Narrow Results

The sensitivity of different parameters used in probabilistic seismic hazard calculation is investigated by different logic tree runs with alternative magnitude sets, source zone models and attenuation relations, and with different sets of values for the seismicity parameters and the σ-value. Also the influence from the different parameters on the hazard uncertainty, represented by fractiles, is investigated. The calculations are made for peak ground acceleration at a site near Aachen in the Lower Rhine Embayment. The model where the site is located in a larger source zone gives lower hazard values. This is typical for the case where the seismicity near the site is high relative to its surrounding. The hazard curves for the different attenuation functions are similar, an effect of the similarity of the functions themselves. However, a large sensitivity of this parameter is indicated for small mean return periods. An increased σ-value implies a moderate increase of the hazard at long mean return periods. The hazard is increasing for decreased focal depth, decreased β-value and increased maximum expected magnitude, respectively. However, the effects are noteworthy only at low hazard levels for variations in the focal depth and to some extent in the maximum expected magnitude. Finally, decreasing the minimum magnitude thought to be of engineering relevance causes a drastic increase of the hazard at small mean return periods.

The city of Oslo, Norway, was affected by a magnitude 5.4 earthquake in 1904 causing widespread minor damage. The earthquake occurred around 100 km south of Oslo within the Permean rift structure that runs North-South along the Oslofjord, and deep clay deposits under the city contributed to the damages. A seismic risk scenario including soil amplification and buildings classifications has been conducted with two earthquake sources, one very close to the city and one near the 1904 epicenter. Both scenarios exhibit strong dependencies on the soft clays underlying large parts of Oslo. The results confirm the 1904 effects, but also show a strong dependency on the applied attenuation functions. All computations are based on the capacity-spectrum method, and the predefined pushover curves and vulnerability functions were adopted from the HAZUS code. With this basis, the computational scheme was developed independent from the GIS framework, and a weighted logic tree formulation was implemented for appropriate treatment of epistemic uncertainties.