Estimation of S-Wave Site Response in Anchorage, Alaska, from Weak-Motion Data Using Generalized Inversion Method

Abstract

The generalized data inversion method has been used to estimate the spatial variation of site response (SR) in the Anchorage basin in south-central Alaska. The data analyzed represents weak motion of the horizontal components of S waves for local earthquakes. They were recorded by a temporary portable 22-station network (IRIS-PASSCAL equipment) that was operated for about six months. Setting the path effect from independent sources, the inversion of the data was carried out to determine SR. The SR values thus obtained were averaged (logarithmically) for two frequency bands, namely, a low frequency band (LFB) and a high frequency band (HFB) from 0.5 to 2.5 Hz and 3.0 to 7.0 Hz, respectively, with center frequencies of about 1.0 Hz and 5.0 Hz. In LFB, SR increases from 1.0 along the foothills of Chugach Mountains in the east to about 3.5 in the west in and around Campbell Lake and Government Hill. The areas with SR > 3.0 are associated with extensive ground failure during the Prince William Sound earthquake ( M w = 9.2) of 1964. In HFB, there are two small areas adjoining Tumagain Arm in the southern side of Anchorage with SR greater than 3.0. A comparison of the SR values obtained from the inversion with those reported by others for standard spectral ratio (SSR) and horizontal to vertical spectral ratio (HVR) showed (1:1) correspondence with the values of SSR but larger HVR values by a factor of about 2. Areas of soil class D in the study area are characterized by SR > 2.0, while those in C are characterized by SR ≤ 2.0 in LFB, but soil class and SR lack correlation in HFB.