No description available.

No description available.

No description available.

No description available.

The 2010-2011 Canterbury, New Zealand, Earthquake Sequence (CES) resulted in a liquefaction dataset of unprecedented size and quality, presenting a truly unique opportunity to assess and improve the efficacy of liquefaction-analytics in the field. Towards this end, the study presented herein develops and analyzes a database of 10,000 high-quality liquefaction case histories resulting from the CES. The objectives of these analyses are varied, but underlying each is the desire to more accurately assess liquefaction hazard for civil infrastructure (i.e., to predict both the occurrence and damage-potential of soil liquefaction). Major contributions from this work include, but are not limited to: (1) the Liquefaction Potential Index (LPI), the state-of-practice framework for assessing liquefaction hazard, is shown to produce erroneous predictions for a significant percentage of the assessed case histories; (2) the cause of poor predictions is rigorously investigated and specific shortcomings of the LPI framework are identified; (3) based on the limitations identified, and using insights from historical data, a revised liquefaction hazard framework is developed; and (4) the revised framework is shown to assess liquefaction hazard more efficiently relative to both LPI and a competing alternative framework newly proposed in the literature. Ultimately, significant room for improvement remains with respect to accurate assessment of liquefaction hazard. The findings presented in this dissertation thus form the basis for future development of a further-improved framework. Moreover, a methodology is proposed by which improvements can be measured in a standardized and objective manner. / Ph. D. / Soil liquefaction is a common cause of ground failure during earthquakes and is directly responsible for tremendous damage to civil infrastructure. Manifestations of liquefaction include the occurrence of sand blows and lateral spread failures, settlement and tilting of structures, cracking of pavements, and failure of buried lifelines due to flotation or differential settlements, among others. These effects were vividly displayed during the 2010-2011 Canterbury, New Zealand, Earthquake Sequence (CES), which resulted in a liquefaction dataset of unprecedented size and quality, presenting a truly unique opportunity to advance the science of liquefaction hazard. Towards this end, the study presented herein develops and analyzes a database of 10,000 high-quality liquefaction case histories resulting from the CES. The objectives of these analyses are varied, but underlying each is the desire to more accurately assess liquefaction hazard for civil infrastructure (i.e., to predict both the occurrence and damage-potential of soil liquefaction). Major contributions from this work include, but are not limited to: (1) the Liquefaction Potential Index (LPI), the state-of-practice framework for assessing liquefaction hazard, is shown to produce erroneous predictions for a significant percentage of the assessed case histories; (2) the cause of poor predictions is rigorously investigated and specific shortcomings of the LPI framework are identified; (3) based on the limitations identified, and using insights from historical data, a revised liquefaction hazard framework is developed; and (4) the revised framework is shown to assess liquefaction hazard more efficiently relative to both LPI and a competing alternative framework newly proposed in the literature. Ultimately, significant room for improvement remains with respect to accurate assessment of liquefaction hazard. The findings presented in this dissertation thus form the basis for future development of a further-improved framework.

White Island and Ruapehu are currently the most active volcanoes in New Zealand. During non-eruptive periods, intense quiescent degassing through fumaroles can occur. The current project studies the quiescent degassing plumes, including aerosol sampling on White Island and dispersion modelling of SO₂ and PM₁₀ from White Island and Ruapehu volcanoes. Aerosol sampling from fumaroles at the crater floor on White Island volcano was carried out on 9 February and 6 April 2005. The exposed filters were analysed for various anions and cations and the particle mass concentration and molar concentration determined. Major elemental constituents were sodium and chlorine (Na⁺: 413 µg m⁻³, Cl⁻: 1520 µg m⁻³), which show best correlation at both sampling sessions. Other ions detected, with little correlation, are Ca²⁺, PO₄³⁻ and to a certain extent Mg²⁺. Other constituents found, which cannot correlate explicitly to other ions, are K⁺, NH₄⁺, NO₃⁻, and SO₄²⁻. SEM study of one exposed filter was performed and mainly NaCl particles could be distinguished due to their well-defined cubic shape. The Air Pollution Model (TAPM) was used for dispersion modelling of SO₂ (models 1-4) and PM₁₀ (models 5 and 6) from White Island and Ruapehu volcanoes. Annual modelling was performed using different parameters of emission rate, exit temperature and exit velocity. The resulting plume dispersions show relatively low concentrations at ground level ≤10 m), particularly for the models of PM₁₀ dispersion. TAPM calculated the highest SO₂ ground level concentrations with model 4, where the NES values of 350 and 570 µg m⁻³ were exceeded several times. The data was then used for detailed hazard assessment of urban population in the North Island. The meteorological data from annual modelling was used for model evaluation and compared with observation data from different weather stations by statistical calculations. Overall, TAPM performed well with most good and very good results. To evaluate SO₂ dispersion modelling, airborne plume measurements were carried out on 22 November 2006 by plume traverses at 3, 10 and 20 km. Although there is some variation, the calculated correlation coefficients indicate good model results for two plume traverses at 3 and 20 km and one plume traverse at 10 km. The meteorological data was also used for model evaluation, and the results indicate good model performance. TAPM is therefore suggested for future studies when more observation data are available to verify the calculated model data.

Debris-flows pose serious hazards to communities in mountainous regions of the world and are often responsible for loss of life and damages to infrastructure. Characterised by high flow velocity, large impact forces and long runout, debris-flows have potential discharges several times greater than clear water flood discharges and possess much greater erosive and destructive potential. In combination with poor temporal predictability, they present a significant hazard to settlements, transport routes and other infrastructure located at the drainage points (fan-heads) of watersheds. Thus, it is important that areas vulnerable to debris-flows are identified in order to aid decisions on appropriate land-uses for alluvial fans. This research has developed and tested a new GIS-based procedure for identifying areas prone to debris-flow hazards in the Coromandel/Kaimai region, North Island, New Zealand. The procedure was developed using ESRI Arc View software, utilising the NZ 25 x 25 m DEM as the primary input. When run, it enabled watersheds and their associated morphometric parameters to be derived for selected streams in the study area. Two specific parameters, Melton ratio (R) and watershed length were then correlated against field evidence for debris-flows, debris-floods and fluvial processes at stream watershed locations in the study area. Overall, strong relationships were observed to exist between the evidence observed for these phenomena and the parameters, thus confirming the utility of the GIS procedure for the preliminary identification of hydrogeomorphic hazards such as debris-flow in the Coromandel/Kaimai region study area. In consideration of the results, the procedure could prove a useful tool for regional councils and CDEM groups in regional debris-flow hazard assessment for the identification of existing developments at risk of debris-flow disaster. Furthermore, the procedure could be used to provide justification for subsequent, more intensive local investigations to fully quantify the risk to people and property at stream fan and watershed locations in such areas.

The Canterbury Region is susceptible to a variety of natural hazards, including earthquakes, landslides and climate hazards. Increasing population and tourism within the region is driving development pressures and as more and more development occurs, the risk from natural hazards increases. In order to avoid development occurring in unacceptably vulnerable locations, natural hazard assessments are required. This study is a reconnaissance natural hazard assessment of Lakes Lyndon, Coleridge and Tekapo. There is restricted potential for development at Lake Lyndon, because the land surrounding the lake is owned by the Crown and has a number of development restrictions. However, there is the potential for conservation or recreation-linked development to occur. There is more potential for development at Lake Coleridge. Most of the land surrounding the lake is privately owned and has less development restrictions. The majority of land surrounding Lake Tekapo is divided into Crown-owned pastoral leases, which are protected from development, such as subdivision. However, there are substantial areas around the lake, which are privately owned and, therefore, have potential for development. Earthquake, landslide and climate hazards are the main natural hazards threatening Lakes Lyndon, Coleridge and Tekapo. The lakes are situated in a zone of active earth deformation in which large and relatively frequent earthquakes are produced. A large number of active faults lie within 15 km of each lake, which are capable of producing M7 or larger earthquakes. Ground shaking, liquefaction, landslides, tsunami and seiches are among the consequences of earthquakes, all of which have the potential to cause severe damage to lives, lifelines and infrastructure. Landslides are also common in the landscape surrounding the lakes. The majority of slopes surrounding the lakes are at significant risk from earthquake-induced failure under moderate to strong earthquake shaking. This level of shaking is expected to occur in any 50 year period around Lakes Lyndon and Coleridge, and in any 150 year period around Lake Tekapo. Injuries, fatalities and property damage can occur directly from landslide impact or from indirect effects such as flooding from landslide-generated tsunami or from landslide dam outbreaks. Lakes Lyndon, Coleridge and Tekapo are also susceptible to climate hazards, such as high winds, drought, heavy snowfall and heavy rainfall, which can lead to landslides and flooding. Future climate change due to global warming is most likely going to affect patterns of frequency and magnitudes of extreme weather events, leading to an increase in climate hazards. Before development is permitted around the lakes, it is essential that each of these hazards is considered so that unacceptably vulnerable areas can be avoided.

The impetus for this research was the increasing threat of catastrophic wildfires resulting from the accumulation of fuels across the West. Guided by the priorities, goals, and guiding principles outlined by the national fire plan (NFP), the objective was to identify those areas within a pinyon-juniper woodland-dominated landscape with the highest hazard of catastrophic wildfire. The intent was to help managers prioritize proactive fuels management efforts outside of the wildland urban interface (WUI). Based on a management paradigm, constraints were placed on the data collection, analysis, and model development. A geographic information system (GIS) was used to create a hazard assessment at a landscape scale in Tintic Valley, Utah. Hazard categories were a classification of fuels based on crown cover of pinyon-juniper trees, utilizing remotely sensed data. The data set consisted of digital orthophoto quadrangle (DOQ) images from 1993. The methods were developed in three phases. Phase One resulted in a hazard assessment protocol. In Phase Two, data layers were created to further divide the hazard categories into more tractable management units. Phase Three, through the retrospective examination of recent wildfires, indicated the limitations and utility of the assessment technique. The protocol presented provides a relatively fast, inexpensive, and timely hazard classification technique for pinyon-juniper woodlands at a watershed level. It is intended to be used for coarse-scale assessments of fuel hazards for strategic planning purposes. While not appropriate for fire behavior predictions, this assessment can focus managerial efforts for additional tactical planning.

This thesis reports a rock slope hazard investigation located in the Central Andes of Chile, where two significant rock mass wasting events were recognized. Dating using cosmogenic nuclide for ³⁶Cl showed that the deposits were post-glacial in age, corresponding to the Upper Pleistocene Portillo Rock Avalanche (PRA) and a Holocene rock slump and rockslide. The pre-historic landslide deposits underlie both a key transportation route between Chile and Argentina and an important ski resort. The purpose of this research was to investigate the likely failure mechanism and characterise the runout path and volume of the PRA. The insights gained on the back analysis of the slope were used in later stages to assess the hazard potential of a recurring major rockslide. The distinct element code UDEC was used to evaluate the failure mechanism. Elasto-plastic modelling results showed that sliding and shearing along the bedding planes together with brittle fracturing and shearing through the toe of the slope likely had occurred. Runout simulations were carried out using DAN3D. Combinations of rheologies were tested and ranked based on their ability to represent the current distribution of the debris by means of pre-failure topography reconstruction and volume estimates of the deposits. Results showed that the best basal rheological combination for the PRA was frictional during the rockslide and Voellmy when entrainment became important. In contrast, a constant frictional basal rheology best represented the Holocene rock slump. The performance of the present-day state of the slope was tested under different scenarios. Under static condition the slope proved to be stable indicating a stabilized geometrical profile with time. Also, the slope proved to be stable under increased pore water pressures at its toe. Finally the modelled slope was subjected to a seismic load (M=7.8) and its crest failed due to an outward rotation of blocks, probably aided by topographic amplification. The runout simulations showed that the leading edge of the flow could override part of the International Santiago-Mendoza Corridor with no direct impact to the Portillo Ski Resort. Overall, though, under this highly unlikely dynamic condition for the site, the hazard level is very low.