In recent years, GEM has been collaborating with Taiwan Earthquake Model (TEM) in the development of the new Taiwan seismic hazard model.
The recent visit to the GEM offices by Dr Chung-Han Chan of TEM provided an excellent opportunity to continue building on the GEM-TEM collaboration, following Dr Chan's previous visit in December 2012. The primary objective of the visit was to study the new Taiwan seismic hazard model (Cheng et al. 2014). The visit provided also a suitable forum for knowledge exchange on the usage of OpenQuake engine and the Hazard Modelling Tools, exploring new features implemented since Dr Chan's previous visit.
Fig. 1 Earthquake Catalogue of Taiwan overlain by the active shallow area sources
As part of the active progress, and while final results are still pending, Dr Chan and GEM’s scientists Marco Pagani, Graeme Weatherill, and Vitor Silva are are glad to share some of the preliminary results finalized during the latest visit of Dr Chan.
By working together, they are moving towards implementation of the forthcoming Taiwan seismic hazard model, a state-of-the-art model containing detailed seismogenic source characterisation and an extensive representation of epistemic uncertainties, in the OpenQuake engine software. Recent developments include:
- The use of the Hazard Modeller's Toolkit (hmtk) to analyse the earthquake catalogue and estimate the seismicity activity across Taiwan.
- The integration of several seismogenic sources (Figure 2) and corresponding parameters, for the production of a probabilistic seismic hazard map (Figure 3).
Preliminary results have shown a high hazard along some of active faults in inland Taiwan.
Fig. 2 Distribution of the seismogenic sources determined by the Sino tech Consultant Inc
Fig. 3 Seismic hazard map for Taiwan considering only active shallow sources
Scenarios for the rupture of the Meishan fault  were also investigated, and the corresponding ground shaking evaluated. By further considering recent distribution of population it was possible to estimate the expected number of fatalities, were the same event to occur today (Figure 4). By comparing the two scenario cases, a larger number of fatalities over a wider region would be expected when the scenario rupture is longer. These analyses might provide essential information on earthquake preparation and devastation estimations.
Fig. 4 Comparison of the expected fatalities for the case of the 1906 Meishan earthquake rupture (left) and for a similar magnitude rupturing a larger section of the fault (right)
Additional consideration was given to the estimation of earthquake consequences using an empirical vulnerability model derived based on data from the Chi Chi earthquake (1999). This model was employed to calculate the distribution of fatalities due to a possible repetition of the Meishan earthquake of 1906. This exercise represents the initial steps of a larger endeavour to develop a comprehensive seismic risk model for Taiwan.
Fig. 5 Estimated ground shaking based on the 1906 Meishan-earthquake rupture scenario (right). Estimated fatalities, given current exposure, from a 1906 Meishan earthquake rupture scenario (left).
The rupture of this fault in 1906 resulted in more than one thousand fatalities