| Summary: | Coastal hazard assessments commonly require the integration of different data sets that, in turn, require different statistical techniques to reveal underlying processes. In this thesis, I present three studies in which I apply and develop different statistical methods for a broad sedimentological data set and sea-level time series. In the first study, I use a number of statistical analyses to examine the use of grain size parameters, mineral composition and trace element geochemistry in determining the provenance of tsunami (the 2004 Indian Ocean Tsunami (IOT) and three paleo-tsunami) deposits and the 2007 storm surge deposit on Phra Thong Island, Thailand. I also evaluate whether the 2004 IOT tsunami and 2007 storm deposits could be discriminated using grain size and geochemistry. The key findings are that geochemistry data are statistically inadequate to distinguish the provenance of modern storm and tsunami deposits, but the mean grain size is probably a good discriminator. In addition, the sediment sources of each of the overwash deposits are diverse. In the second study, I investigate interannual sea-level variations using sea-level records at ten tide gauges (TGs) around the South China Sea (SCS). The results reveal that the ENSO significantly contributes to interannual sea-level signals but that its influences vary across the study area, whereas the winter monsoon clearly impacts sea level in the northern SCS. In the third study, I examine spatio-temporal variation of extreme sea level (ESL) around the SCS. The results reveal contrasting mechanisms for generating surges in the northern and southern SCS; and that the changes in ESL are broadly consistent with the changes in mean sea level. I also calculate the temporal variability of the return levels and show they are significantly correlated with climatic variability in the region.
|
|---|