The research area of earthquake hazard is driven by the need for reliable forecasts on earthquake shaking. While we can not in the foreseeable future forecast the earthquake occurrence, we can forecast earthquake shaking intensities with appropriate mathematical/statistical tools. The basis tools were pointed at in the late 1960’s by L. Estava and later A. Cornell who is today regarded as the founder of probabilistic seismic hazard analysis. The probabilistic seismic hazard provides the annual exceedence probability for given ground motions in hazard curves (see Figure 1), and this methodology has today practically substituted the classical deterministic methods that determine ground shaking from mapped faults assuming a certain rupture.
The seismic hazard is a direct result of the needs of engineers for “baselines” for the structures exposed to earthquake shaking: once a ground motion level has been defined, it is a question of engineering techniques to design a structure that can withstand the shaking.
An earthquake emits shaking energy over a wide frequency band, and it is now evident that the structures performance is strongly frequency dependent. This is the basis for the representation of earthquake shaking through response spectra. A response spectrum is the collection of the responses of one-degree-of-freedom oscillators with different resonance frequencies subjected to the shaking from a given earthquake. The elastic response spectrum (Figure 2) represents a new baseline for the engineering design, and is nowadays developed for building regulation codes. The “equal hazard response spectrum” can be developed when combining the hazard curves (Figure 1) for all frequencies of interest such that the probability of exceedence is equal for all frequencies and the associated ground motion levels. It is nowadays also common to conduct a deaggregation analysis as shown in Figure 3.
The brief outline of the statistical framework is available here.
Within the above research area there are several topics of major importance and that are currently under focus as research themes. One of the most important themes is to reduce the scatter in the ground motion prediction equations. These equations are based on observed ground motion, and demonstrate a large scatter for earthquakes having similar magnitudes and distances. One of the currently debated issues is how many sigma values (representing the scatter) that should be included in the hazard computations. Figure 3 demonstrates the deaggregation as a function of varying epsilon which is the factor that determines the number of sigmas to include in the computation.
 |
|
 |
|
Figure 1: Example of a seismic hazard curve where the probability decreases for increased ground motion shaking. |
|
Figure 2: Example of an equal hazard response spectrum. The representation can be given in a tripartite plot (as above) or as normal Sa (spectral acceleration) as function of frequency. |
 Figure 3: Example of a deaggregation of the hazard results as function of magnitude and distance and epsilon value (which compares to the sigma value in the attenuation relation). See text above for details. |
 |
|
|
|
