Multiparametric quantification of volcanic hazards for eruption forecasting and communication
When a volcano enters a period of unrest it is necessary to assess the threat an eruption will pose. This requires the integration of multiparametric data sets. I present two examples of how different data sets with different levels of uncertainty can be combined to better constrain volcanic hazards...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Theses and Dissertations |
| Language: | English |
| Published: |
2016
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/66030 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | When a volcano enters a period of unrest it is necessary to assess the threat an eruption will pose. This requires the integration of multiparametric data sets. I present two examples of how different data sets with different levels of uncertainty can be combined to better constrain volcanic hazards, and also how well this information is currently being communicated. Firstly, I develop a new database of plinian eruption characteristics and use it to inform a Bayesian Belief Network that calculates the probability of an eruption becoming plinian in style. I show that the presence of an excess gas phase is one of the key factors. Secondly, I integrate geologic field studies with pyroclastic flow and ash fall modeling to improve hazard assessment at Gede volcano, Indonesia. I use this new set of data to produce an event tree for future activity at Gede. Finally, I perform an analysis of the issuance of volcano alert levels worldwide. I find that current success rates of alert issuance are low and suggest that increasing monitoring networks to an optimal level is likely to improve this. |
|---|