Theme 1 Modelling: where have we come from and where are we going?
Review of the achievements that have been made during the last ~30 years on which we build our modelling efforts, highlighting areas which need continued development. How can we use a diversity of models and approaches in a combined framework to develop a holistic view of marine ecosystems and exploitation, and the consequences of ecosystem and societal change for the services and benefits that people derive from the oceans?
This introductory session will feature 3 invited speakers:
Further details on these speakers are available on the keynotes page.
Convenors
This introductory session will feature 3 invited speakers:
- Stephanie Dutkiewicz from MIT in the USA
Modelling diverse phytoplankton Communities
The ecosystem modelling group at MIT has worked to incorporate aspects of phytoplankton diversity in numerical models, and have used these models along with theory to understand some of the controls, such as resource supply ratios, grazing, and the role of a moving ocean, on phytoplankton communities. I will provide a synthesis of our trait-based approach to differentiate phytoplankton types, as well as some of the results of how different dimensions of diversity combine to set the seasonal and regional phytoplankton communities. - Kevin Flynn of Swansea University, UK. Kevin's will talk on 'From sublime simplicty to enigmatic models: sharing knowledge and responsibilities'
- Beth Fulton from CSIRO in Australia
Integrated modelling – the challenge is making it relevant
Beth will give an opening talk discussing the increasingly complexity of marine ecosystems and their competing and intense uses. This complexity lends itself to the use of systems approaches, which have become increasingly sophisticated over the last 50 years. Nevertheless, while the science has matured a good deal in this time the eternal challenge remains making models relevant to stakeholders – “but what will it do for me?”. Drawing on integrated modelling experience from around the globe, this overview will highlight some of the high and low points of practical integrated management and where ongoing challenges remain – particularly around distrust, accessibility, communication and scale.
Further details on these speakers are available on the keynotes page.
Convenors
- Hermann Lenhart – Germany
- Icarus Allen – United Kingdom
Theme 2 Making an Impact
Marine ecosystem models have a vital role to play in decision-making, but there are significant challenges surrounding the uptake and use of complex models in support of marine policy, management or operations. One of the key requirements is the need for a stronger link between models of biological, social and economic systems; another challenge is the communication of model outcomes through to policy. This session seeks submissions from across the globe on the practical applications of ecosystem models that have had real impact. This will include traditional ecosystem, hydrodynamic and biogeochemical models as well as integrated models that include biology, social, and economic processes. Contributors are encouraged to address how the managing process was entered, if a benchmarking process was required (potentially involving stakeholders or managers), and report on the fit between managers needs' and what is provided by their model. Case studies that identify success stories and synthesise learning that could allow others to increase their impact are encouraged.
Topics include but are not limited to:
Convenors
Topics include but are not limited to:
- Multiple uses of the marine environment
- Ecosystem based management
- End-to-end ecosystem services
- The blue economy and the circular economy
- Including the human dimension in modelling
- Pollution management and monitoring
- Dredging
- Marine use planning including energy extraction
- Aquaculture
- Climate and climate change mitigation
- Improving collaboration with emerging economies
Convenors
- Kieran Hyder - United Kingdom
- Marie Savina-Rolland - France
- Paul Somerfield - United Kingdom
- Jerry Blackford – United Kingdom
- Beth Fulton - Australia
Theme 3 Building blocks: truth or dare?
As one of the conference's key themes, we welcome abstracts that focus on state-of-the-art developments in “how to model”. Truth refers to the ability of the component models to describe all those facets of biology and ecology that are vital to an empiricist. The dare looks to the necessity of sometimes taking risks to explore system and feedbacks, proposing component models based on limited or no empirical evidence. Further details provided in the session descriptions below.
Session 3.1 Plankton in motion
The vertical distribution of plankton can have profound effects on growth and predation rates, as well as retention and dispersal due to vertical shear in currents. Unlike dissolved nutrients, some plankton can move independently from the physical flows, by regulating their buoyancy, sinking and swimming. Behavioural capabilities, environmental cues (e.g. light, chemistry, temperature, conspecifics, predators), and vertical migration time-scales (e.g. diel, ontogenic) vary among plankton functional groups and life stages, and it can be challenging to characterize these in a modelling context. Horizontal processes such as larval advection, biome contraction/expansion and invasive species spread are also challenging to model and under-represented in current applications. This session welcomes both new applications and reviews on the optimal behavioural strategies of movement as well as the sensitivity of modelled ecosystem dynamics to different parameterizations of planktonic motion. We will consider how coupled physical-behavioural transport shapes biogeochemical cycling and the linkages between vertical movement and horizontal spatial patterns. We encourage development and implementation of novel modelling approaches and computational algorithms (e.g. game theory, optimization, advanced IBM, etc.).
Session 3.2 Model experiments and experimental models
Improving the representation of the physiological mechanisms that regulate the response of marine organisms to changing environmental conditions and multiple stressors is one of the fundamental challenges for modellers. Small scale laboratory experiments are the only tools providing the adequate level of detail to inform model formulations able to capture the variety of the responses exhibited by marine organisms when exposed to environmental stressors (e.g. changes in temperature, pH, nutrient regimes, food availability etc.). On the other hand, models can be used to mechanistically interpret experimental datasets and to assess the importance of laboratory findings at the ecosystem scale. We welcome studies combining experiments and models with the aim to develop new formulations, test experimental hypotheses, and/or contextualise dynamics observed in the laboratory in an ecosystem framework.
Session 3.3 Building the zoo
This session reflects the central role of zooplankton in top down and bottom up control, but extends to all inter-trophic level interactions and end to end systems. The "Grand challenge" in marine ecosystem modelling is to achieve realistic end-to-end 3D models where the entire ecosystem is simulated from the physics all the way up through the foodweb. Zooplankton play a crucial role in the ecosystems as the link between the lower and the higher trophics, and one of the key challenges for end-to-end ecosystem models are representing the proper zooplankton diversity with regards to size, function, and parameterization of key processes like growth, reproduction, stress responses, predation and prey of zooplankton. In this session we will focus on presenting new ideas for improving the modelling with different functional complexity, varying from populations to communities that are influenced by bottom-up (predator) and/or top-down (prey) forcing.
Convenors
Kevin Flynn - United Kingdom
Wendy Gentleman – Canada
Andrew Morozov – United Kingdom
Luca Polimene - United Kingdom
Morten Skogen - Norway
The vertical distribution of plankton can have profound effects on growth and predation rates, as well as retention and dispersal due to vertical shear in currents. Unlike dissolved nutrients, some plankton can move independently from the physical flows, by regulating their buoyancy, sinking and swimming. Behavioural capabilities, environmental cues (e.g. light, chemistry, temperature, conspecifics, predators), and vertical migration time-scales (e.g. diel, ontogenic) vary among plankton functional groups and life stages, and it can be challenging to characterize these in a modelling context. Horizontal processes such as larval advection, biome contraction/expansion and invasive species spread are also challenging to model and under-represented in current applications. This session welcomes both new applications and reviews on the optimal behavioural strategies of movement as well as the sensitivity of modelled ecosystem dynamics to different parameterizations of planktonic motion. We will consider how coupled physical-behavioural transport shapes biogeochemical cycling and the linkages between vertical movement and horizontal spatial patterns. We encourage development and implementation of novel modelling approaches and computational algorithms (e.g. game theory, optimization, advanced IBM, etc.).
Session 3.2 Model experiments and experimental models
Improving the representation of the physiological mechanisms that regulate the response of marine organisms to changing environmental conditions and multiple stressors is one of the fundamental challenges for modellers. Small scale laboratory experiments are the only tools providing the adequate level of detail to inform model formulations able to capture the variety of the responses exhibited by marine organisms when exposed to environmental stressors (e.g. changes in temperature, pH, nutrient regimes, food availability etc.). On the other hand, models can be used to mechanistically interpret experimental datasets and to assess the importance of laboratory findings at the ecosystem scale. We welcome studies combining experiments and models with the aim to develop new formulations, test experimental hypotheses, and/or contextualise dynamics observed in the laboratory in an ecosystem framework.
Session 3.3 Building the zoo
This session reflects the central role of zooplankton in top down and bottom up control, but extends to all inter-trophic level interactions and end to end systems. The "Grand challenge" in marine ecosystem modelling is to achieve realistic end-to-end 3D models where the entire ecosystem is simulated from the physics all the way up through the foodweb. Zooplankton play a crucial role in the ecosystems as the link between the lower and the higher trophics, and one of the key challenges for end-to-end ecosystem models are representing the proper zooplankton diversity with regards to size, function, and parameterization of key processes like growth, reproduction, stress responses, predation and prey of zooplankton. In this session we will focus on presenting new ideas for improving the modelling with different functional complexity, varying from populations to communities that are influenced by bottom-up (predator) and/or top-down (prey) forcing.
Convenors
Kevin Flynn - United Kingdom
Wendy Gentleman – Canada
Andrew Morozov – United Kingdom
Luca Polimene - United Kingdom
Morten Skogen - Norway
Theme 4 Data, data everywhere, yet no time to think
Session 4.1 Novel use of observations in models
In this session we encourage contributions that use challenging data sets to inform models and/or models that include detailed process representation to make use of esoteric data sets.
A wide range of observations (including non-parametric) are available to inform biogeochemical and ecological models, however typically only the fraction of these that characterise ‘standard’ biogeochemical or ecological properties are used by modellers. Observations of specific or unusual analogues (e.g. pulsed fluorescence, ETS, bacterial characteristics), diverse data types (e.g. species, microbial or genetics), non-parametric data (e.g. presence of mixotrophs), and spatial and temporal scales (e.g. glider data) not typically represented by standard models and scales precludes direct reconciliation with some models. In this session we encourage contributions that use challenging data sets to inform models and/or models that include detailed process representation to make use of esoteric data sets. Motivation for the topic is to demonstrable enhancements to model performance and associated social license to use models for applied science and management.
Session 4.2 Data assimilation
Understanding and management of marine ecosystems can be advanced by integrating marine models with the last generation of biogeochemical observations, by using Data Assimilation (DA) techniques. DA aims to improve the simulation of ecosystems by accounting for errors in models and data. Applications advancing marine ecosystem understanding (e.g. biogeochemical fluxes, food-web dynamics), operational oceanography (quasi-real-time predictions and reanalysis simulations), supporting societal needs (e.g. aquaculture, marine policy) at a variety of time and spatial scales (from daily to interannual to climate scale simulations in estuary, shelf or global scales) are all welcome in this session.
Session 4.3 Statistics, bioinformatics and machine learning
With increasing volumes and varieties of information flowing from our model systems, we require smart automated methods to discover and display the patterns buried in our data and assess their validity. Consequently there has been a resurgence of interest in machine learning including techniques such as Bayesian analysis, automated neural nets, genetic algorithms parameter estimation, similarity matching and so on. This session welcomes all aspects of machine learning applied to marine data, including both methodologies that could revolutionise our analysis and novel understanding that has arisen from applications specific applications.
Convenors
Stefano Ciavatta - United Kingdom
Gianpiero Cossarini - Italy
Mike St. John - Denmark
Karen Wild Allen - Australia
In this session we encourage contributions that use challenging data sets to inform models and/or models that include detailed process representation to make use of esoteric data sets.
A wide range of observations (including non-parametric) are available to inform biogeochemical and ecological models, however typically only the fraction of these that characterise ‘standard’ biogeochemical or ecological properties are used by modellers. Observations of specific or unusual analogues (e.g. pulsed fluorescence, ETS, bacterial characteristics), diverse data types (e.g. species, microbial or genetics), non-parametric data (e.g. presence of mixotrophs), and spatial and temporal scales (e.g. glider data) not typically represented by standard models and scales precludes direct reconciliation with some models. In this session we encourage contributions that use challenging data sets to inform models and/or models that include detailed process representation to make use of esoteric data sets. Motivation for the topic is to demonstrable enhancements to model performance and associated social license to use models for applied science and management.
Session 4.2 Data assimilation
Understanding and management of marine ecosystems can be advanced by integrating marine models with the last generation of biogeochemical observations, by using Data Assimilation (DA) techniques. DA aims to improve the simulation of ecosystems by accounting for errors in models and data. Applications advancing marine ecosystem understanding (e.g. biogeochemical fluxes, food-web dynamics), operational oceanography (quasi-real-time predictions and reanalysis simulations), supporting societal needs (e.g. aquaculture, marine policy) at a variety of time and spatial scales (from daily to interannual to climate scale simulations in estuary, shelf or global scales) are all welcome in this session.
Session 4.3 Statistics, bioinformatics and machine learning
With increasing volumes and varieties of information flowing from our model systems, we require smart automated methods to discover and display the patterns buried in our data and assess their validity. Consequently there has been a resurgence of interest in machine learning including techniques such as Bayesian analysis, automated neural nets, genetic algorithms parameter estimation, similarity matching and so on. This session welcomes all aspects of machine learning applied to marine data, including both methodologies that could revolutionise our analysis and novel understanding that has arisen from applications specific applications.
Convenors
Stefano Ciavatta - United Kingdom
Gianpiero Cossarini - Italy
Mike St. John - Denmark
Karen Wild Allen - Australia
Theme 5 Wild Card
In order not to exclude any interesting novel work we are convening a wild card session. If you cannot see an obvious place for your work within the current themes please submit an abstract under the Wild Card theme. We reserve the right to move such presentations into one of the other sessions if we consider that that is most appropriate.
Convenors will be confirmed once abstract submitted.