nsmp.wr.usgs.gov ~ftp\READ.ME last update: November 4, 1998 USGS National Strong-Motion Program Recently-Digitized Strong-Motion Records Table of Contents ================= 1.0 Introduction 2.0 Time-Series Data File Format 3.0 Data Processing Software 4.0 Digitized Records 4.1 From the 17jan94 Northridge CA EQ 4.1.1 Uncorrected data 4.1.2 Corrected data 5.0 Other Sources of Strong-Motion Data 5.1 USGS Strong-Motion CD-ROM 5.2 Surfing the Internet for Strong-Motion Data Address questions to: Chris Stephens Kent Fogleman e-mail: cdstephens@usgs.gov fogleman@usgs.gov telephone: (650) 329-4752 (650) 329-4745 post office address: USGS, MS 977 345 Middlefield Road Menlo Park, CA 94025 USA 1.0 Introduction ================= The USGS has recently purchased a scanner-based digitizing system with which to digitize our incoming analog strong-motion accelerogram records. Some of the records we have been unable to digitize for the past 5 years, primarily those subsequent to the 1989 Loma Prieta earthquake, were digitized for us by the digitizer vendor as part of the test of the new system. We will post these digitized records, and those we digitize in-house in the future, here on the "nsmp.wr.usgs.gov" computer, after we have given them a preliminary inspection. As time and staffing permit, we will add more information to the headers of the preliminary, right-out-of- the-digitizer time-series data files. We also plan to make "corrected" versions of the time series available in addition to these "uncorrected" versions. The addition of corrected data has commenced with this update. We will update Section 4.0 of this file as new files are added to, or changes are made to existing files in, this collection. The term "uncorrected" indicates that a digitized (or digitally recorded) time series has received no processing that involves any hypotheses as to the character or frequency content of the ground motions or recording instruments. An "uncorrected" analog-recorded time series has been corrected only for uneven film transport with time and for transverse motion of the film as it moved longitudinally through the recorder; it has been shifted to have zero mean; and it has been translated from digitization units to units of cm/sec/sec (ordinates) and seconds (abscissas). The "correction" process applies bandpass filters (removing noise contamination) and instrument correction (removing the effects of frequency-dependent instrument response) to a time series. Software, named "BAP", that will correct the time series generated by the new digitizer is also available from this "nsmp.wr.usgs.gov" computer. See section 3.0. 2.0 Time-Series Data File Format ================================= The "smcfmt.doc" file describes the format of the time-series data files available on the "nsmp.wr.usgs.gov" computer. The file should be found in the anonymous FTP home directory on that computer, the same directory from which the READ.ME file you are reading now was (probably) retrieved. 3.0 Data Processing Software ============================= Software, named BAP, that will process the time series available from "nsmp.wr.usgs.gov" is available from the same computer in subdirectories below "/software/bap/". Introductory information about the software and information about how one can acquire a printed user's manual is given in the file at "/software/bap/README.bap". BAP will calculate velocity and displacement from an input acceleration time series or it will calculate acceleration and displacement from an input velocity time series. The program will make linear baseline corrections, apply instrument correction, filter high frequency and/or low frequency content from the time series, calculate the Fourier amplitude spectrum, and calculate response spectra. It will also plot the results after each processing step. 4.0 Digitized Records ====================== 4.1 From the 17jan94 Northridge CA EQ ====================================== 4.1.1 Uncorrected data This preliminary data is made available with the understanding that its useful frequency range might be narrower than that of subsequent corrected data. There has been no high-frequency filtering or instrument correction, and no filtering for long- period noise removal. Initial inspection indicates that the high- frequency content of these uncorrected time series is satisfactory to about 15 Hz, and that the long-period content is generally good to about 10 seconds for records within an epicentral distance of 35km. a. The currently available records are all located under the directory 940117:1 b. All SMA-1 records are located in the sub-directory sma-1. c. All CRA-1 records are located in the sub-directory cra-1. d. The indices which cross-reference the station names with their respective file names are in 940117:1 as files sma-1.index, and cra-1.index e. At the present time, the uncorrected records from stations within an epicentral distance of 120km are available. A few low- amplitude records in this category were not digitized. As processing continues, we will make corrected acceleration, velocity, displacement, and response spectra available also. f. All data is digitized at 600 or 1200 sps but is output here evenly sampled at 200sps. g. Comments on the Uncorrected Data (updated 12/13/94): g.1 General: The sensitivity for the digitized amplitudes is 1200 dots per inch. Amplitude steps of this size become visible on a computer plot when low-amplitude records are magnified sufficiently. For example, the step amplitude measures 1 mm on the plot if the amplitude scale is chosen at 0.012 g per cm, that is, 12 cm/s/s per cm. g.2 Specific comments on individual structures, records, or traces: 1045b.smc Diemer Filter Plant, Reservoir, Roof: A prominent 16 Hz frequency is present in the vertical component. 1069, 1070, 1071 Whittier, 7215 Bright Ave, Basement, 5th and 10th floors: Long-period P waves are evident in the horizontal components, prior to the S-wave arrival, more prominently in the basement record than at upper levels. 1276 Santa Susana, ETEC Building 462, 6th floor: At triggering, the film transport is faulty, and the first 1/4 second may eventually be deleted. 1278c.smc Santa Susana, ETEC Bldg 026, Ground level: The 3rd trace, 235 deg, was faulty and was not digitized. 1453 Malibu Canyon, Monte Nido Fire Station, Ground level: Prominent high frequencies, up to 20 and 25 Hz, are present on all 3 components. 2019 and 929 Los Angeles, 1955 1/2 Purdue Ave, 1st and 3rd levels: The vertical component on each of these records is impossible to follow and is not included. The two horizontal components on each record are difficult to follow and to check. The third component peak value on the 3rd level record conflicts with the hand-scaled value. Very high frequencies affect the accuracy of the digitization of the normal frequency content. These two instruments are above the base isolators installed on this structure. 384c.smc Carbon Canyon Dam, Right Abutment: The 3rd component has a faulty sensor; the 25 Hz content needs to be removed. 386a.smc Brea Dam, Crest: The digitized peak value of 0.24 g conflicts with the hand-scaled value on p 14 of OFR 94-141. 419a.smc Norwalk, 12400 Imperial Highway, North Ground Site: The 1st component has a faulty sensor; the 25 Hz content needs to be removed. 4222 and 6697 San Juaquin Reservoir, Left Abutment and Crest: The peak accelerations listed in OFR 94-141 are interchanged. 4980c.smc Wadsworth VA Hospital, North Ground Site: A long-period P-wave or ground tilting is evident in the 3rd component, prior to the S-wave arrival. 6756b.smc Jensen Filter Plant, Reservoir roof: The vertical component has a digitized peak value of 644.55 cm/s/s, conflicting with the hand-scaled value of 0.51 g (p 7 of OFR 94-141). 6757 Jensen Filter Plant, Generator Building, Ground level: A disturbance of the instrument, possibly not seismic, at 6.7 s after triggering, may affect processing accuracy. The vertical component peak conflicts with the hand-scaled value. 259 Jensen Filter Plant, Administration Building, Basement: The first horizontal component (22 deg) contains a possible 6-second, long-period P-wave arrival, or a ground tilting, prior to the S- wave arrival. 823 Norwalk, 12400 Imperial Highway, South Ground Site: This record was not available for listing in OFR 94-141 h. ERRATA: June 29, 1995: Station Number 5246, Newport Beach, 840 Center Drive had incorrect orientations in file = cra-1.index. This has been corrected, and subsequently revised again. CH# OLD VALUE NEW VALUE REVISED VALUE 1 360 020 20 2 UP UP UP 3 090 110 110 4 360 110 110 5 360 020 20 6 090 110 290 7 090 110 110 8 360 020 20 9 090 110 110 10 360 110 110 11 270 020 290 12 270 110 20 4.1.2 Corrected data General Preliminary tests of noise levels on a representative set of records indicated that a 10-sec long-period cut would be suitable for stations out to 35 km, an 8-sec cut for 50 km, 6 sec for 75 km, 4 sec for 85 km, and 3 sec for the furthest digitized records at 120 km. 4.1.2.1 Specific comments on individual structures, records, or traces: 1453a, b, c: Malibu Canyon, Monte Nido Fire Station, Ground level: Long-period noise at the period of the 10-sec filter forced a reprocessing with a cut at 4 sec. 3822a, b, c: Los Angeles, Griffith Observatory, Ground level: Reprocessed at a long-period cut of 6 sec. 931a, b, c: Los Angeles, 4929 Wilshire Blvd., Roof (11th level): Long-period noise forced reprocessing at 4 sec. 923a, b, c: Los Angeles, 1526 Edgemont St., Roof (8th level): Noise with an 8-sec period forced reprocessing at 6 sec. To ensure consistency of the long-period content in the following records, namely, 1111 Sunset, 333 S Hope, 500 and 520 S Grand, and 1150 S Hill, all in Los Angeles, they were reprocessed at 6 sec. Their instrument number file names commence with 1074, 1075, 1076, 1631, 3844, 6316, and 7194. To ensure consistency of the long-period content in the records at Brea Dam and the Orange County Reservoir, they were reprocessed at 3 sec. Their instrument number file names commence with 385, 386, 387, 1046, and 6696. The following structures, records or traces are not yet included because of high-frequency jolts and/or difficulty with terminating digitization while amplitudes are still high: 6146: Burbank, 3601 W Olive, Roof (9th level); 929, 2019, and 3914: Los Angeles, 1955 Purdue Ave; 5701: Los Angeles, 444 S San Vicente, Roof (12th level); 2691a, b, c: Los Angeles, 2005 N Highland Ave, Roof (8th floor): A serious but unidentified incompatability between the 10-sec filter and all three traces forced reprocessing at 1 sec. The vertical component peak acceleration is totally controlled by a high- frequency spike, and the record consequently is not yet included. 6063, 6064, 6065: Los Angeles, 1100 Wilshire, Bsmt instruments; 4224: Los Angeles, 6101 Century Blvd, 15th level; 1069, 1070, 1071: Whittier, 7215 Bright Avenue, 3 instruments; 824, 2218: Norwalk, 12440 Imperial Highway, Gd and Bsmt; 4.2 (no other data available here yet) ======================================= 5.0 Other Sources of Strong-Motion Data ======================================== 5.1 USGS Strong-Motion CD-ROM ========================= A USGS CD-ROM that includes all available ground-level, uncorrected, digitized (or digitally recorded) strong-motion records from North America written prior to the end of 1986 is available from the USGS Information Services (1-800-USA-MAPS; the same organization from which printed user's manuals for the BAP software are available.) The CD-ROM includes records retrieved by the USGS and many other agencies. The format used for the files on the CD-ROM is consistent with the strong-motion time-series files available from the "nsmp.wr.usgs.gov" computer. More information about the CD-ROM and how it can be acquired is given in the "discinfo.txt" file in the anonymous FTP home directory on "nsmp.wr.usgs.gov". 5.2 Surfing the Internet for Strong-Motion Data ========== David Wald of the USGS has a WWW page which lists numerous strong-motion resources (databases, data sources and/or archives, and products and information. His web site can be viewed online at http://www-socal.wr.usgs.gov/smdata.html.