Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
A generic probabilistic model for natural hazard assessment
Thesis (PhD)--University of Pretoria, 2019.
Saved in:
| Other Authors: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
University of Pretoria
2019
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867613464130748416 |
|---|---|
| access_status_str | Open Access |
| author2 | Kijko, Andrzej |
| author_browse | Kijko, Andrzej |
| author_facet | Kijko, Andrzej |
| collection | Thesis |
| dc_rights_str_mv | © 2019 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria. |
| description | Thesis (PhD)--University of Pretoria, 2019. |
| format | Thesis |
| id | oai:repository.up.ac.za:2263/70291 |
| institution | University of Pretoria (South Africa) |
| language | English |
| last_indexed | 2026-06-10T12:36:33.567Z |
| license_str | Other — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from UPSpace — University of Pretoria Institutional Repository |
| publishDate | 2019 |
| publishDateRange | 2019 |
| publishDateSort | 2019 |
| publisher | University of Pretoria |
| publisherStr | University of Pretoria |
| record_format | dspace |
| source_str | UPSpace — University of Pretoria Institutional Repository |
| spelling | oai:repository.up.ac.za:2263/70291 A generic probabilistic model for natural hazard assessment Kijko, Andrzej ansie.smit@up.ac.za Stein, Alfred Smit, Ansie UCTD Mathematical Statistics Natural hazard assessment Thesis (PhD)--University of Pretoria, 2019. A generic methodology for probabilistic natural hazard assessment is presented. Three area-characteristic recurrence parameters are defined by combining a Poisson process with the relevant natural-hazard-frequency–event-size power law. The distribution of the Poisson process describes the temporal characteristics present in the data and the power law describes the relationship between the frequency of events and the event sizes. The estimates for the mean rate of occurrence λ and the power law parameter b are based on empirical datasets consisting of extreme prehistoric and historical data, along with more-recent instrumental data. Likelihood functions are defined to allow for datasets to be combined and for the application of both maximum likelihood estimation (MLE) and Bayesian inference (BI). The proposed methodology accounts explicitly for aleatory and epistemic uncertainty by making provision for incomplete datasets, uncertainty associated with the observed event sizes, uncertainty associated with the parameters of the applied occurrence and event size distributions, and uncertainty associated with the occurrence of events in the dataset. These types of uncertainty are introduced in the modelling process through convolution and mixture distributions, as well as weighted likelihood functions. Existing techniques to assess the third recurrence parameter, the maximum possible event size x_max, are discussed briefly. The applicability of the proposed methodology is demonstrated by using a synthetic earthquake dataset, real earthquake datasets for Central Italy and the Ceres–Tulbagh region in South Africa, tsunami data for three tsunamigenic regions in the Pacific Ocean, and HAILCAST ensemble re-analysis hail data for Gauteng province, South Africa. Various combinations of the different types of assumptions, data, and uncertainty are investigated. The methodology shows the universality of the power law in assessing natural hazards. In practice, the methodology is not restricted to natural hazard assessment, but can be applied to any instance in which the frequency–event-size relationship follows a power law distribution. To illustrate this statement, financial vehicle loss information related to hail damage, obtained from a short-term insurer in South Africa, is analysed. The versatility of the modelling process provides the researcher with various options to account for incomplete data, as well as data and parameter uncertainty. • Research and travel were supported by the South African National Research Foundation and the South Africa Statistical Association under the SASA-NFR Grant for Vulnerable Discipline — Academic Statistics 2017. This work is based on research supported wholly or in part by the National Research Foundation of South Africa (Grant Numbers 76906, 96412, 94808 and 103724). • University of Pretoria Natural Hazard Centre, University of Pretoria, Department of Geology. Statistics PhD Unrestricted 2019-06-25T11:46:49Z 2019-06-25T11:46:49Z 2019 2019 Thesis Smit, A 2019, A generic probabilistic model for natural hazard assessment, PhD Thesis, University of Pretoria, Pretoria, viewed yymmdd <http://hdl.handle.net/2263/70291> S2019 http://hdl.handle.net/2263/70291 en © 2019 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria. application/pdf University of Pretoria |
| spellingShingle | UCTD Mathematical Statistics Natural hazard assessment A generic probabilistic model for natural hazard assessment |
| title | A generic probabilistic model for natural hazard assessment |
| title_full | A generic probabilistic model for natural hazard assessment |
| title_fullStr | A generic probabilistic model for natural hazard assessment |
| title_full_unstemmed | A generic probabilistic model for natural hazard assessment |
| title_short | A generic probabilistic model for natural hazard assessment |
| title_sort | generic probabilistic model for natural hazard assessment |
| topic | UCTD Mathematical Statistics Natural hazard assessment |
| url | http://hdl.handle.net/2263/70291 |