Use this URL to cite or link to this record in EThOS:
Title: An evaluation of structural loop analysis on dynamic models of ecological and socio-ecological systems
Author: Abram, Joseph James
ISNI:       0000 0004 7656 2771
Awarding Body: University of Southampton
Current Institution: University of Southampton
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
This thesis evaluates a modelling analysis technique known as Loop Eigenvalue Elasticity Analysis for its utility and application to system dynamic models of ecological and socio-ecological systems. The technique acts as a structural analysis of the interactions within the system and is capable of identifying feedback loops as structural drivers of dynamic behaviour. With adverse behaviours of many ecological systems known to be driven by feedback mechanisms, Structural Loop Analysis could become an important method for increasing our understanding and control over the systems on which we so greatly depend. Within this thesis, a detailed account of the methodology and application of structural loop analysis to ecological dynamic models is undertaken. The focus of the thesis is an assessment of the technique for its ability to improve model design, to increase understanding of system behaviour and ultimately to evaluate if it could be used for the design and implementation of policy surrounding complex ecological and socio-ecological systems. Dynamic system models are predominantly used for exploring the interactions which occur within and between systems. Dynamic system models are used across a wealth of academic fields and, much like the purpose of other models, allow the user to explore and manipulate a system where tests on its real-world equivalent would be impractical or unethical to carry out. Through the exploration of components interactions it is possible to learn about, observe and simulate endogenous drivers of systems as causes of dynamic behaviour and change. While the development and simulation of a dynamic system model can provide a wealth of information over a target system, model output alone can often generate more questions than were initially being asked. Converting a real world system to model format can often lead to black box models, where the combination of multiple system components and interactions between them generate unexpected dynamics, even when interactions at a local level are well understood. The complexity that is inherent to our worldly systems can often translate into the models used to represent them. Within the fields of ecology and socio-ecology, the occurrence of black box models is common and seldom a surprise to the multi-disciplinary approach to system understanding. Ecological and socio-ecological systems are highly complex, naturally incorporating social aspects of human activity and decision making with the natural world, generating an array of human-environment interactions and forming multiple feedback mechanisms between the two spheres. Models of these systems can quickly become just as difficult to interpret as the real world systems, limiting our ability to run and understand sensitivity analysis, conduct meaningful scenario testing or use these models to reflect on policy implementation. Maintaining ecological systems in desirable states is key to developing a growing economy, alleviating poverty and achieving a sustainable future. While the driving forces of an environmental system are often well known, the dynamics impacting these drivers can be hidden within a complex structure of causal chains and feedback loops. It is important that we are always on the lookout for new modelling methods, developing and learning new ways to represent the dynamics and behaviours capable by our target system. Modelling analysis tools are an important step in the modelling process, able to extract additional information of a target system that is often unavailable from model output alone. Exploring analysis tools can bring new techniques and new understanding to our model systems which translates to a greater knowledge and understanding of the target system.
Supervisor: Dyke, James ; Hutton, Craig ; Dearing, John Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available