Course overview
As the impacts of climate and environmental change become increasingly apparent, the need to better understand the likely impact of these changes, as well as the relative effectiveness of potential mitigation strategies, is paramount. As these changes occur in the future under conditions that have not been experienced previously, we need to rely on models to make these assessments under considerable amounts of uncertainty. Consequently, it is essential to understand different modelling approaches, how relevant models are developed, what potential pitfalls are, how to deal with uncertainty, and how to best use models for assessing the impact of climate and environmental change and identify the most effective mitigation strategies. In order to equip course participants with the skills for achieving these outcomes, this course addresses the major steps in the development of environmental models, and how they are used for decision-making, with a particular emphasis on water quality and responding to potential climate change impacts. Topics to be covered include one or more of the following: model specification (types of models e.g. process-driven models, artificial neural networks, environmental processes, model complexity, model application), model calibration (different optimisation methods, including gradient methods and evolutionary algorithms), model validation (structural, replicative and predictive validity) and stochastic modelling (types of uncertainty, random variables, risk-based performance measures and reliability analysis, including Monte Carlo simulation), environmental decision-making (multi-objective trade-offs, multi-criteria decision analysis). These topics are explored through a project on managing dissolved oxygen and salinity in a river system under climate and population change.
Course learning outcomes
- Recognise, discuss, apply, test and critically evaluate different model types (e.g. data-driven (machine learning), process-driven).
- Recognise, discuss, apply, test and critically evaluate the different steps in the development of models (e.g. model specification, calibration and validation) and the methods used in each of these steps.
- Develop, test and apply process-driven dissolved oxygen and data-driven (machine learning) salinity models in river systems.
- Distinguish between sources and different types of uncertainty, explain their potential origins and discuss how they might impact engineering modelling and decision-making.
- Recognise, interpret, discuss, apply, test and critically evaluate different approaches to incorporating uncertainty into engineering modelling and decision-making.
- Use models and multi-criteria decision analysis approaches to solve complex engineering problems that examine the trade-offs between economic, environmental and social outcomes in an uncertain environment,including the development of solutions to adapt to climate change impacts.
- Describe, discuss and critically evaluate modelling and management processes, findings and decisions.
- Apply an integrative or systems approach to solving engineering problems.
- Use computers and information technology effectively.