Soil Dynamics and Earthquake Engineering I1 (1992) 101-110 A note on scaling peak acceleration, velocity and displacement of strong earthquake shaking by Modified Mercalli Intensity (MMI) and site soil and geologic conditions M.D. Trifunac & V.W. Lee University of Southern California, Civil Engineering Department, Los Angeles, California 90089-2531, USA (Received 1 March 1991; revised version received 14 May 1991; accepted 16 May 1991) It is shown that the overall trends of peak amplitudes of strong motion accelera- tion, velocity and displacement versus site intensity in the Western United States are consistent with our previous inferences, t8 but the larger present data base allows us to begin with identification of the additional changes of average peak amplitudes also in terms of the local soil and local geologic site conditions. These trends display a reversal of peak amplitude trends in going from small to large intensities, making it difficult to reconcile the simple scaling of design spectra in terms of peak ground acceleration. INTRODUCTION More than 15 years ago, we presented a study on the relationships between the peaks of the recorded strong motion amplitudes and the Modified Mercalli Intensity (MMI) at the recording site. 18 In that work, we found general agreement of our results with the overall trends reported by others, and somewhat larger peak ampli- tudes for strong ground motion recorded in the Western United States. Our study in 1975 used 186 strong motion accelerograms, recorded during 57 earthquakes which occurred between 1933 and 1971. During the late 1970s and through the 1980s, the uniformly processed strong motion data basC grew steadily and, at present, it exceeds 1000 records. The process of collecting the local soil and geologic site con- ditions at all recording stations has been slow and some, although incomplete, data is now available only for the recordings up to about 1984. ~4 For consistency, and to enable direct comparisons with other related results, W5'17 we will use this same data base here also. Between 1984 and present, excellent additional recordings have been obtained, but it may take years before the uniform data for local soil and local geologic site conditions are gathered and become available for use in this type of analysis. Soil Dynamics and Earthquake Engineering 0267-7261 •92]$05.00 © 1992 Elsevier Science Publishers Ltd. Since the late 1950s, when the response spectrum was first proposed for use in the design of nuclear power plants, through the 1960s, 1970s and 1980s, the spectral characterization of strong earthquake staking ampli- tudes for earthquake resistant design become more wide spread. It was originally recommended that the scaling of response amplitudes be done in terms of spectral intensity. 7 However, the empirical equations on how spectral intensity should scale with magnitude and distance or with the MMI at the site, became available only some 20 years later. 19'24"25 On the other hand, the direct and simple relationship between the high fre- quency spectral amplitudes and the recorded peak acceleration could be used immediately. In the following years, this scaling in terms of peak ground acceleration was used by many, both in theoretical and in practical engineering uses of strong motion data. With this back- ground, one aim of this paper is to examine whether this approach can be reconciled with what we know about the strong earthquake motion today, with emphasis on the scaling in terms of MMI at the site. Through the 1970s and the 1980s, numerous contri- butions to detailed scaling of Fourier and response spectra of strong ground motion have been presented. It was shown that the spectral amplitudes and the spectral shapes depend on the size of the earthquake (scaled either by magnitude or by intensity), local geologic and local soil conditions, 12'W5 vertical versus horizontal com- 101