Seismic Intensity, PGA and PGV for the South Napa Earthquake, August 24, 2014

Numerous studies have investigated the statistical relationship between Modified Mercalli Intensity (MMI) and peak ground acceleration (PGA) and peak ground velocity (PGV). The Mw 6.0 South Napa (California) earthquake of August 24, 2014 provides valuable data to examine these relationships for both urban and rural environments within northern California. The finite fault model (D. Dreger, 2014) indicates that the fault rupture propagated predominantly NNW and up-dip from a 11-km-deep hypocenter. The epicentral location was 8 km SSW of downtown Napa. Recorded PGA north of the epicenter is as high as 600 cm/s2 and PGV locally reaches 80 cm/s. Field studies by two teams of investigators were conducted on August 24-26 and September 8-11, 2014 to assign MMI values at 108 sites. The highest MMI values are strongly localized along the NNW-SSE rupture trend north of the epicenter. Parts of the city of Napa and some communities several km farther north on Dry Creek Road were found to have experienced shaking intensities of MMI VII to VIII. The observed effects include houses moved off their foundations, chimney collapse or damage, cracked foundations and/or interior walls, broken windows, and the lateral displacement of heavy furniture. Communities to the east and west of this zone of high seismic intensity reported significantly lower values of MMI, typically IV and V even at distances as close as 10 km from the mapped surface rupture. In comparison with previous estimates by Wald et al. (1999) and Dangkua and Cramer (2011), we find that MMI III-VIII uniformly correspond to significantly larger (>150%) PGA and PGV values, as reported by the Center for Engineering Strong Motion Data (2014). We attribute this observation to two primary factors: (1) improved earthquake engineering in the post-Loma Prieta earthquake era that has led to building construction, both new and retrofitted, that is more resistant to earthquake strong ground motions; and (2) a frequency band relevant to these MMI estimates that contains less energy than that leading to the PGA or PGV values. The latter would primarily be a source effect.