School of Geography - Theses

Permanent URI for this collection

Search Results

Now showing 1 - 1 of 1
  • Item
    Thumbnail Image
    Study of a quantitative method of delineating physical-geographic regions in interdisciplinary integrated survey: the grid-point method
    Massey, Jack ( 1975)
    Interdisciplinary surveys involve mapping physicalgeographic regions according to readily observable criteria for the purpose of assisting in making a variety of decisions about land use. Difficulties are associated with traditional surveys, which rely on aerial photographic interpretation and purposive sampling, due to the lack of a standardized, quantitative methodology. The grid-point method is proposed as a rigorous, quantitative approach and was applied in the Grampians north-east section (507 sq.km.), the upper Barwon River catchment (977 sq.km.), and on French Island (172 sq.kin.). Variables used reflect obvious distributions of landform, soil, and native vegetation, and quasi-random sampling patterns with circular sampling units and densities between 1 sampling unit per .89 and 2.42 sq.km. were employed. Landform data were gathered from contour maps, and soil samples and vegetation data were gathered in the field; Boil data were obtained by laboratory analyses. Data were subjected to principal components analysis, and standardized data were scored on the first three scaled and first three orthoreormalized vectors. Sampling units were classified on the basis of the scores and then regionalized by SYM�-V proximal mapping. These regionalizations provided a sound basis for physical-geographic description. Inherent features of the grid-point method overcame difficulties associated with traditional surveys. Statistical analyses indicated that sampling densities as low as 1 sampling unit per 4 sq.km. are sufficient to generate interpretable regionalizations of the Grampians north-east section. Comparisons of regionalizations with the land systems map of this area produced by traditional methods reveals that the grid-point method is capable of generating regions similar in level of generalization to the land systems. With respect to selected variables,the regionalizations are in the majority of cases of higher quality than the land systems map. The cost of survey at 1 sampling unit per .97 sq.km. of this area is within the limits of most survey organizations and estimates for surveys at densities of 1 sampling unit per 1.14, 1.51, 2.07, 3.49, 4.15, and 9.74 sq.km. reveal that although there is a significant reduction in cost with decrease in sample size, an economy of scale factor operates. Because the grid-point method is orientated to field data gathering, it is least expensively applied in study areas characterized by undulating plains as well as flat plains and hills, which are for the most part cleared with a dry surface. Estimates for five study areas previously surveyed by the Soil Conservation Authority indicate that the cost of application of the grid-point method at sampling densities of 1 sampling unit per 1 ?q.km. and lower is not excessive. Efficiency of the grid-point method may be enhanced by purposivesystematic sampling near roads and vehicle tracks and purposive stratification of the study area prior to application. The grid-point method should be applied in interdisciplinary surveys carried out in south-eastern Australia. Applications should take the form of reco??aissan?e surveys involving relatively low density sampling. The resulting regionalizations, although probably somewhat coarse, should provide a sound framework for general pbysical-geographic description.