next up previous contents
Home: Berkeley Seismological Laboratory
Next: Outreach Activities Up: Operations Previous: Data Archive and Distribution:

Network Data Analysis

Lind Gee, Doug Neuhauser, Steve Fulton, Rick McKenzie, Asya Kaverina, Doug Dreger, Hrvoje Tkalcic, Bob Uhrhammer


Rapid Earthquake Data Integration System

Rapid access to reliable information is critical for any emergency response effort. In the case of a major earthquake, the mobilization of local, state, and federal disaster operations can be greatly enhanced by dependable, near real-time estimates of location, magnitude, mechanism, and extent of strong-ground shaking. This information can be used to identify endangered communities, to evaluate the impact on lifelines, and to provide input for damage and loss estimation programs. Current applications of rapid earthquake information include the emergency services, transportation, utilities, telecommunications, and insurance industries.

Over the last 6 years, the BSL has invested in the development of the hardware and software necessary for a rapid earthquake notification system. The Rapid Earthquake Data Integration (REDI) project is a research program at the BSL for the rapid determination of earthquake parameters with three major objectives: to provide near real-time locations and magnitudes of northern and central California earthquakes; to provide estimates of the rupture characteristics and the distribution of ground shaking following significant earthquakes, and to develop better tools for the rapid assessment of damage and estimation of loss. A long-term goal of the project is the development of a system to warn of imminent ground shaking in the seconds after an earthquake has initiated but before strong motions begin at sites that may be damaged.

Current Status

On April 18, 1996, the BSL and the USGS announced the formation of a joint notification system for northern and central California earthquakes (Gee et al., 1996b). The system merges the programs in Menlo Park and Berkeley into a single earthquake notification system, combining data from the NCSN and the BDSN. On the USGS side, incoming analog data from the NCSN are digitized, picked, and associated as part of the Earthworm system (Johnson et al., 1995). Preliminary locations, based on phase picks from the NCSN, are available within seconds, based on the association of a few arrivals, while final locations and preliminary coda magnitudes are available within 2-4 minutes. Earthworm reports events - both the "quick-look" 25 station hypocenters (without magnitudes) and the more final solutions (with magnitudes) to the Earlybird alarm module in Menlo Park. This system sends the Hypoinverse archive file to the BSL for additional processing, generates pages to USGS and UC Berkeley personnel, and updates the northern California earthquake WWW server.

On the UC Berkeley side, the Hypoinverse archive file is used to drive the REDI processing system. This is a modification of the original REDI design, which identified and located events using raw phase data from the BDSN and NCSN (Gee et al., 1996a). In the revised REDI processing, the magnitude of each hypocenter is assessed. If the coda magnitude is greater than or equal to 3.0, waveforms from the BDSN are analyzed to estimate local magnitude. Once an updated magnitude is obtained, the event information is paged to USGS and UC Berkeley personnel, notification is sent to emergency response agencies, and the revised magnitude is transmitted to the Earlybird system in Menlo Park. The earthquake is then evaluated for the next level of REDI processing. If the local magnitude is greater than 4.5, waveforms from the BDSN strong-motion instruments are analyzed to determine peak ground acceleration, velocity, and displacement and to estimate the duration of strong shaking. When the strong-ground motion processing is complete, these values are distributed by email and pager and the event is scheduled for moment tensor estimation. In this stage of REDI processing, both the waveform modeling method of Dreger and Romanowicz (1994) and the surface wave inversion technique of Romanowicz et al. (1993) are run for every qualifying event (earthquakes with ML greater than 3.5). Each algorithm produces an estimate of the seismic moment, the moment tensor solution, the centroid depth, and solution quality. The REDI system uses the individual solution qualities to compute a weighted average of moment magnitude, to compare the mechanisms using normalized root-mean-square of the moment tensor elements (Pasyanos et al., 1996), and to determine a "total" mechanism quality.

In the current implementation, the resulting estimate of moment magnitude is used in preference to local magnitude when the solution quality exceeds $50\%$. The moment tensor is distributed automatically only when the correlation between the two methodologies exceeds 0.5 and the solution quality is greater than $50\%$. The moment magnitude information is routinely distributed as part of the REDI system, while the mechanism information is distributed to individuals such as David Hill (in order to assist with the rapid assessment of the Long Valley status) and Jack Boatwright (for his development of methodologies for predicting strong ground shaking) of the USGS and to the OES. Analysts at the BSL review the automatic moment tensor solutions within minutes following an earthquake and distribute updated information.

At present, two Earthworm-Earlybird systems in Menlo Park feed two REDI processing systems at UC Berkeley (Figure 8.1). One of these systems is the production or paging system; the other is set up as a hot backup. The second system is frequently used to test new software developments before migrating them to the production environment. In addition, the BSL operates a third system, which uses BDSN picks to form an independent list of associated events. This third system provides redundancy in case the communication links with the USGS Menlo Park are disrupted. A fourth system, discussed below, is installed in Sacramento (Figure 8.2) in order to provide a redundant notification facility outside of the Bay Area.

  
Figure 8.1: Schematic diagram illustrating the connectivity between the real-time processing systems at the USGS Menlo Park and UC Berkeley. The combined system forms the joint earthquake notification project in Northern California. Courtesy of L. Gee.
\begin{figure}
\begin{center}
\epsfxsize=7cm \epsfbox{lsg2.ps}\end{center}\end{figure}


  
Figure 8.2: Map illustrating the distribution of BDSN stations (squares) used in REDI processing at UC Berkeley and at the redundant processing facility in Sacramento. 6 stations transmit data to both sites. Courtesy of L. Gee.
\begin{figure*}
\begin{center}
\epsfig{file=lsg1.ps}\end{center}\end{figure*}

This structure has greatly expedited automatic earthquake processing in northern California (Table 8.1). The dense network and Earthworm-Earlybird processing environment of the NCSN provides rapid and accurate earthquake locations, low magnitude detection thresholds, and first-motion mechanisms for smaller quakes. The high dynamic range dataloggers, digital telemetry, and broadband and strong-motion sensors of the BDSN and REDI analysis software provide reliable magnitude determination, moment tensor estimation, peak ground motions, and source rupture characteristics. Robust preliminary hypocenters are available about 25 seconds after the origin time, while preliminary code magnitudes follow within 2-4 minutes. Estimates of local magnitude are generally available 30-120 seconds later, and other parameters, such as the peak ground acceleration and moment magnitude, follow within 1-4 minutes. Earthquake information from the joint notification system is distributed by pager, e-mail, and the WWW. The first two mechanisms "push" the information to recipients, while the current Web interface requires interested parties to actively seek the information. Consequently, paging and, to a lesser extent, e-mail are the preferred methods for emergency response notification. The Northern California Web site has enjoyed enormous popularity since its introduction and provides a valuable resource for information whose bandwidth exceeds the limits of wireless systems and for access to information which is useful not only in the seconds immediately after an earthquake, but in the following hours and days as well.


 
Table 8.1: Current REDI processing stages. The location procedures in Stage 1 are skipped in normal operation of the joint system, but are utilized if problems prevent the flow of hypocenter information from the USGS. The items in italics are planned extensions to the REDI system.
Stage Qualification Processing
1 All events Location
2 M > 3.0 Local magnitude, Energy magnitude
3 M > 4.5 Peak ground motion, Centroid estimation
4 M > 3.5 Moment tensor
5 M > 6.0 Line source finite fault
6 M > 6.0 2D finite fault
 

System Performance

The August 12, 1998 earthquake near San Juan Bautista provides a simple example of the joint notification system (Uhrhammer et al., 1998). Although this is relatively minor event (5.3 Md, 5.3 Ml, 5.2 Mw; the only reported injuries were a teenager falling from a bunk bed and a painter falling from a ladder (San Francisco Chronicle, 8/13/98)), it is notable as the first event for which the REDI system paged strong-ground motion data.

The first alert of the event at the BSL came when the threshold signal detector at the station BKS triggered (except, of course, for those personnel who felt the earthquake) at 22 seconds after the origin. The BSL monitors the threshold detector at BKS as a preliminary indicator of an event of interest. Triggering at a velocity of 10 microns/sec, the "seismic alarm" notifies BSL personnel by pager and serves as a "wake-up call" (literally, in this case) that an event is in progress. This seismic alarm page is one example of continuous data processing at the BSL.

A "quicklook" location from the Earthworm system, utilizing data from the Northern California Seismic Network at the USGS Menlo Park, was produced within 30 seconds, based on P-wave detections at the nearest 25 stations. A final automatic solution with a coda magnitude of 5.3 was transmitted to the REDI processing system at UC Berkeley 217 seconds at the event occurred and posted on the Web page.

The REDI processing system utilizes data from the Berkeley Digital Seismic Network for the determination of local magnitude, ground motions, and the seismic moment tensor. Using the preliminary magnitude as a guide, the REDI system processed BDSN waveforms from 14 stations to determine a local magnitude of 5.4. The Earthworm location and the REDI local magnitude were paged 272 seconds after the event and the magnitude information was transmitted to the USGS. Data from BDSN strong motion instruments were processed and peak ground accelerations that exceeded 0.5% "g" were distributed by pager (12% g was observed at the BDSN station SAO, which is located  1 km from the epicenter). In the final step of automated processing for this event, the seismic moment tensor and moment magnitude were determined. The automatic moment magnitude of 5.1, the strike-slip mechanism (130/ 89/-164) and the centroid depth of 8 km compare extremely well with the compare extremely well with the reviewed solution of magnitude 5.1, mechanism (129/ 85/-171) and depth of 8 km. All REDI processing was complete a total of 11 minutes after the origin of the event.

Although a small event, this earthquake was widely felt. And the automatic earthquake information produced by the UCB/USGS system provided valuable information to the emergency response managers. For example, the Peninsula Commute Service, which operates train on the San Francisco Peninusla was able to decide that only 1 train of the 14 running from Gilroy to San Francisco needed to be stopped while the track was inspected.

Local and Teleseismic Picker

We are in the final stages of the development of a real-time phase picking algorithm using 3-component broadband waveforms from the BDSN. Recent efforts have focused on improvements to the picker to enhance the performance of the association algorithm. We have encountered two problems in this area. First, the association algorithm can be confused by a multiple P-picks at the same station. In this case, the picker "repicks" immediately following a phase detection. This is a well-known problem and one end-member solution is to reject redetections within some fixed length window following a pick. This approach is typically employed in event detectors, such as the Murdock-Hutt-Halbert algorithm (Murdock and Hutt, 1983) installed on the Quanterra dataloggers. We wish, however, to retain sensitivity to the situation where a small event preceeds a larger event and to avoid the problem of "event echos" which can occur when the detection algorithm turns back on. Second, in view of the sparseness of the BDSN, S-wave picks are critical for the location of events outside the network.

In order to address these problems, we have reformulated the characteristic function to include a polarization term (e.g., Montalbetti and Kanasewich, 1970). We have also continued to experiment with filters. After some trial and error, we have thrown away the bandpass filter that formed the basis of our previous work. Although the filter enhanced the signal strength, particularly for small events, we have found that the polarization term compensates for this loss with fewer problems from crustal P-waves.

We are currently using a characteristic function of the form:

$f(t) = av(t)^2 + b(\delta{v(t)}/\delta{t})^2 +
c(\delta{\theta(t)}/\delta{t})^2 +
d(\lambda(t))^2 $

where v(t) is the velocity, $\theta(t)$ is the change of the solid angle of the velocity vector, $\lambda(t)$ is maximum eigenvalue from the polarization decomposition, and a, b, c, and d are weighting factors. A high-pass filter at 100 sec is applied to the data in order to remove long-period noise from the timeseries before forming the characteristic function.


  
Figure 8.3: Figure comparing the results of different versions of the picker algorithm, plotted as travel time versus distance, for 128 earthquakes, and the UC Berkeley analyst's picks for the same events. A) The original characteristic function combined with a bandpass filter. B) The original characteristic function without the bandpass filter. Notice the the decrease in the number of P and S picks, as well as the decrease in the "post-P" detections. C) The new characteristic function without the bandpass filter. D) UC Berkeley analyst's picks for this data set. Courtesy of S. Fulton.
\begin{figure*}
\begin{center}
\epsfxsize=14cm \epsfbox{lsg4.ps}\end{center}\end{figure*}

We are currently running the picker as part of a REDI environment on a test system. The picker is feeding the association algorithm and we are "tuning" the process. The binder algorithm was designed for a dense network such as the NCSN (Johnson et al., 1995) which does not typically observe S phases. Over the last few years, we have gained some experience in using binder with the Murdock-Hutt event detections from the BDSN. However, we are currently experimenting with the balance in weighting S arrivals relative to P and incorporating the "quality" of the arrival. Since the day-to-day seismicity does not, in general, provide us with events of the necessary size for testing, we have developed the capability to feed archive waveform data into the picker.

We have also examined the picker's ability to estimate the azimuth of an incoming arrival based on the use of the polarization filter. We have found that while the picker produces computationally stable estimates of azimuth, comparison with the observed azimuth shows an uncertainty of +/- 20o (Figure 8.4). Although most of the scatter is in the picker estimates, the analyst estimates show considerable variation. We believe that some of the scatter is due to 3-D structure, although a significant component may be introduced by the FIR filters in the Quanterra dataloggers. The FIR filters can also influence the determination of the pick time and we are currently reviewing the possibility of removing the FIR filter in near real-time.


  
Figure 8.4: Plot of azimuth residual, displaying the scatter from the picker. The "Human Residual" is the difference between the azimuth determined by an analyst and the true source-receiver azimuth while the "Picker Residual" is the difference between the picker estimate and the true source-receiver azimuth. Although most of the scatter is in the picker estimates, the analyst estimates show considerable variation. Courtesy of S. Fulton.
\begin{figure}
\begin{center}
\epsfxsize=7cm \epsfbox{lsg5.ps}\end{center}\end{figure}

Energy Magnitude

One of the important parameters generated by the REDI system is an estimate of earthquake size. The current processing provides estimates of local magnitude, based on Wood-Anderson synthetic amplitudes, and moment magnitude, based on moment tensor procedures. The problem of the saturation of the local magnitude scale is well known and we have relied on moment tensor algorithms to provide a more robust estimate of earthquake size. However, energy magnitude may a better predictor of potential damage than moment magnitude, since the energy represents the integral of the spectrum over the entire frequency band. Consequently, we are preparing to expand the capabilities of the REDI system to include energy magnitude.

We have calibrated the algorithm of Kanamori et al. (1993) for northern California. Using a suite of 152 earthquakes, we have determined estimates of seismic energy release using:

$e_{ij} = 23.6x10^5 r^2 [r_{o}q(r_{o})/rq(r)]^2 \int {\sum {v_{ij}(t)}^2 dt}$

where eij is the energy of earthquake j at station i, r is the hypocentral distance, ro is 8 km, q(r) is the attenuation curve for distance r, vij is the observed velocity, and the expression includes correction terms for the radiation pattern. We found that the revised attenuation relation of Kanamori et al. (1993) provides a good fit to the northern California data and that no addition distance correction are required (Figure 8.5) . On the other hand, significant station variations are observed. In order to reduce the influence of site effects, we fixed the energy estimate at a single station (BKS) and determined site-specific scale factors for the network: Ej = eij si where Ej is the correct estimate of the seismic energy release by earthquake j and si is the correction factor for station i.


  
Figure: Figure illustrating the deviation of $\log {E}$ as a function of distance, before (left) and after (right) the determination of station corrections. Courtesy of S. Fulton.
\begin{figure*}
\begin{center}
\epsfxsize=18cm \epsfbox{lsg6.ps}\end{center}\end{figure*}


 
Table 8.2: Station corrections for determination of energy release. For comparison, the ML station adjustments are included in the table ( Uhrhammer et al., 1996). * indicates a preliminary estimate of the adjustment.
Station si No. of obs ML adjustment
ARC 0.2808 126 +0.209
BDM 0.5177 49 -0.023 *
BKS 1.0000 127 -0.035
BRIB 0.7515 89 +0.176
BRK 1.6369 97 +0.198
CMB 2.8105 127 +0.240
CVS 1.0835 78 -0.189 *
FARB 1.9465 92 -0.150
HOPS 1.1581 114 +0.324
JRSC 0.8514 124 +0.139
KCC 3.6114 96 +0.369
MHC 0.8681 126 +0.128
MIN 0.5680 121 -0.107
ORV 3.7033 127 +0.428
PKD 0.6153 96 -0.033 *
PKD1 0.1420 27 -0.198
POTR 0.2266 41 +0.159 *
SAO 1.3975 125 +0.314
STAN 0.1668 3 -0.233
WDC 2.5639 126 +0.484
WENL 0.4048 82 +0.089 *
YBH 2.1010 126 +0.499
 

The correction factors range from 0.14 to 3.7 and correlate, in general, with the local magnitude station corrections (Table 8.2). This is a larger variation in these corrections than observed by southern California, which may reflect the greater variability of geologic conditions at BDSN sites. The coastal stations tend to have similar corrections to BKS, but the stations in the Sierra stand out with their large corrections (CMB, KCC, ORV, WDC, YBH). Figure 8.5 compares the individual estimates of energy as a function of distance with and without the station corrections. The station corrections have significantly reduced the scatter in the energy estimates, although some site effects remain.

Using southern California events, Kanamori et al. (1993) obtained a linear relationship between the $\log {E}$ and local magnitude ( $M_{E} = 0.51(\log {E} - 9.05) $). We have found that this describes the northern California events well and we plan to implement energy magnitude estimation to the current stage for local magnitude. Under this revised processing, both ML and ME would be calculated for every qualified event (earthquakes with coda magnitude greater than 3.0).

Finite Rupture Parameters

Much of the current REDI processing is based on the use of the high-frequency hypocenter location. While this is adequate for most events, a finite rupture becomes increasely important for earthquakes of interest - particular from the perspective of emergency response. We are in the process of expanding the REDI system to include procedures for the estimation of the rupture centroid as well as the determination of finite-fault parameters.

Rupture Centroid The hypocenter location estimated using high-frequency travel times generally represents the point of rupture initiation. For a finite rupture, the centroid of the faulting can provide more useful information for rapid response. One approach to this problem is to estimate the centroid location as part of the moment tensor procedures. However, in the short term, we are exploring the use of amplitude data to estimate a "strong-motion centroid" (Kanamori, 1993). This method uses the decay of amplitude with distance in order to determine the most probable event location.

We have tested the procedure on a number of magnitude 5+ events with three types of amplitude data: peak ground acceleration (PGA), peak ground velocity (PGV), and Wood-Anderson synthetic amplitude (WAS). We have experienced some difficulty in using PGA and PGV with the attenuation relations of Joyner and Boore (1981), suggesting that these relationships do not hold for moderate earthquakes in northern California (Figure 8.6), especially at distances greater than 100 km. We have had more success using the WAS for this procedure, based on the attenuation relation of Bakun and Joyner (1984). This is perhaps not surprising, since we use the Wood-Anderson amplitudes to estimate local magnitude, but we use the Nordquist nomograph rather than this analytic relationship. The TriNet project is also using Wood-Anderson amplitudes to determine the strong-motion centroid (Wald et al., 1999).

We have found that this procedure is strongly dependent on the distribution of stations. The case of the 8/12/1998 M5.3 earthquake in San Juan Bautista illustrates this effect (Figure 8.7). The centroid estimation procedure works well for the San Juan Bautista earthquake (since the nearest BDSN station is within a few kilometers). However, if the data from the closest station is removed, then a significant degradation is observed (Table 8.3). In this example, the PGA estimate of the centroid is within 2 km of the high-frequency location, while the PGV and WAS estimates are closer to the end of the aftershock zone, approximately 8 km to the northwest. When the closest station is removed, however, all three estimates move west, with the PGA estimate moving over 135 km.

Similar behavior has been observed in other recent events, such as the 08/18/1999 Bolinas earthquake where the nearest BDSN station was over 35 km away and the distribution of stations was rather one-sided. As a result, a poor estimate of the centroid is obtained.


 
Table 8.3: Centroid locations obtained from peak ground acceleration (PGA), peak ground velocity (PGV), and Wood-Anderson amplitudes (WAS) from BDSN stations. For each amplitude type, the centroid is estimated with and without the nearest station, SAO. The centroid estimate is strongly dependent on the station distribution.
Case Latitude Longitude Dist (km)  
Epicenter 36.7527 -121.4627 4  
PGA 36.80 -121.53 10  
PGV 36.80 -121.50 10  
WAS 36.80 -121.53 10  
PGA 36.05 -122.70 67  
PGV 36.75 -121.75 67  
WAS 36.80 -121.70 67  
 


  
Figure 8.6: Figure illustrating the fit of BDSN observations to attenuation relationships for peak ground acceleration, peak ground velocity, and Wood-Anderson amplitudes for two earthquakes for the M5.3 San Juan Bautista earthquake. Courtesy of L. Gee.
\begin{figure*}
\begin{center}
\epsfxsize=18cm \epsfbox{lsg7.ps}\end{center}\end{figure*}

We are currently adding the centroid estimation procedure to the existing REDI stage of strong-ground motion processing. However, the results of this alogrithm need to be treated judiciously, depending on the distribution of stations. As the density of digital instrumentation increases in the San Francisco Bay Area, this sensitivity will be reduced. However, this is likely to remain a problem in less populated areas where the density of instrumentation may not significantly increase.


  
Figure 8.7: Map displaying different centroid locations for the San Juan Bautista earthquake (circle) determined from WAS, PGA, and PGV. The filled stars are solutions obtained using data from SAO (square) while the open stars are solutions with data from SAO. Note that the PGA solution without SAO does not locate on the map. Courtesy of L. Gee.
\begin{figure}
\begin{center}
\epsfxsize=7cm \epsfbox{lsg8.ps}\end{center}\end{figure}

Finite Fault Parameterization We have made progress toward the goal of expanding REDI processing to include the estimation of finite-fault parameters. When completed, this development will add two new stages to the REDI processing structure with the goal of automatically estimating the distribution of slip on a finite-fault plane from regional-distance seismograms.

As the first step in this development, we are automating the existing codes (Dreger and Kaverina, 1999 - described in detail in the research section). Programs developed for interactive use in a research environment require significant effort to structure and harden for an operational environment and our recent efforts have focused in this direction. We have divided the finite-fault estimation procedure into 2 REDI stages. In the first stage, BDSN regional waveform data are prepared for inversion and preliminary or initial estimates of the rupture parameters are derived using empirical scaling relationships.

These empirical estimates are then checked by testing a line-source rupture either along the rake or in the horizontal plane. Studies of the Landers and Northridge earthquakes (Dreger and Kaverina, 1999) suggest that the horizontal line source may work better. The line source calculation can be computed quite rapidly and it is possible to test a range of rupture velocities as well as both possible fault planes in order to identify the causative fault. When completed, this REDI stage will yield preliminary estimates of the rupture length, width, rupture velocity, dislocation rise time, and probable fault plane. These values then become input to the next stage for refinement. Stage 5 will perform the full 2-dimensional inversion for distributed slip on both fault planes and will produce a slip map for the preferred plane.

The automation of these codes is nearly complete and we anticipate having them fully incorporated within the REDI structure before the end of 1999. These stages will be executed for earthquakes of moment magnitude 6.0 and higher.

In parallel with this development, we are also working toward the parameterization of fault rupture using GPS. As planned, this system would parallel the REDI processing and provide an independent estimate of the source parameters.

"Shake Maps"

The quantification of a finite rupture is important for emergency response as it may be used to identify communities directly affected by the faulting process. However, these parameters may be used to predict areas of strong shaking following a major earthquake. These "Shake Maps" are highly valued by emergency response personnel (Wald et al., 1999) and have become a routine operation in southern California as part of the TriNet Project. In northern California, we are working with the USGS and the CDMG to implement capabilities similar to those in southern California. The USGS has taken the lead role in adapting the southern California software to northern California (in particular, correcting for site effects and attenuation relations). At the moment, we are working to development mechanisms to import data from the REDI stage of strong-motion processing to the USGS for map generation.

In the longer term, we plan to implement the generation of shake maps as part of the REDI system. In particular, we intend to both implement the current "approved" procedures (Shake Map Ver 1.0) as well as new methods that take the finite-fault parameters estimated by the REDI system and use the theoretical Green's functions to generate synthetic seismograms for a grid of points. Maps of predicted peak ground acceleration, peak ground velocity, and spectral acceleration will be generated.

Redundant Processing

In a further step to insure that information is available following a major Bay Area earthquake, we have established a backup processing facility in the Sacramento headquarters of the California Office of Emergency Services. This site receives data from 6 BDSN stations and operates independently of the UC Berkeley notification system. Although utilizing a small subset of the BDSN, the system has the capability to detect and locate moderate-size earthquake.

To setup this redundant processing facility, we installed a Sun Ultra 140 with a Vanguard 320 FRAD. A 56-Kbps frame-relay connection was established and divided into 7 8-Kbps circuits. 6 of these circuits are used for data acquisition from the stations YBH, ORV, CMB, SAO, BKS, and HOPS, while the 7th circuit is used for a connection between the OES site and UC Berkeley (Figure 8.2). Although 8-Kpbs is a fairly "narrow" telemetry pipe (and we initially experienced some problems in establishing these circuits because of the limited bandwidth), it is sufficient to transmit continuous data at 100 sps from the strong-motion sensors, and 20 and 1 sps from the broadband seismometers installed at these sites. In effect, these 6 stations are supporting 2 data acquisition processes, one in Berkeley and one in Sacramento. The data acquisition in Sacramento does not depend on any operations in Berkeley. The OES system has been operational since late September 1998, using the same REDI software installed at UC Berkeley.

Preliminary analysis indicates that the facility is operating well. In the 4 months from September to December 1998, a number of small events and 8 earthquakes of magnitude 4 and higher were processed. In nearly all cases, the location based on the subset of BDSN stations was within a few km of the joint notification determination. Problems have been experienced with events significantly outside the network and part of this difficulty is due to the lack of S-wave picks to constrain the location outside of the network. Our work with a 3-component picker holds promise for rectifying this situation as well as providing estimates of arrival azimuth that can be used to further constrain the location.

Year 2000 Compliance

We are in the process of updating the REDI processing software for Y2K compliance. Since most of the REDI development has taken place in the last 5 years, the Y2K "problem" is limited to 3 areas.

The first and primary source of non-Y2K compliant codes was the Earthworm related software. Since Earthworm was developed initially as an exact replacement for the RTP systems at USGS/Menlo Park, the software was non-compliant from the start. The USGS installed an updated version of the Earthworm software this summer and we have modified our codes appropriately.

The 2nd problem involved some aspects of the moment tensor codes. These stemmed from issues related to naming data files and the problem has been addressed.

The 3rd - and most challenging - problem involves the standard earthquake location program used in REDI: st-relp. This program was developed on a PC-system in the 1970s and became the standard location program used at the BSL. The program was automated as part of the initial REDI development. Because of the complexities in the code, we have decided against updating the program. Instead, we are working with bw-relp - the new location program developed by Bob Uhrhammer - to replace st-relp. Bw-relp was primarily designed to incorporate the use of azimuth information for earthquake location. However, it emulates the performance of st-relp when using travel-time data alone. We will replace the st-relp elements in REDI in November with the new code and will transition to bw-relp in our standard earthquake location procedures at the beginning of the new year.

Routine Earthquake Analysis

On a daily basis, the BSL continues to locate and determine the magnitude of earthquakes in northern California and adjacent regions. As a general rule, events are analyzed if their magnitude is greater than 2.8 in the Central Coast ranges, greater than 3.0 in all of northern California, or greater than 3.8 in the bordering regions. Traditionally, these events were located using hand-picked arrival times from the BDSN stations in conjunction with P-arrival times from the NCSN using the algorithm st-relp. Over the past several years, the BSL has made a transition in the daily analysis to take advantage of the automatic processing system. As part of this transition, events which have been processed by the automatic system are not generally relocated, although phase arrivals are still hand-picked and the synthetic Wood-Anderson readings are checked. Instead, analysts are focusing on the determination of additional parameters, such as phase azimuth, and measures of strong ground shaking. In addition to the routine analysis of local and regional earthquakes, the BSL also processes teleseismic earthquakes. Taking advantage of the CNSS catalog, analysts review teleseisms of magnitude 5.8 and higher. All events of magnitude 6 and higher are read on the quietest BDSN station, while events of magnitude 6.5 and higher are read on the quietest station and BKS. Earthquakes of magnitude 7 and higher are read on all BDSN stations.

The locations and magnitude determined by the BSL are cataloged on the NCEDC. The phase and amplitude data are provided to the NEIC, along with the locations and magnitudes, as contributions to the global catalogs, such as that of the ISC.

Regional Moment Tensor Estimation

Two independent moment tensor inversion methodologies are currently being used in routine earthquake analysis at the Berkeley Seismological Laboratory: a) a regional surface wave inversion method, which is a two-step frequency domain inversion that uses 15 to 50 second surface waves and is adapted from a method originally developed for a global data (Romanowicz, 1992) and b) a regional time-domain inversion method using complete waveforms at frequencies between 0.01 and 0.10 Hz (Dreger and Helmberger, 1993). Both methods have been fully automated (e.g. Pasyanos et al., 1996) and integrated into REDI (Gee et al., 1996). Human reviewed solutions are still sent out to the community via an e-mail distribution list and also the website: www.seismo.berkeley.edu/ dreger/mtindex.html We estimated 49 moment tensor solutions for events which occurred in and around northern and central California between August 1998 and September 1999 (Figure 8.8 and Tables 8.4 and 8.5). They were determined using one of the methods mentioned above.


  
Figure 8.8: Moment tensor solutions for events which occurred in and around northern and central California between August 1998 and September 1999 determined by both moment tensor inversion methodologies. The locations of the BDSN stations are shown (squares).
\begin{figure*}
\begin{center}
\epsfig{file=hrvoje99_3_1.ps,width=18cm}\end{center}\end{figure*}


 
Table 8.4: Moment tensor solutions for significant events from August 1998 to September 1999 using both regional methodologies. Epicentral information from the UC Berkeley/USGS Northern California Earthquake Data Center. Moment is in dyne-cm and depth is in km. Key to methods: (1) Complete waveform fitting inversion; (2) Regional surface wave inversion.
Location Date UTC Time Lat. Lon. Depth Ml Mw Mo Str. Dip Rake Method
Nevada 08/26/1998 15:50:12.0 37.114 -117.712 8.0 4.0 3.7 4.05e21 341 38 -114 1
B. Columbia 08/30/1998 11:33:34.0 51.1 -130.4 10.0 6.1 6.1 1.56e25 62 79 -43 1
San J. Baut. 09/02/1998 11:24:07.0 36.839 -121.298 8.0 3.9 4.0 10.0e21 165 81 190 2
Bishop 09/11/1998 11:38:42.0 37.388 -118.689 11.0 3.8 3.8 5.80e21 334 83 169 1
Parkfield 09/16/1998 09:00:44.0 35.959 -120.508 14.0 3.5 3.5 2.17e21 154 87 178 1
Gilroy 10/10/1998 06:50:31.0 36.97 -121.58 5.0 4.0 3.8 5.51e21 296 72 172 1
          5.0 4.0 3.9 6.80e21 299 86 -167 2
Arcata 10/15/1998 05:05:46.0 40.840 -123.571 30.0 4.0 4.2 2.69e22 345 75 -74 1
          20.0 4.0 4.1 2.4e22 348 61 -79 2
Portola 10/21/1998 08:31:01.0 39.725 -120.677 11.0 4.4 4.1 1.73e22 180 88 341 1
          22.0 4.4 4.2 2.5e22 280 89 349 2
Truckee 10/30/1998 09:53:30.0 39.303 -119.977 11.0 5.3 4.8 1.77e23 30 85 -3 1
          24.0 5.3 5.0 3.5e23 32 88 -1 2
Redding 11/26/1998 19:49:53.0 40.63 -122.42 24.0 5.4 5.2 6.40e23 21 86 67 1
          24.0 5.4 5.1 4.80e23 289 77 -46 2
Offshore 11/27/1998 00:43:48.0 40.67 -125.38 5.0 5.2 5.4 1.40e24 129 58 -162 1
Berkeley 12/04/1998 12:16:07.0 37.92 -122.29 11.0 4.1 3.9 9.26e21 139 69 174 1
          20.0 4.1 4.2 2.10e22 228 90 -8 2
Mammoth 12/07/1998 16:15:54.0 37.64 -118.93 11.0 3.7 3.4 1.53e21 146 78 -27 1
Mammoth 12/14/1988 04:14:02.0 37.53 -118.798 8.0 4.0 4.0 8.10e21 222 84 17 1
          6.0 4.0 4.0 2.00e22 206 80 0 2
Mammoth 12/17/1998 10:32:12.0 37.524 -118.798 8.0 4.1 3.9 7.14e21 219 73 13 1
          20.0 4.1 4.0 2.10e22 214 75 21 2
Ben Lomond 12/29/1998 12:38:12.0 37.093 -122.041 11.0 3.9 3.9 6.93e21 149 64 106 1
          8.0 3.8 3.9 5.18e21 140 49 115 1
Tres Pinos 01/18/1999 08:48:36.0 36.648 -121.268 11.0 4.3 4.3 2.98e22 48 88 -4 1
          16.0 4.3 4.4 5.20e22 131 88 175 2
Fort Bragg 01/24/1999 15:34:56.0 39.563 -123.767 5.0 4.0 4.1 1.47e22 311 49 115 1
          4.0 4.0 4.0 2.50e22 271 27 47 2
Morgan Hill 01/27/1999 03:58:42.0 37.253 -121.638 8.0 3.8 3.7 4.04e21 324 78 -174 1
Morgan Hill 02/04/1999 00:19:36.0 37.160 -121.554 5.0 3.8 3.5 2.32e21 153 88 -166 1
          6.0 3.8 3.9 8.70e21 51 89 4 2
Morgan Hill 02/04/1999 00:21:41.0 37.162 -121.554 8.0 3.7 3.5 2.05e21 329 80 175 1
          10.0 3.7 3.8 6.20e21 142 88 173 2
Gorda Plate 02/09/1999 14:32:47.7 40.70 -124.96 18.0 4.7 4.1 1.89e22 146 73 156 1
          10.0 4.7 4.1 1.40e22 333 87 -166 2
Punta Gorda 02/15/1999 06:00:04.0 40.28 -124.39 27.0 3.7 4.7 1.25e23 101 80 170 1
          30.0 3.7 4.6 1.00e23 276 90 -178 2
Geysers 02/18/1999 08:58:41.0 38.796 -122.739 5.0 3.9 4.1 1.83e22 167 77 -161 1
          22. 3.9 4.3 3.10e22 333 72 44 2
Parkfield 02/26/1999 15:25:56.7 35.95 -120.50 11.0 4.0 4.0 1.33e22 148 88 174 1
          24. 4.0 4.2 2.30e22 243 73 -19 2
Crescent City 03/16/1999 18:06:31.3 42.062 -125.718 8.0 4.0 4.3 3.13e22 8 66 -51 1
          6.0 4.0 4.4 5.20e22 349 56 -73 2
Tres Pinos 03/23/1999 18:36:40.0 36.676 -121.300 8.0 4.1 4.1 1.72e22 310 84 139 1
Geysers 04/04/1999 06:00:36.0 38.840 -122.754 5.0 3.8 3.9 7.59e21 191 51 -111 1
Pinnacles 04/15/1999 03:20:25.5 36.583 -121.175 6.0 3.8 3.8 6.50e21 314 89 -173 2
Mammoth 04/24/1999 07:56:27.0 37.545 -118.864 8.0 3.9 3.8 2.31e21 343 66 -70 1
 


 
Table 8.5: Moment tensor solutions for significant events from August 1998 to September 1999 using both regional methodologies ( continued). Epicentral information from the UC Berkeley/USGS Northern California Earthquake Data Center. Moment is in dyne-cm and depth is in km. Key to methods: (1) Complete waveform fitting inversion; (2) Regional surface wave inversion.
Location Date UTC Time Lat. Lon. Depth Ml Mw Mo Str. Dip Rake Method
Mammoth 05/15/1999 13:22:10.0 37.533 -118.833 8.0 6.0 5.5 2.31e24 203 85 -11 1
          8.0 6.0 5.7 4.30e24 203 79 23 2
Mammoth 05/15/1999 17:54:08.0 37.507 -118.832 8.0 5.3 4.6 9.49e22 201 87 -28 1
          8.0 5.3 4.7 2.40e23 202 88 -2 2
Petrolia 05/15/1999 20:58:00.0 40.366 -125.133 11.0 4.0 4.0 1.27e22 279 82 -168 1
Mammoth 05/17/1999 06:37:19.1 37.512 -118.826 4.0 4.3 4.0 1.13e22 343 46 -116 1
          4.0 4.3 4.0 2.10e22 347 46 -97 2
Mammoth 05/25/1999 16:42:46.4 37.529 -118.817 6.0 3.9 3.8 6.20e21 7 60 -80 2
Mammoth 05/26/1999 03:53:53.4 37.556 -118.803 5.0 4.2 3.9 8.89e21 14 55 -22 1
          8.0 4.2 4.4 4.00e22 250 45 -91 2
Mammoth 05/26/1999 18:04:07.2 37.546 -118.806 5.0 4.2 3.8 4.83e21 9 84 -26 1
          4.0 4.2 3.8 6.50e21 3 69 -25 2
Mammoth 05/26/1999 21:22:28.3 37.523 -118.825 5.0 4.4 3.9 9.25e21 5 62 16 1
          6.0 4.4 3.9 9.40e21 358 88 42 2
Mammoth 05/28/1999 20:45:54.2 37.558 -118.801 5.0 3.8 3.6 3.03e21 130 80 -153 1
          4.0 3.8 3.6 2.80e21 290 86 -156 2
Petrolia 06/02/1999 21:26:30.0 40.455 -124.641 11.0 3.5 3.9 8.26e21 208 65 10 1
Mammoth 06/03/1999 21:36:27.7 37.537 -118.806 5.0 4.4 4.2 2.34e22 112 76 -135 1
          4.0 4.4 4.3 3.70e22 209 72 -142 2
Mammoth 07/05/1999 03:06:29.0 40.316 -124.528 4.1 3.7 4.1 1.42e22 129 61 115 1
Isabella 07/11/1999 18:20:46.0 35.735 -118.485 8.0 4.3 4.3 2.76e22 345 39 -93 1
          8.0 4.3 4.4 3.11e22 345 39 -92 2
Petrolia 07/24/1999 00:38:41.0 40.391 -125.127 5.0 4.4 4.5 5.65e22 359 64 -18 1
          4.0 4.4 4.6 9.40e22 217 62 -97 2
Geysers 07/29/1999 04:52:26.6 38.796 -122.733 5.0 3.6 3.7 4.30e21 341 65 -149 1
          8.0 3.6 4.4 4.00e22 359 47 -100 2
CA-NV bdr 08/01/1999 16:06:22.0 37.39 -117.07 0.0 5.6 5.7 3.60e24 64 273 39 1
          8.0 5.6 5.9 7.00e24 216 45 -92 2
CA-NV bdr 08/02/1999 06:05:13.4 37.361 -116.949 11.0 5.5 5.1 4.90E23 251 82 -44 1
          8.0 5.5 5.1 5.60e23 200 55 49 2
Bolinas 08/18/1999 01:06:19.0 37.907 -122.686 8.0 5.0 4.5 7.25e22 115 49 69 1
          6.0 5.0 4.8 2.60e23 221 51 86 2
San Simeon 09/21/1999 15:42:48.0 35.799 -121.255 8.0 5.6 3.6 2.70e21 119 72 81 1
Santa Rosa 09/22/1999 22:27:13.1 38.393 -122.633 5.0 4.2 4.0 9.68e21 51 50 27 1
          6.0 4.2 4.0 1.00e22 113 49 100 2
Nevada 09/26/1999 20:11:22.0 37.43 -117.080 8.0 4.1 4.2 2.53e22 261 72 -92 1
          6.0 4.1 4.0 20.00e21 208 61 -83 2
Petrolia 09/29/1999 06:22:03.0 41.357 -123.424 39.0 4.0 4.0 1.07e22 188 41 -95 1
          40.0 4.0 4.0 2.20e22 9 52 -78 2
 

References

Bakun, W., and W. Joyner, The ML scale in central California, Bull. Seis. Soc. Am., 74, 1827-1843, 1984.

Cohee, B., and G. Beroza, A comparison of two methods for earthquake source inversion using strong motion seismograms, Ann. Geophys., 37, 1515-1538, 1994.

Dreger, D., and A. Kaverina, Development of procedures for the rapid estimation of ground shaking, PGE-PEER Final Report, 1999.

Dreger, D., and B. Romanowicz, Source characteristics of events in the San Francisco Bay region, USGS Open-File-Report 94-176 , 301-309, 1994.

Gee, L., D. Neuhauser, D. Dreger, M. Pasyanos, B. Romanowicz, and R. Uhrhammer, The Rapid Earthquake Data Integration System, Bull. Seis. Soc. Am., 86, 936-945,1996a. Gee, L., A. Bittenbinder, B. Bogaert, L. Dietz, D. Dreger, B. Julian, W. Kohler, A. Michael, D. Neuhauser, D. Oppenheimer, M. Pasyanos, B. Romanowicz, and R. Uhrhammer, Collaborative Earthquake Notification in Northern California (abstract), EOS Trans. AGU, 77, F451, 1996b.

Johnson, C., A. Bittenbinder, B. Bogaert, L. Dietz, and W. Kohler, Earthworm: A flexible approach to seismic network processing, IRIS Newsletter, XIV (2), 1-4, 1995.

Joyner, W., and D. Boore, Peak horizontal accelerations and velocity from strong-motion records including records from the 1979 Imperial Valley, California, earthquake, Bull. Seis. Soc. Am., 71, 2011-2038, 1981.

Kanamori, H., Locating earthquakes with amplitude: application to real-time seismology, Bull. Seis. Soc. Am., 83, 264-268, 1993.

Kanamori, H., J. Mori, E. Hauksson, T. Heaton, L. Hutton, and L. Jones, Determination of earthquake energy release and ML using TERRASCOPE, Bull. Seis. Soc. Am., 83, 330-346, 1993.

Montalbetti, J., and E. Kanasewich, Enhancement of teleseismic body phases with a polarization filter, Geophys. J. R. Astr. Soc., 21, 119-129, 1970.

Murdock, J., and C. Hutt, A new event detector designed for the Seismic Research Observatories, USGS Open-File-Report 83-0785, 39 pp., 1983. Romanowicz, B., D. Dreger, M. Pasyanos, and R. Uhrhammer, Monitoring of strain release in central and northern California using broadband data, Geophys. Res. Lett., 20, 1643-1646, 1993.

Pasyanos, M., D. Dreger, and B. Romanowicz, Toward real-time estimation of regional moment tensors, Bull. Seis. Soc. Am., 86, 1255-1269, 1996.

Uhrhammer, R., S. Loper, and B. Romanowicz, Determination of local magnitude using BDSN broadband records, Bull. Seis. Soc. Am., 86, 1314-1330, 1996.

Uhrhammer, R., L. S. Gee, M. Murray, D. Dreger, and B. Romanowicz, The Mw 5.1 San Juan Bautista, California earthquake of 12 August 1998, Seismol. Res. Lett., 70, 10-18, 1999.

Wald, D., T. Heaton, and K. Hudnut, The slip history of the 1994 Northridge, CA, earthquake determined from strong motion, teleseismic, GPS and leveling data, Bull. Seis. Soc. Am., 86, s49-s70, 1996. Wald, D., V. Quitoriano, T. Heaton, H. Kanamori, C. Scrivner, and C. Worden, TriNet "ShakeMaps": Rapid generation of peak ground motion and intensity maps for earthquakes in southern California, Earthquake Spectra, 15, 537-556, 1999.


next up previous contents
Next: Outreach Activities Up: Operations Previous: Data Archive and Distribution:

The Berkeley Seismological Laboratory, 202 McCone Hall, UC Berkeley, Berkeley CA 94720
Questions and comments to www@seismo.berkeley.edu
Copyright 1999, The Regents of the University of California.