Date of Award
Bachelor of Applied Sciences Honours
Faculty of Science and Technology
Dr James Cross
In environmental studies, many programs have been implemented for the purpose of examining the influence of factors on the natural environment from various sources. In the past, predictions have been obtained from a simple regression model of the data from these programs to aid in deciding whether an intervention is required in the program or policy. However, most environmental time series involve correlated dependent variables which makes it difficult to apply conventional regression analysis. This thesis compares the application of the class of transfer function models, where high correlation between dependent variables is allowed, to that of conventional regression analysis. Regression analysis, is a method of analysis that can be used to examine data. It can also be used to help determine if dependency relationships do exist in the data. Regression models CM also be used to help determine if dependency relationships do exist in the data. Transfer function modelling was first introduced by Box and Jenkins in 1976. The Box-Jenkins methodology is a systematic procedure. This involves a prior analysis of the data under consideration. This methodology comprises three main stages: identification, estimation and diagnostic checking. The identification stage is considered to be the most difficult stage in the model building. Due to lack of computer software and the complexity of the underlying theory, transfer function modelling has nm been practically applied to analysis of environmental data in Australia. This thesis explores each class of time series models in depth, in order to investigate the performance of each technique on a current environmental data set. It also shows the significant advantages transfer function modelling has over regression analysis. The data being examined is collected from the Western Australia Marine Research Laboratories. This thesis examines if the environmental factors involved in this data set, the Leeuwin Current and the westerly winds, have a significant effect on levels of puerulus settlement, at two coastal sites. Dongara and the Abrolhos Islands, in the western rock lobster fishery.
Farag, S. A. (1994). Application of time series models to sets of environmental data. Retrieved from http://ro.ecu.edu.au/theses_hons/633