Sea advective transports are not negligible. Nevertheless, physical properties like sea surface temperature (SST) can be hindcasted with sufficient precision by vertical process water column models. Annual cycles of SST in the southern, central, and northern North Sea can be simulated using physical upper layer models with relatively small RMS errors. For the Fladenground Experiment (FLEX’76) in the northern North Sea the RMS error is less 0.3 °C for the 2 months of the experiment. This justifies the initial use, at least, of vertical process water column models in simulations for investigating transfer processes in the planktonic ecosystem. Experiments have shown that the simulated entrainment velocities at the bottom of the mixed layer during summer are critically dependent on the resolution of the forcing variables. The effects of this resolution on the annual phytoplankton dynamics will be discussed. Phytoplankton dynamics are strongly influenced by those of the zooplankton, and vice versa. Several field investigations have shown that, seemingly, phytoplankton cannot sustain the observed stock of zooplankton in the northern North Sea: there exists a gap between the abundance of phytoplankton and the need for it to maintain the zooplankton. Revisiting FLEX’76, the simulations with water column models of increasing complexity concerning detritus suggest that pelagic detritus can fill the gap in food availability for the zooplankton. If it is assumed that the zooplankton feeds also on detritus, the zooplankton experiences no food shortage.

Modelling water column processes in the North Sea

In the North Sea advective transports are not negligible. Nevertheless, physical properties like sea surface temperature (sst) can be hindcasted with sufficient precision by vertical process water column models. Annual cycles of sst in the southern, central, and northern North Sea can be simulated using physical upper layer models with relatively small rms errors. For the Fladenground Experiment (FLEX’76) in the northern North Sea the rms error is less 0.3 °C for the 2 months of the experiment. This justifies the initial use, at least, of vertical process water column models in simulations for investigating transfer processes in the planktonic ecosystem. Experiments have shown that the simulated entrainment velocities at the bottom of the mixed layer during summer are critically dependent on the resolution of the forcing variables. The effects of this resolution on the annual phytoplankton dynamics will be discussed.

Advective contributions to the heat balance of the german bight (LV Elbe 1) and the central north sea (OWS Famita)

The mean annual cycle of the net energy flux through the sea surface and of the heat storage are investigated in detail using observations of the Light Vessel LV Elbe 1 for the period 1962-1986 in the German Bight and at Ocean Weather Ship OWS Famita for the period 1965-1978 in the central North Sea. The investigation confirms the general geographical picture of the heat budget of the German Bight that shows a net loss to the atmosphere by a long-term mean of -15 W m-2. The radiative surface input of 113 W m-2 is balanced by -62 W m-2 net back radiation, -51 W m-2 of latent heat flux and -15 W m-2 of sensible heat flux. The heat advection calculated as the residual of the heat storage rate and surface energy balance is 16 W m-2. The mean annual cycles of heat storage and surface energy balance are nearly equal, and the temperature variations are mainly driven by local heat input. The small differences build up the annual advection cycle. Warm water advection occurs from October to April and cold water advection in summer from May to September. The seasonal advection variability is extreme in winter and summer, and the ranges slow down in spring and autumn, when the sign of the heat balance changes. The OWS Famita is situated also in an area of net energy loss to the atmosphere, showing a long-term annual mean loss of -16 W m-2. The surface radiation input of 105 W m-2 is mainly balanced by outgoing long wave back radiation of -60 W m-2 and a latent heat flux of -49 W m-2. A minor contribution to the heat balance is the sensible heat flux of -12 W m-2. Warm water advection occurs in winter and spring. Variability is greater than at LV Elbe 1.

A NMR-spectra-based scoring function for protein docking

A well studied problem in the area of Computational Molecular Biology is the so-called Protein-Protein Docking problem (PPD) that can be formulated as follows: Given two proteins A and B that form a protein complex, compute the 3D-structure of the protein complex AB. Protein docking algorithms can be used to study the driving forces and reaction mechanisms of docking processes. They are also able to speed up the lenghty process of experimental structure elucidation of protein complexes by proposing potential structures. In this paper, we are discussing a variant of the PPD-problem where the input consists of the tertiary structures of A and B plus an unassigned 1H-NMR spectrum of the complex AB. We present a new scoring function for evaluating and ranking potential complex structures produced by a docking algorithm. The scoring function computes a “theoretical” 1H-NMR spectrum for each tentative complex structure and subtracts the calculated spectrum from the experimental spectrum. The absolute areas of the difference spectra are then used to rank the potential complex structures. In contrast to formerly published approaches (e.g. Morelli et. al. [38]) we do not use distance constraints (intermolecular NOE constraints). We have tested the approach with the bound conformations of four protein complexes whose three-dimensional structures are stored in the PDB data bank [5] and whose 1H-NMR shift assignments are available from the BMRB database (BioMagResBank [47]). In all examples, the new scoring function produced very good rankings of the structures. The best result was obtained for an example, where all standard scoring functions failed completely. Here, our new scoring function achieved an almost perfect separation between good approximations of the true complex structure and false positives. Unfortunately, the number of complexes with known structure and available spectra is very small. Nevertheless, these experiments indicate that scoring functions based on comparisons of one- or multi-dimensional NMR spectra might be a good instrument to improve the reliability and accuracy of docking predictions and perhaps also of protein structure predictions (threading).

Variabilität der Primärproduktion aus dreidimensionalen Modellrechnungen für die Nordsee mit ECOHAM1

Mit dem dreidimensionalen Modell ECOHAM1 (Ecological North Sea Model, Hamburg, Version 1) wurde die jahrliche Primarproduktion fur die Jahre 1985 bis 1994 regional mit einem 20 km Gitterabstand quantifiziert, um die jahrliche Variabilitat zu untersuchen. Die Phytoplanktonproduktion wird durch das Lichtangebot, den steuernden Nahrstoff Phosphat und den Zooplanktonfras begrenzt. Fur die Simulation wurden aktuelle Antriebe der Jahre 1985–1994 verwendet, um den Einfluss von kurzzeitigen physikalischen Effekten bei der Primarproduktion zu erfassen. Die vom Modell berechnete mittlere Jahresproduktion betragt 124 gC m−2 y−1. Die jahrliche Variabilitat liegt bei 15% und wird durch die Verteilung fur zwei extreme Jahre gezeigt.