ML20128B335
| ML20128B335 | |
| Person / Time | |
|---|---|
| Issue date: | 12/29/1992 |
| From: | Zurflueh E NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | Murphy A NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| References | |
| NUDOCS 9302030007 | |
| Download: ML20128B335 (9) | |
Text
_,
fkk
.)
J DF.C-0 0 DN 1
MEMORANDUM FOR:
Andrew J. Murphy, Chief SSEB/DE/RES THRU:
Nilesh C. Chokshi, Leader i
Seismic Section, SSEB/DE/RES FROM:
Ernst G. Zurflueh, Geophysicist q
SSEB/DE/RES
SUBJECT:
FALL 1992 MEETING OF THE AMERICAN GE0 PHYSICAL UNION i
Enclosed is a report on special topics r. resented at the fall-meeting of-the American Geophysical Union which I attended from Decet.ber 7 to I
December 11, 1992. Also included in th report are notes on an ad' hoc meeti_ng for regional and local seismic networks, and on a meeting at'LLNL.concerning the seismic hazard project.
"* D id by
-.":.hth3di Ernst G. Zurflueh, Geophysicist SSEB/DE/RES
Enclosure:
As stated cc w/ encl:
R.1McMullen R. Rothman
.G. Giese-Koch
-P. Sobel.
B. Ibrahim DISTRIBUTION:- RESReading EZurflueh NChokshi AMurphy RBosnak LShao &PDREt SSEB/DE/RES EZurflgeh:fkm If>--
9302030007 921229 12/2fJ/92 ;
{4 l
z
~1 H
@', e y;
,?'
FALL 1992 MEETING OF THE AMERICAN GEOPHYSICAL UNION San Francisco, California, December 7 - 11 At this year's AGU f all Meeting a large number of interesting -papers. were presented, ano the meeting included a number of sessions on special topics, such -
as the Mendocino triple junction and the Landers -- Big Bear earthquakes. About-half the papers presented were in the form of posters.
In many instances, the posters repeated or expanded on subjects coveNd in regular oral paper sessions, thus allowing attendees to concentrate on thesa subjects in greater detail, in the session on the Mendocino triple junction, Zandt et al. presented information on the tectonics of the region based on tomography and earthquake source mechanisms.
The south edge of the Gorda plate under the 'cortinental crust is flexed downward and forms a linear continuation of the-offshore seismicity trend.. Inland from the triple junction, an area of crustal thickening occurs, and east of that area (east of 123 W) there is an area. of manti_e -
upwell ing. Other than at this-location there is little or no mantle upwelling.
The Petrolia and Honeydew earthquakes are probably related to subduction and occurted on subhorizontal faults in the accretionary wedge above the Gorda plate.
Castillo and Ellsworth consider the transform boundary in northern California to be a 100 to 150 km wide zone, which includes the Maacama-Rodgers Creek-Hayward-Calaveras fault system. The Maacama fault is a non-vertical strike slip fault that dips northeast. The stress on the Maacama is purely right lateral. -Farther south the fault system steepens and.the Hayward and Calaveras faults are nearly vertical. While the Maacama fits a strong fault model, the stress on the Hayward.
and Calaveras is about normal'to the-fault trend, consistent with a weak fault '
model.
A paper by Oppenheimer and Eaton and posters-by S. Schwartz, Velasco et al., and others, analyzed the fault motion of the Petrolia earthquake and its aftershocks.
The main shock had a subhorizontal thrust motion, although the fault orientation is still somewhat in debate, and the subsequent magnitude 6 and 6.5 events:had:
a' strike slip motion.
Lajoie compared the central Aleutian arc with the southern rascaata subduction zone-and reached the conclusion that a 240 to 280 hn serwfrom Cape Mendocino northward may act as a unit, potentially generating earthquakes of magnitude 8.25-to 8.5.
This is based on Late Cenozoic structure and deformation..The author-considers it possible that an even larger segment may have ruptured in the late:
Holocene, based on paleoseismic evidence.
In the latter case, mac" tudes-may.
have been as large as 9.
A poster by Sydnor of the California Strong Motion Program showed that the CSMIP' Cape Mendocino station can be considered a hard rock 'ite. that. has moved
~
integrally with the bedrock during the Petrolia earthquake. No-local' sources of ground motion cre, therefore, assumed to exist, and the recorded 'high acceleretions are attributed to the main earthquake source.
A poster by Atwater et al. described the results of precise tree-ring age dating.
at Humboldt Bay. The tree ring ages show that the Mad River Slough subsidence
' happened at 1690-1710. This is the same as-estimates obtained for subsidence on
-,s 3
g g-.
2 the southern coast of Washington.
The evidence, then, suggests that a giant-earthquake occurred at that time that ruptured more than 600.km of. subduction zone.
Alternatively, it is possible that. two large earthquakes occurred; separately within'about 20 years.
A-paper by Rajendran and Talwani postulated that the Bluff ton area of South Carolina contains a seismic source that is separate from the Charleston source and currently inactive. This conclusion is based on dating of paleoliquefaction features calibrated with dendrochronology and on stratigraphic evidence for-.a buried, 30 km long, east-west trending fault.
~
Kelson et al. investigated the central Reelfoot scarp in Tennessee.
Trenching has produced no evittence for a fault at the surface, whereas seismic reflection profiles show r.<erse faulting at depth.
The trench did expose liquefaction features, attributed to the 1811-1812 earthqakes, and colluvium from an earlier:
event dated at A.D. 1310 to 1540.
The recurrence time between the two most recent large earthquake episodes is, therefore, estimated to be between 200 and 600 years.
The Little Skull Mountain earthquake, a m=5.6 event.that occurred less than 20 km from Yucca Mountain, was discussed in a paper by Brockman et al. This is the largest known earthquake to have occurred in the southern Great Basin. The event is described as northeast trending normal faulting with southeast dip at a depth-of 8 to 11 km. The event did not activate a previously mapped fault, it happened 22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br /> after the Landers eartWake and was probably triggered by that event.
Aftershocks are distributed in a aortheast-southwest trending region.
A whole session on Tuesday morning was reserved for a discussion of the Landers earthquake.
A paper by Hauksson et al. served as an introduction to set the stage for the session. The Landers earthquakes form a series,- starting with-the Joshua Tree event that. happened just north of the San Andreas fault. T_he series continued with the main Landers event and with the Big Bear aftershock. This was the largest earthquake sequence in California in 40 years, it consisted predominantly of strike-slip motion, and the-three events uniform _ly ruptured northward, away from the San Andreas fault.. The numerous aftershocks form a -
trend striking north 14' west, following the: eastern California shear zone or Mojave shear zone.
tie earthquake sequence forms a link that connects the San Andreas with the shear zone.
Jackson et al. presented a paper evaluating the seismic ~ hazard in southern California, reaching the conclusion that the hazard estimate has about doubled after the occurrenc'e of the Landers earthquake sequence. This is based on stress distribution considerations and on the fact that seismicity in the area has~-
increased by approximately a factor of two since 1985.
Kanamori et al. analyzed TERRAscope data for the Landers earthquake. They also found a south to north directivity of the fault rupture.
In addition, they
- postulate 'that there were two distinct subevents about 30 km apart, with the second event.about twice the size of the first one.
They consider the-Landers event to'be a high stress drop earthquake on a fault with long recurrence: time.
- Sauber et al. have looked at geodetic measurements and have concluded that' slip'-
e
~,,,
w
3 on the northwest striking faults of the central Mojave is kinematically related to the right lateral faults bounding the region of extension in the southern Basin and Range. They also found that recurrence on one segment of the northwest trending fault system hould be in the order of 3700 to 6400 years.
With multiple segments, the recurrence may be lowered to the order of decades.
Jcume and Sykes postulate that both stress and fluid pressure are migrating towards the San Andreas fault.
The segments of the San Andreas that were thus influenced have not ruptured in a great earthquake for about 300 years.
The stress changes induced by the Landers events are equivalent to several decades of slow stress accumulation.
A Tuesday afternoon poster session was also dedicated to the Landers earthquake.
The posters included a number of papers on coseismic deformation measured by GPS.
A poster by Shen et al. analyzed post-seismic deformation. Deformation occurs at an approximately constant rate of 1 mm/ day on the Big Bear fault and the southern portion of the landers rupture zone, in the northern portion of the rupture zone, the rate is about 2 mm/ day.
In general, the post-seismic deformation mimics the pattern of coseismic deformati Post-seismic deformation amounts to about 10% of the coseismic moment release on the Camp Rock-Landers fault and 5% of costismic release on the Big Bear fault.
An interesting demonstrttion by Crippen and Blom showed a video display of SPOT images taken before and after the Landers earthquakes.
By toggling the two images, a visible display of fault deformation can be achieved.
The authors claim that in spite of a 10 m pixel size, deformations as small as 1 m can be seen.
Some critics sthted that it is very difficult to get a sufficiently accurate registration of the images to see such movement. However, it appeared to me that indeed the movement shown was different from one side of the f ault to the othe., thus substantiating the authors' claims.
A poster by Lizarralde and Holbrook on the structure of the southeastern U.S.,
based on offshore reflection profiles, showed an abrupt change in the velocity of the upper crust across the Brunswick magnetic-anomaly. Potential field data also indicate an abrupt change between the Brunswick and East Coast magnetic anomalies. The authors concluded that the crust south of the Brunswick anomaly was altered by rifting and ultramafic intrusion. An alternative explanation is that the change represents a suture zone.
A paper by Karlin and Abella described the latest results from the investigation of sediments in Lake Washington. Sediment cores were analyzed for sedimentology, geochemistry, X-ray response, and magnetic susceptibility.
These data permit correlation of layers across the entire lake down to the Mazama ash layer, which forms a marker bed 6850 years in age.
In general, logs of susceptibility and alumina content correlate, whereas carbon content is inversely correlated.
Correlation of cores demonstrates that turbidites have been deposited.
episodically throughout the lake.
These layers do not seem to have been climatically induced, but their characteristics are most consistent with slumping induced by earthquakes. Their chemistry indicates a sudden disturbance in the sedimentary balance and then a slow recovery. A prominent turbidite layer seems
'o have been deposited in reponse to a large e:rthquake 1100 years ago. Sunken forests, found at three separate locations in the lake, have also been dated at i
.e 4
1100 years.- This date corresponds to a major event found-by Atwater on~ the Washington coast.
Other possible events in the lake were-dated at. 300,1600--
1700, and 2800-3100 years ago.
Again, many of these -dates correlate with equivalent ones found by Atwater on the coast, In all,14 events' have been found in the interval from the Mazama ash to present, suggesting i. recurrence interval-of about 500 years.
A poster paper by Thio and Kanamori described a method of calculating moment tensor inversions based on surface waves recorded by TERRAscope.
The method..
works in the spectral domain, is fast, and can be used for magnitudes between'3.0 and 6.5.
Focal mechanisms, moments and local magnitudes derived from the method are consistent with those determined using first motions.and body waves.
A Thursday afternoon session again focused on the Landers earthquakes.
The general conclusion of many of. the papers was that this is the first well documented case showing triggered seismicity at large distances from the original earthquake. Reasenberg et al. reported that widespread and sudden increase in seismicity in many parts of the western U.S. establishes the fact that seismic triggering occurred at distances of 3 to 17 source dimensiois-(1250 km), whereas usually there is no effect beyond 2 source dimensions. - The Little. Skull Mountain earthquake and a Yellowstone event are among the triggered earthquakes.
Gomberg et al. analyzed the Little Skull Mountain earthquake and consider it to j
have been triggered by the Landers event. They concluded that magmatic processes.
cannot account for the triggering mechanism, but that sympathetic slip and ground
.)
water may be involved. They also noted that seismic hazard estimates must take triggering into account.
Qian and Jones looked at the effect of the Landers earthquake on the southern' San
- Andreas fault. - Except for the San Bernardino segment, stress on the' San Andreas fault has not' changed since 1986. Network data indicate that seismicity on the southern San ' Andreas increased after Landers, the largest increase having occurred on the San Bernardino segment.
Many other experts at the meeting
=t pointed out that this segment is the one associated with the highest earthquake
-hazard.
A poster by Braile et al. showed the results of 400 km of seismic profiles in the New ' Madrid area using refraction, reflection and-strong motion data.
The profiles show a crust under the Reelfont rif t that is about 42 km thick,. and the top of a 7.4 km/s layer, interpreted as a rift pillow, is clearly imaged.
Some-U fault offsets have aise been noted in the subsurface in these preliminary interpretations.
Results of a Mini-Sosie reflection survey of the Crittenden County fault. zone' were described in a poster by Nicholas e.t al.
Combined with previous surveys, 30 km of the fault zone have now been imaged; the northeast extent of the fault-zone remains unknown. The fault zone consists of a monoclinal flexure with minor a
high-angle reverse faulting. Offsets decrease southwestward.
The majority ofi
,a the structural relief was formed before deposition of the Eocene Memphis sand,.
although minor movement in the Quaternary section cannot be ruled out.
Hopkins. et al. showed results of a seismic reflection profile across the Valley.
n d
5 and Ridge province of eastern Tennessee.
The crystalline basement is highly reflective with numerous west dipping reflectors, except for non-reflective zones in the central and western portions of the profile.
Earthquake hypocenters projected onto the profile coincide mostly with the reflective basement.
The reflective crust may contain more numerous deformation features that may be activated if oriented favorably with respect to the stress field.
Honjas et al. reported results of new imaging of the nosgri fault.
They distinguish three fault planes: A central portion, dipping east, at 70 near the surface and 60 at depth; a west branch with an east dip of 60 ; and a near vertical east bre..ich.
Most of the vertical slip occurred during or before the Miocene; neotectonic motion on the Hosgri is predominantly strike slip.
{
A Quing et al. poster described studies of the Juan de Fuca plate using broadband stations in restern Oregon. They place the oceanic Moho beneath Corvallis at a depth of 40 to 45 km with a Poisson's ratio of 0.31 to 0.33, indicating ultramafic composition. The dip of the Moho near Corvallis is 10-15 east, with probably higher (20 ) dip near the Coast Range. A low velocity layer parallel to the Moho may be subducted oceanic crust.
No indication of a shallower continental Moho has been found.
A poster by Benson et al. presented age determtaations of trees living in tidal swamps of southern Washington.
These trees were damaged in a past subsidence event, the damage be:ng expressed in tree rings whose age can be counted.
The ages found for the event are slightly over 290 years, which corresponds with the prominent 300 year event found by numerous other investigators in the region.
A session on Friday morning was dedicated to results from the BASIX experiment in the San Francisco Bay region. A paper by Jones et al. showed Sierran basement underlying the region above the Moho.
In the western portion of the area, the Pacific plate is underthrust under the fold and thrust belt visible at the surface. Above the Pacific plate and the Sierran basement, there is a brittle-ductile transition zone, which is also called a ductile decollement.
Based on hypocenter plots, the authors contend that the faults in the area, including the San Andreas, are not vertical and throughgoing, but flatten out westward into the decollement zone. This general view was also supported by other papers, but some scientists disagree with this view.
According to Jones, with progressing-shortening of the crust, the faults had to be deformed also.
In general, the view is that there is some sort of subduction of the Pacific plate going on, some called it a pseudo-subduction that happens very slowly, perhaps by grain boundary recrystallization.
This is an oblique subduction with a large northward component. According to this view, the pinte boundary would be the mid-crustal decollement, not the San Andreas fault.
w o
.r Ad Hoc Meetino for Reoional and local Seismic' Networks ~
An impromptu meeting was held at the meeting location on December 8 at 5:45PM.
A meeting announcement is attached. The meeting began with'a discussion by-John:Filson of the USGS on a charter for the National Seismic System Coordinating Council (NSSCC). He mentioned'that_there is a1 Federation of Digital Seismic Networks that has operated for 3. years-now and is'only now starting to establish bylaws.
The unresolved question concerning the NSSCC is' that of how member organizations are to be represented.
Filson outlined two possible modes of operation, (a) each. institution has one representative (this is how IRIS is organized), (b) each network is represented, which-gives some institutions multiple representation. The conclusien of the meeting was that mode (a) would be simpler to implement, although that would restrict' the~ USGS to one representative. John Filson did most of the talking and thinking'on' this subject and he leans-toward the one rep esentative per institution model but seems to be still undecided because the LSGS would not have much clout in-this_ situation.
It was also said that anyone can attend meetings of this proposed organization, but that_ voting members must have a stake in networks.
Walter Arabasz proposed setting up a small working _ group to get things going.
Other questions that came up were whether a vote-by-the NSSCC would be binding on the networks and how often they should meet. Most of the-questions.were-only raised but not answered and, concerning the binding resolutions, it was suggested that a network should "try to carry out' a resolution."
The second major subject was a discussion by Al Lindh on how to distribute 4
network data in a timely fashion. He suggested that, for the time being, some data could be distributed by E-mail using-a simple keyword format.
This-would allow anyone to send data to anyone else.
The third subject was a more general discussion of what types of network data are-available and how they can be obtained.
The subject was' introduced.by Bill Ellsworth who described the new Seismic III display system.
The>systemL requires a PC with VGA and beeper.
Later, the Internet will be used-instead of a beeper. He said that the_ Federation Archive at_the' IRIS data' center in Seattle'has the codes.
They can be obtained;by E-mail from' Tim @ IRIS.EDU.
Ray Buland stated that hypocenters have been available from this -address at '
IRIS for 6 months now. 'He said that 5-letter ~ station codes can be used, but-that for international reasons 4 letters should.be used.
The question.was raised as to how data can be distributed via'~Internet, and it was stated that; Lament has a system for distributing data on this network.
Tom Hanks' suggested that the Seismic Research Letters should be 'used to distribute new data.
Someone then mentioned'that_it is easy to get hypccenters,,but not auxiliary' data, such as _ station locations and velocity structure. -Again, it appears that Lamont has-such: data available in the file pub /gb. tar.Z at' lamont.LDG0. columbia.EDU.
Finally, John Filson-asked-that anyone who wants to be on -the mailing listL for the NSSCC-should send name and. address to him.
.e November 24,1992 Meeting Notice Ad Iloc Meeting for Regional & Local Seismic Networks AGU 1992 Fall Meeting, San Francisco Tuesday, December S,1992 Civic Auditorium, Room 413-5:45 - 6:45 p.m.
Purpose his improm;'tu meeting is being organized (1) to review the status of USGS efforts to -
l create a National Geismic System by coordinating regional and local seismic networks with the U.S. National Seismograph Network and (2) to explore some immediate steps that can be taken toward routine and rapid sharing of earthquake data among network seismologists. Operators of regional and local seismic networks -- and all interested persons -- are invited.
l t
Primary Agenda items l
l Discussion of a charter for the National Seismic System's coordinating council, t
-- John Filson Seismic-network data -- What kinds of data are available from various networks and how can the data be accessed? (Exploring interest in creating a " data access yellow pages and :
handbook.")
- Bill Ellamrrh Rapid sharing and pooling of earthquake data - especially for events not inside a discrete l
network. (Exploring the possibility of automatically mailing arrival time and amplitude information to interested parties for rapid post-earthquake assessments.)
- Al Lindh l
For more information, contact;
-Walter Arabasz (University of Utah)
(801) 581-6274:
L Bill Ellsworth (USGS, Menlo Park)
(415) 329-4784 ~
j; John' Filson (USGS, Reston)
(703) 648-6785:
Al Lindh (USGS, Menlo Park)
(415) 329-4778
F...
Meetino on the Seismic Hazard Pro.iect at LLNL On the afternoon of-December 9, 1992, I met with. Don
- nreuter and Jean Savy at LLNL.. The first point' of discussion was progress -. ' assembling the. Senior' Seismic Hazard Analysis Committee (SSHAC).
I was' informed that-Allin-Cornell has agreed to work for what the DOE _ auditors are willing to-pay him. With-this problem out of the way, Allin should be _under contract approximately byf
)
the end of January.
[
While this is a step forward, a new problem has surfaced in that DOE has issued a new order, prohibiting contracting with-other government agencies.
This makes it uncertain as to how Dave Boore can be brought on board.
This j
question is to be resolved by senior contracts experts at LLNL.
It was mentioned that Mike Bohn at Sandia is about 1 month behind LLNL in
'l establishing his consultant contracts.
Among the obstacles he faces is a-
~
large consulting fee for Peter Morris.
LLNL has Bob Budnitz under contract now and expects Geomatrix (Kevin Coppersmith) to be under contract by about the end of January.
Overall, it 1
appears that an SSHAC meeting should be planned sometime in February.
By that i
time the NRR-sponsored' work should be finished, and Jeff Kimball's project 1
should be at a more advanced stage.
At the time of this discussion, they had not yet received a response _from Carl Stepp on the objections we had to some aspects.of-the EPRI-DOE technical agreement, in the meantime, Carl seems to have accepted our_idear, and-we should have a copy of the revised agreement-soon.
All of us (Bernreuter, Savy and I) felt that there should be a meeting of L hel t
Senior Management Group (SMG) before the SSHAC meeting.
It was concluded-that' such an -SMG meeting could be held on the afternoon of January 13,.when a, half-day is available after the meeting in Bethesda concerning the NRR Lproject.
The SMG meetingishould include not only DOE, LPRI, and NRC, but also-Don Bernreuter, Mike Bohn and Bob Budnitz as chairman of the SSHAC. The SMG1.
meeting should serve to_ iron out any remaining ambiguities.so thatLthe future -
work will not be hindered by misunderstandings.
Also discussed were the revised 189, which I should receive soon, and' monthly.
management letters which have been held up because of a personnel change-at LLNL.
S 4
-w.,
-~e
,m.
e y