Natural Hazard Assessment and Mapping - Wellington Region

Abstract

This thesis concerns the hazardousness of the Wellington Region, New Zealand. As no single source exists which documents the natural hazard experience of the Region, many diverse records have been brought together, scrutinized, analysed and compared. The research examines descriptions and spatial distributions of about 18,000 natural hazard impacts that have been found in a search of insurance, city council and other historical records covering the period 1968 to 1986. This period is too short to have included impacts from the full range of hazards which threaten the Region. However, records from this 19 year period show impacts from earthquakes, high winds, landslides, flooding, marine erosion and ice on roads. Despite the difficulty of obtaining records, and the variability in the information they provide, it has been possible to rank the significance of different natural hazards in the Region. The greatest potential hazards are those posed by earthquakes, tsunamis and overbank flooding of the Hutt River. Lesser, more frequent, hazards are posed by high winds, surface flooding and landslides. Minor hazards arise from ice on roads, wave-impact and volcanic ash. For most hazards, variations of occurrence in space have been assessed and can be accounted for largely by differences in local geology, landform, topography, land-use and types of urban development. A landform classification for the Wellington Region has been produced as part of this spatial analysis. A second part of this research identifies methods by which landslide hazard mapping or zonation might be carried out in the Wellington Region. Trials of many different zonation methods identified from the literature have been carried out in a sample area. The methods of landslide hazard zonation most suitable for regional-scale application in the Wellington Region are: (1) an approach involving the geomorphic analysis of successive stages of hillslope evolution, and (2) factor overlay methods which involve the construction of many maps of factors that promote or inhibit landslide occurrence, and superimposing these to produce maps of relative landslide susceptibility. The abilities of the resulting zonation maps to predict spatial differences in landslide occurrence have been compared quantitatively.