nisee National Information Service for Earthquake Engineering
University of California, Berkeley

Seismicity in the Southern Santa Cruz Mountains in the 20-Year Period Before the Earthquake.

J. A. Olson, USGS, Menlo Park, California.

Examined are the spatial distribution of the well-located earthquakes in the twenty years preceding the Loma Prieta earthquake and their association with faults along a 100 km-long extent of the Southern Santa Cruz Mountains.

I. Summary

   Several files are archived here.  The principal data files are master event
files from which the other data files were derived.  These master files, named 
"archive" files (described below), are primarily earthquake and quarry blast 
hypocenter and traveltime information for the period 1969 through 1991 located 
by the U.S. Geological Survey's Northern California Seismic Network (Calnet).  
All events in the Calnet catalog within a 100-km-long by 30-km-wide rectangle 
which is oriented northwest-southeast and approximately centered on the Loma 
Prieta main shock are included in the "archive" data files.  The corner 
coordinates of this rectangle are:

        Latitude        Longitude
        deg (N) min     deg (W) min
        37      14.00   122     20.00
        37      25.30   122      6.20 
        36      48.70   121     18.00
        36      37.10   121     32.20
   
   The Calnet catalog event recording, processing and location methods are
summarized in Oppenheimer, Klein and Eaton (1992), which includes many
references for further, specific information about the catalog data.  The 
events are located by Calnet using the FORTRAN computer program HYPOINVERSE 
(Klein, 1989).  The "archive" files archived here (files with suffix "ARC") are
in the HYPOINVERSE output "full archive" format (Klein, 1989; p. 44) and each of
the parameters listed in those files is described in detail in that report.  
Briefly, the data in the "archive" files for each event consist of one line 
which summarizes the hypocenter information (HYPOINVERSE "summary" output 
format of Klein (1989; p. 42) including the event's date, time, magnitude, 
location, etc., followed by many lines which list the information obtained 
from recordings at individual stations including the station code name, arrival 
times, first motion polarity, coda duration, etc.  For convenience, these 
"archive" files are divided into five-year periods except for the period 1989 
through 1991.  The file names are:
        (1) LP69to73.ARC
        (2) LP74to78.ARC
        (3) LP79to83.ARC
        (4) LP84to88.ARC
        (5) LP89.ARC 
        (6) LP90to91.ARC
The corresponding hypocenter file with all the Calnet catalog data within the
rectangle described above is called:
            LP69to91CN.SUM.
   
   Other data archived here include hypocenter and fault plane solution data
for subsets of the Calnet catalog data (the master "archive" files) which are 
reported by Dietz and Ellsworth (1993) and Olson and Hill (1993).  The location
and magnitude determination methods and/or selection criteria used to obtain 
these subsets are described below.  The main difference between the hypocenter
data reported in those papers is that Olson and Hill (1993) used the Calnet
catalog hypocenters and Dietz and Ellsworth (1993) relocated those hypocenters
with a slighty different velocity model.  The hypocenter data are in the 
HYPOINVERSE output "summary" format (Klein, 1989; p. 42) and these files have 
the suffix "SUM", the same format as the first line for each event in the 
"archive" files.  The fault plane solution data were determined with the 
FORTRAN computer program FPFIT by Reasenberg and Oppenheimer (1985) using the 
first-motion polarities from the Calnet catalog data in the "archive" files.  
These data are in the HYPO71 "extended summary" format (Reasenberg and 
Oppenheimer, 1985; p.13), which includes the dip angle, dip direction and rake 
for one of the nodal planes and other parameters of the nodal plane 
determinations in the columns following the hypocenter information.  The 
hypocenter and fault-plane solution data files archived here are:

        (1) LP69to91DE.SUM (hypocenter data reported in Dietz and Ellsworth, 1993)
        (2) LP89to91DE.FPS (fault-plane solution data reported in same)
        (3) LP69to89OH.SUM (hypocenter data reported in Olson and Hill, 1993)
        (4) LP88to89OH.FPS (fault-plane solution data reported in same)

                       
II. Data Files from Dietz and Ellsworth (1993)

A. Location Method

Hypocenters from Dietz and Ellsworth (1993) were derived by relocating the
Calnet archive data (described above) with HYPOINVERSE using the velocity 
model described in Dietz and Ellsworth (1993).  To account for geologic 
differences across the San Andreas fault, the velocity model uses a separate 
1-dimensional velocity structure for the crust on each side of the fault.  
Velocities within each layer are constant. The focal mechanisms described below 
(section II.D) were calculated by FPFIT from Calnet first motions having 
takeoff angles derived from this model. 

B. Magnitude Determination

Magnitudes (MF) reported in Dietz and Ellworth (1993) were calculated by
HYPOINVERSE from measurements of record duration using the method and site
corrections developed by Eaton (1992).

C. Selection Criteria for Hypocenters

1. Area: Polygon with coordinates:

        Latitude        Longitude    
        deg (N) min     deg (W) min    
        37      20.00   122     10.00
        37      20.00   122      2.00
        36      57.20   121     29.40
        36      45.00   121     29.40
        36      45.00   121     43.00
        37       3.70   122     10.00
2. Time Period: January 1, 1969 through October 17, 1991 (G.m.t.)
3. Magnitudes: duration magnitudes (MF) 1.0 and greater
4. RMS (root mean square traveltime residual): 0.15 s and less
5. ERH (horizontal error): 1.0 km and less
6. ERZ (vertical error): 2.0 km and less
7. Exclusion of known and suspected quarry blasts:
   All events labeled as quarry blasts by Calnet ("Q" in col. 77) were excluded
   from LP69to91DE.SUM.  Additionally, all hypocenters near two known quarry
   sites were screened carefully for unlabeled blasts.  The area screened 
   around each site encompasses the epicenters of labeled blasts from that
   site. Those events which fell into the range of depth, magnitude, and
   time-of-day of known quarry blast hypocenters were labeled as suspected
   quarry blasts and were excluded from the summary file.  The screened areas
   are defined by these coordinates:

        Latitude        Longitude    
        deg (N) min     deg (W) min    
   (1)  37      23.00   122      1.00
        37      16.00   122      1.00
        37      16.00   122     10.00
        37      23.00   122     10.00
  
   (2)  36      54.50   121     36.00
        36      53.00   121     36.00
        36      53.00   121     38.50
        36      54.50   121     38.50
 
D. Selection Criteria for Hypocenters with Fault-Plane Solutions:

1. Area: same as section II.C.1.
2. Time Period: October 18, 1989 through October 17, 1991 (G.m.t.)
3. Magnitude: MF 1.5 and greater
4. RMS, ERH, ERZ: same as sections II.C.4-7.
5. Number of First-Motion Readings: 30 and greater
6. Fj (FPFIT misfit value): 0.15 and less
7. Exclusion of multiple fault-plane solutions:
   In the cases where FPFIT returned multiple solutions for one event, all
   mechanisms for that event were examined visually. A single "best" solution
   was chosen and saved in LP89to91DE.FPS for each event. The decision about
   which mechanism to save was based on comparisons of factors such as the
   distribution of arrivals about the nodal planes, the weights of misfit
   arrivals, the estimated errors in the nodal plane orientations, etc. 


III. Data Files from Olson and Hill (1993)

A. Location Method

   Hypocenters reported in Olson and Hill (1993) are those in the Calnet
   catalog as described in Oppenheimer et al. (1992).  These data, located in
   the southern Santa Cruz Mountains area, were located with a model (Olson
   and Hill, 1993, fig. 3) based on the model of Dietz and Ellsworth (1990).  
   The main difference between these models is that the Calnet model has a 
   linear velocity gradient within each layer and the Dietz-Ellsworth model has 
   homogeneous layer velocities.

B. Magnitude Determination

   The determination of magnitudes for Calnet catalog hypocenters is described
   in Eaton (1992).  The hypocenter data archived for Olson and Hill (1993)
   uses the Calnet magnitude determinations with one exception: the magnitudes 
   for hypocenters before the Loma Prieta main-shock have values for U.C. 
   Berkeley amplitude magnitudes (MX) substituted in the columns for duration 
   magnitude (MF) values (cols. 68-69) for events with MF 4.0 and greater; in 
   these cases, the MX values more closely approximate ML (J.P. Eaton, oral 
   commun.).  These substitutions were made merely to facilitate plotting.

C. Selection Criteria for Hypocenters

1. Area: same as rectangle described for selection of archive files (section I)
2. Time Period: January 1, 1969 through December 31, 1989 (G.m.t.)
3. Magnitudes: all magnitudes in Calnet catalog
4. RMS (root mean square traveltime residual): 0.20 s and less 
5. ERH (horizonal error): 1.0 km and less
6. ERZ (vertical error): 2.0 km and less
7. NST (number of stations): 6 and greater 
8. Exclusion of hypocenters near quarry blast sites:
   All events identified by Calnet to be quarry blasts (events with "Q" in
   column 77) were omitted from "LP69to89OH.SUM".  In addition, all remaining 
   events near quarry blast sites were excluded from that file, thus, if 
   earthquake hypocenters were near the site, these were excluded as well as 
   those of quarry blasts.  The size of the area near each quarry site within 
   which hypocenters were excluded depended on the hypocenter locations of 
   known quarry blasts, as identified by Calnet readers (events with "Q" in 
   col. 77).  There were five such sites in the 100-km-long rectangle described
   above (section I)with hypocenters near them.  In each case, hypocenters were excluded 
   within a polygon, the shape and size of which were chosen on the basis of 
   the location of hypocenters of known quarry blasts.  The depths beneath 
   these polygons within which hypocenters were excluded were also chosen on 
   the basis of the depths of known quarry blast hypocenters.  These depth 
   ranges were 0-5 km, with the exception of quarry number 2, below, in which 
   case there were no nearby faults or suspected earthquake hypocenters and all
   events beneath the polygon were excluded.  The area for each site within 
   which hypocenters were excluded are described by the following polygons:

        Latitude        Longitude    
        deg (N) min     deg (W) min    
   (1)  37       9.31   121     59.80    
        37       9.51   121     51.84
        37      17.96   121     44.30
        37      19.23   121     53.89

   (2)  36      42.05   121     39.64     
        36      42.65   121     32.52
        36      44.98   121     32.50
        36      48.59   121     35.11
        36      47.76   121     40.13
        36      45.23   121     41.20

   (3)  36      52.36   121     40.05     
        36      52.34   121     34.76
        36      55.73   121     35.01
        36      55.75   121     39.77

   (4)  37      17.10   122      8.81     
        37      16.89   122      2.41
        37      20.94   122      1.61
        37      20.94   122      9.07
   
   (5)  36      45.81   121     29.86     
        36      46.01   121     26.96
        36      48.96   121     26.67
        36      49.41   121     30.88

D. Selection Criteria for Hypocenters with Fault-Plane Solutions

1. Area: Lake Elsman area - rectangle with coordinates:

        Latitude        Longitude
        deg (N) min     deg (W) min
        37       8.98   121     58.93
        37      11.28   121     57.07
        37       8.50   121     51.58
        37       6.18   121     53.45
2. Time Period: June 27, 1988 through September 30, 1989 (G.m.t.)
3. Magnitude: MF 1.5 and greater (see "magnitude determinations", section III.B)
4. Number of First-Motion Readings: 15 and greater
5. RMS, ERH, ERZ: same as section III.C.4-6.
   
IV. References Cited

Dietz, L.D., and Ellsworth, W.L., 1990, The October 17, 1989 Loma Prieta,
  California, earthquake and its aftershocks; geometry of the sequence from
  high-resolution locations: Geophysical Research Letters, v. 17, no. 9, 
  p. 1417-1420.

Dietz and Ellsworth, 1993, Aftershocks of the 1989 Loma Prieta, California
  earthquake and their tectonic implications: U.S. Geological Survey
  Professional Paper I - Earthquake Occurrence, Chapter on Aftershocks
  and Postseismic Effects, P.A. Reasenberg, ed.

Eaton, J.P., 1992, Determination of amplitude and duration magnitudes and site
  residuals from short-period seimsographs in northern California: Seismological
  Society of America Bulletin, v. 82, no. 2, p. 533-579.

Klein, F.W., 1989, User's guide to HYPOINVERSE, a program for VAX computers
  to solve earthquake locations and magnitudes: U.S. Geological Survey Open-
  File Report 89-314, 58 p.
 
Olson, J.A., and Hill, D.P., 1993, Seismicity in the southern Santa Cruz
  Mountains in the 20-year period before the earthquake, in The Loma Prieta, 
  California Earthquake of October 17, 1989: U.S. Geological Survey Professional
  Paper I - Earthquake Occurrence, Chapter C. - Preseismic Observations, M.J.S.
  Johnston, ed. 

Oppenheimer, D.H., Klein, F.W., and Eaton, J.P., 1992, The first 20 years of
  CALNET, The Northern California Seismic Network: U.S. Geological Survey Open-
  File Report 92-209, 33 p.

Reasenberg, P., and Oppenheimer, D.H., 1985, FPFIT, FPPLOT, and FPPAGE, FORTRAN
  computer programs for calculating and displaying earthquake fault-plane 
  solutions: U.S. Geological Survey Open-File Report 85-739, 109p.

Return to Table of Contents