- Ground-motion model is:
σ = 0.314 for regional model and 0.331 for local model. No coefficients are reported.
- Use 2 site classes:
- Use coefficient c1
- Use coefficient c2
Use V s when available and otherwise geological and geophysical (e.g. H/V ratios) information. Because
most Algerian stations on firm soils/rock model applies to these sites.
- Derive two models: regional (using data from Algeria up to 06/06/2008, Europe and USA) and local (only
data from Algeria).
- Find local model overestimates observations for large earthquakes.
- Examine residuals w.r.t. rhypo and find no bias. Relate large residuals to local site effects.
- Data from: SMA-1 (analogue), SSA-1 and Etna (digital) instruments.
- Instrument-correct data, fit baseline and band-pass Ormsby filter with cut-offs 0.12–0.2Hz with roll-off
width of 0.06Hz for digital instruments and 0.2–0.3Hz with roll-off width of 0.1Hz for analogue instruments
and 25Hz (with 3Hz roll-off width for all instruments). Cut-offs based on signal-to-noise ratio of each
- Algerian data mainly covers 3 ≤ Ms ≤ 6 and 7 ≤ rhypo ≤ 100km. Include foreign data to supplement
Algerian data outside these ranges.