Strong Motion Data Processing and Recording at University of Southern California

A brief review is presented of the Los Angeles and Vicinity Strong Motion Network operated by the University of Southern California (USC), consisting of 80 stations deployed in 1978-1980, and of software developed and used at USC for digitization of accelerograms recorded on film, and for routine and specialized data processing of digitized or digitally recorded accelerograms. The currently used digitization system consists of a flatbed scanner, a PC, and the LeAuto software. The films are scanned at 600 dpi optical resolution, which is less than the limit of the hardware, but was found to be optimal, considering the limitations of the recorders, and benefit versus cost. The standard data processing methods used for further processing are a new generation of the software developed by Trifunac and Lee [1,2]. Baseline correction is performed by high-pass filtering, with cut-off determined for each component separately using a standard noise spectrum so that the recorded signal-to-noise ratio is greater than unity. The high frequency noise is removed by low pass filtering (with ramp 25-27 Hz for film records). In a batch mode of processing, the low frequency cut-off is determined automatically by the program, and is later verified by an operator, who can specify the filter manually if necessary. Filtering is performed in the time domain by Ormsby filter (a non-causal filter that does not introduce phase distortion), after appropriate even extension of the record. Digitally recorded accelerograms are also baseline corrected by high pass filtering, as random piecewise baseline offsets are not uncommon in digital accelerograms, even for small accelerations, and permanent displacements cannot be computed reliably from recorded three components of acceleration alone.