• Login
    View Item 
    •   Home
    • Massey Documents by Type
    • Theses and Dissertations
    • View Item
    •   Home
    • Massey Documents by Type
    • Theses and Dissertations
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    Statistical methods for estimating tephra source and dispersal : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Statistics at Massey University, Palmerston North, New Zealand

    Icon
    View/Open Full Text
    01_front.pdf (229.0Kb)
    02_whole.pdf (7.326Mb)
    Export to EndNote
    Abstract
    Tephra refers to any pyroclastic fragments ejected from a volcanic vent and its dispersal is one of the major hazards with explosive eruptions. The attenuation of tephra fall thickness is most commonly estimated after contouring field measurements into smooth isopachs. I explicitly describe the variability in thickness by using a semiempirical tephra attenuation relation as a link function. This opens the way to fitting models to actual tephra observations through maximum likelihood estimation (MLE). The method is illustrated using data published from the 1973 Heimaey eruption in Iceland. Complex eruptions commonly produce several phases of tephra fall from multiple vents. When attempting to precisely reconstruct past eruptions from the geological record alone, separate phases are often indistinguishable. Augmented by a mixture framework, the MLE attenuation model was able to identify the sources and directions of tephra deposition for the 1977 Ukinrek Maars eruption in Alaska, US, from only the tephra thickness data. It was then applied to the unobserved 1256 AD Al-Madinah eruption in Saudi Arabia. The estimation of the spatio-temporal hazard from a monogenetic volcanic field is criti- cally dependent on a reconstruction of past events. The Auckland Volcanic Field (AVF) has produced about 50 volcanoes in the last 250,000 years. Although inconsistent, age data for many of these volcanoes exist from various dating methods with various re- liabilities. The age order of some pairs is also known due to the overlaying of lavas (stratigraphy). A discussion is provided on how informative priors are obtained via ex- pert elicitation, on both the individual volcano ages, and the reliabilities of the dating methods. A possible Bayesian model for reconciling the available inconsistent volcano age data to estimate the true eruption ages is also discussed. To improve these eruption age estimates, some of the volcanoes can be correlated with the better dated tephra layers recovered from five maars in the field. The likelihood of any combination of volcano and tephra, incorporating the spatial variability based on the attenuation model and temporal components, is evaluated and is maximised numer- ically using linear programming. This statistical matching provides an improvement in the volcano age-order model and age estimates of the volcanoes in the AVF.
    Date
    2016
    Author
    Kawabata, Emily
    Rights
    The Author
    Publisher
    Massey University
    URI
    http://hdl.handle.net/10179/9867
    Collections
    • Theses and Dissertations
    Metadata
    Show full item record

    Copyright © Massey University
    | Contact Us | Feedback | Copyright Take Down Request | Massey University Privacy Statement
    DSpace software copyright © Duraspace
    v5.7-2020.1-beta1
     

     

    Tweets by @Massey_Research
    Information PagesContent PolicyDepositing content to MROCopyright and Access InformationDeposit LicenseDeposit License SummaryTheses FAQFile FormatsDoctoral Thesis Deposit

    Browse

    All of MROCommunities & CollectionsBy Issue DateAuthorsTitlesSubjectsThis CollectionBy Issue DateAuthorsTitlesSubjects

    My Account

    LoginRegister

    Statistics

    View Usage Statistics

    Copyright © Massey University
    | Contact Us | Feedback | Copyright Take Down Request | Massey University Privacy Statement
    DSpace software copyright © Duraspace
    v5.7-2020.1-beta1