Modelling in the Marine Environment: An Overview
Marine systems are complex and dynamic, involving interactions and feedbacks among physical, chemical, biological and human components. Understanding and predicting how marine systems function and respond to local and global pressures is essential for sustainable use and management of marine resources and ecosystems. Modelling is a powerful tool that can help achieve this goal by simulating marine system processes and scenarios based on data, theory and assumptions.
What is Modelling?
Modelling is the process of creating a simplified representation of a system or phenomenon using mathematical equations, algorithms and rules. A model can be used to describe, explain, explore or predict the behaviour of the system or phenomenon under different conditions or inputs. Modelling can also be used to test hypotheses, validate observations, support decision-making and communicate results.
There are different types of models depending on the purpose, complexity and scale of the system or phenomenon being modelled. For example, deterministic models produce a single outcome for a given set of inputs, while stochastic models incorporate randomness and uncertainty in the outcomes. Empirical models are based on observed relationships between variables, while mechanistic models are based on underlying physical or biological principles. Statistical models use mathematical techniques to analyse data and infer patterns or trends, while dynamical models use differential equations to describe how the system changes over time or space.
In the marine environment, modelling can be applied to various aspects of marine systems, such as ocean circulation, waves, tides, sediment transport, water quality, biogeochemical cycles, plankton dynamics, fish stocks, marine habitats, biodiversity, ecosystem services, human activities and impacts. Modelling can also be used to integrate different aspects of marine systems into a holistic framework that captures the interactions and feedbacks among them.
Why is Modelling Important?
Modelling is important for advancing scientific knowledge and understanding of marine systems. By creating models that capture the essential features and processes of marine systems, scientists can gain insights into how marine systems work and what factors influence them. Modelling can also help identify gaps in data or knowledge that need further research or observation.
Modelling is also important for informing policy and management of marine systems. By using models to simulate different scenarios or outcomes of marine systems under various pressures or interventions, decision-makers can evaluate the potential impacts and trade-offs of different options or strategies. Modelling can also help assess the effectiveness and uncertainty of different measures or actions for achieving desired goals or objectives.
Modelling is also important for engaging and educating stakeholders and the public about marine systems. By using models to communicate complex or abstract concepts in a visual or interactive way, modellers can enhance awareness and understanding of marine systems among different audiences. Modelling can also help foster dialogue and collaboration among different stakeholders with different perspectives or interests in marine systems.
How is Modelling Done?
Modelling is a iterative and collaborative process that involves several steps:
– Defining the problem or question: The first step is to identify the purpose and scope of the modelling exercise, such as what system or phenomenon to model, what scale or resolution to use, what data or information to include, what outputs or indicators to produce, what scenarios or experiments to run.
– Developing the model: The second step is to design and construct the model using appropriate methods and tools, such as choosing the type of model (e.g., deterministic or stochastic), selecting the variables and parameters (e.g., physical or biological), defining the equations and rules (e.g., empirical or mechanistic), implementing the algorithms and codes (e.g., numerical or analytical).
– Calibrating and validating the model: The third step is to test and evaluate the model performance using available data or observations, such as adjusting the parameters (e.g., sensitivity analysis), comparing the outputs with reality (e.g., goodness-of-fit), assessing the uncertainty and error (e.g., confidence intervals), identifying the limitations and assumptions (e.g., sources of bias).
– Applying and analysing the model: The fourth step is to use the model to address the problem or question by running simulations or experiments under different conditions or inputs, such as exploring the behaviour or response of the system (e.g., sensitivity analysis), predicting the future state or outcome of the system (e.g., projection analysis), comparing different alternatives or options for the system (e.g., scenario analysis).
– Communicating and disseminating the model: The fifth step is to report and share the model results and implications with relevant stakeholders or audiences, such as summarising the main findings and conclusions (e.g., summary report), explaining the methods and assumptions (e.g., technical report), illustrating the outputs and indicators (e.g., graphs or maps), demonstrating the model functionality (e.g., web interface).
What are the Challenges and Opportunities of Modelling?
Modelling is a challenging and rewarding endeavour that requires creativity, skill and expertise. Some of the common challenges of modelling include:
– Data availability and quality: Modelling relies on data or information to build, calibrate, validate and apply models. However, data may be scarce, incomplete, inconsistent, inaccurate or outdated, especially in the marine environment where observations are difficult and costly to obtain. Therefore, modellers need to use data from various sources and methods, such as in situ measurements, remote sensing, laboratory experiments, literature reviews, expert opinions, etc., and apply techniques to deal with data gaps or uncertainties, such as interpolation, extrapolation, imputation, assimilation, etc.
– Model complexity and uncertainty: Modelling involves simplifying and abstracting reality to create a manageable and tractable representation of the system or phenomenon. However, simplification and abstraction may introduce errors or biases in the model structure or parameters that affect the model accuracy or reliability. Therefore, modellers need to balance the trade-off between model complexity and uncertainty, such as choosing the appropriate level of detail or resolution, incorporating the relevant processes or interactions, estimating the uncertainty or error ranges, etc.
– Model interpretation and communication: Modelling produces outputs or indicators that reflect the model assumptions and inputs. However, outputs or indicators may be misinterpreted or misunderstood by different users or audiences who have different expectations or perspectives of the model. Therefore, modellers need to communicate the model results and implications clearly and transparently, such as explaining the model purpose and scope, describing the model methods and limitations, presenting the model outputs and indicators in a meaningful and accessible way, etc.
Modelling also offers many opportunities and benefits for advancing science and society. Some of the potential opportunities of modelling include:
– Model integration and synthesis: Modelling enables integrating and synthesising data or information from different sources or disciplines into a coherent and consistent framework that captures the whole system or phenomenon. This can help reveal new patterns or relationships, generate new hypotheses or questions, identify new data or knowledge needs, etc.
– Model innovation and development: Modelling stimulates innovation and development of new methods and tools for creating and applying models. This can help improve model performance and functionality, expand model capabilities and applications, enhance model usability and accessibility, etc.
– Model collaboration and participation: Modelling facilitates collaboration and participation among different stakeholders or audiences who have different roles or interests in the system or phenomenon. This can help foster dialogue and exchange of ideas, increase trust and confidence in the model results, promote co-learning and co-production of knowledge, etc.
References
– Marine Systems Modelling – Plymouth Marine Laboratory. https://www.pml.ac.uk/science/Marine-Systems-Modelling
– Robinson N.M., Nelson W.A., Costello M.J., Sutherland J.E., Lundquist C.J. (2017) A Systematic Review of Marine-Based Species Distribution Models (SDMs) with Recommendations for Best Practice. Frontiers in Marine Science 4:421. https://doi.org/10.3389/fmars.2017.00421
– Marine Systems Modelling | National Oceanography Centre. https://noc.ac.uk/science/research-areas/marine-systems-modelling
– Periáñez R. (2002) Modelling the Dispersion of Radionuclides in the Marine Environment: An Introduction. Springer Science & Business Media.
