A mass movement classification for the southern Drakensberg, South Africa
Date
2013-05-29
Authors
Hardwick, Devlyn
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
A variety of mass movement landforms occur in the southern Drakensberg, South Africa, and whilst a number of studies on individual landforms have been conducted, regional scale assessments of the Ukhahlamba Drakensberg Transfrontier Park have been relatively limited. Mass movement has been defined as the downward and outward movement of slope-forming material under the influence of a transporting agent such as water, air, ice or snow (Goudie, 2004). This includes landforms such as landslides, debris flows, terracettes, solifluction lobes and rockfall. Although two landslide risk assessments have been conducted in the region, one was site specific and focussed on shallow, translational slides (Bijker, 2001), whilst the other was at a much larger regional scale and focused on large palaeo-mass movements (Singh, 2008).
Numerous international mass movement classifications have been developed over the years, and one of the primary aims of this research is to develop a classification for mass movement landforms within a southern African context. A number of mass movement landforms were identified, measured and mapped in the field to acquire a better understanding of how the landforms originate. This classification was then further adapted to facilitate the identification of mass movement landforms from orthophotos. Aerial photo interpretation techniques were used to map three terrace-type mass movement landforms and four shear-type mass movement landforms in the Garden Castle State Forest of the Ukhahlamba Drakensberg Transfrontier Park.
A further level of detail was added to the classification by ascribing environmental conditions to the different landform types. A Geographic Information System was used to collate and generate spatial information which could be added to the landforms in the mass movement inventory. These were
then analysed using univariate and multivariate statistical modelling. Histograms, as well as an area-weighted frequency distribution, were used to describe the landforms and then hierarchical partitioning was used to identify the environmental variables associated with each type of landform. One main environmental variable was identified for each type of mass movement. Logistic regression was then used to create probability maps for each type of landform. An average of 30% of the study area has a medium to very high likelihood of developing mass movements, although this percentage varies for each type, whilst rock movement deposits are predicted to occupy more than 80% of the study area. Gradient, altitude and lithology were selected most frequently by the statistical models as influencing landform distribution, whilst distance to a rock exposure had the strongest influence on the location of rock movement deposits. Aspect was selected least frequently by hierarchical partitioning which raises questions about the influence of aspect on valley asymmetry. Various models have been developed which describe slope development in the Drakensberg with reference to slope aspect, however the results of this study suggest that other environmental factors may be more important and that slope development is a complex process.
Description
A thesis submitted to the Faculty of Science, University of the Witwatersrand, in fulfillment of the requirements for the degree of Doctor of Philosophy
Johannesburg, 2012.