ML19275A616
| ML19275A616 | |
| Person / Time | |
|---|---|
| Site: | Seabrook  |
| Issue date: | 09/26/1979 |
| From: | Sheldon K NEW ENGLAND COALITION ON NUCLEAR POLLUTION, SHELDON, HARMON & WEISS |
| To: | |
| Shared Package | |
| ML19275A615 | List: |
| References | |
| NUDOCS 7910110266 | |
| Download: ML19275A616 (21) | |
Text
'
UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE COMMISSION
)
In the Matter of )
)
PUBLIC SERVICE COMPANY OF ) Docket Nos. 50-443 NEW HAMPSHIRE, _e t_ a _l .
) 50-444 (Seabrook Station, Units 1 )
and 2) )
)
NECNP SUPPLEMENTAL MEMORANDUM IN SUPPORT OF PETITION TO REVIEW I. INTRODUCTION AND
SUMMARY
OF THE CASE On August 10, 1977 the New England Coalition on Nuclear Pollution (NECNP or the Coalition) petitioned the Nuclear Regulatory Commission (NRC or Commission) for review of the Appeal Board's decision concerning a number of issues re-lated to the licensing of the Seabrook nuclear power plant.
One of these issues was whether the earthquake design chosen for the Seabrook plant was correct.
Selecting an earthquake design for a nuclear power plant involves two tasks. The first is to determine the maximum credible earthquake that could strike the region in which the plant is located. This earthquake is called the Safe Shutdown Earthquake (SSE). The second task is to predict the level of destructive forces that would be associated with the SSE. This is the " maximum vibratory ground motion" (acceleration) .
- ?
791011 0
A majority of the Appeal Board approved the design proposed by the Applicant and the Staff. They found that a seismic design based on an SSE of Modified Mercalli Intensity VIII (MMI VIII) and an acceleration of 0.25g satisfied the Commission's safety criteria. In the Matter of Public Service Co. of New Hampshire, (Seabrook Station, Units 1 &
- 2) ALAB-442, 6 NRC 33, 54-65, (July, 1977).
The third member of the Appeal Board - Mr. Farrar -
dissented from both findings. However, he did not present his full opinion on the issue in ALAB-442, offering instead an " outline" of his conclusions with a stated intention to prepare a supplemental memorandum on the issue, Id. at 106.
Because it did not have the views of the entire Appeal Board before it at the time of NECNP's petition for review of the seismic issue, the Commission determined that it would " reserve judgment" on the matter. 7 NRC 1, 29, (January, 1978).
On August 3, 1979, Mi. Farrar completed his opinion.
The majority of the Appeal Board subsequently indicated that their views on the issues remained unchanged. ALAB-561, 10 NRC , (September, 1979).
It is now timely and appropriate for the Nuclear Regu-latory Commission to consider the seismic design question.
The Coalition submits this memorandum to supplement its
earlier filing, to support the position taken by Board member Farrar, and to urge the Commission to render its judgment on an issue which is of cr- 'ical importance to the safe operation of the Sea' sok nuclear power plant.
II. ARGUMENT A. The Appeal Board's Decision Was Erroneous The Appeal Board majority found (1) that the SSE for the Seabrook site had a maximum intensity of MMI VIII; and (2) that the NRC Staff had justifiably assigned a value of 0.25g to the maximum vibratory ground motion (acceleration) which might result from such an earthquake. As Mr. Farrar concluded, both findings are in error. The disagreement between the Board and the minority on this question centers on whether alternative approaches to those employed by the Staff and Applicant lead to the equally tenable conclusion that the SSE for Seabrook is a Modified Mercalli Intensity IX. The Board majority held that these alternative approaches are not valia and not acceptable under 10 CFR S100, App. A.
(cited hereinafter as Appendix A) . Mr. Farrar concluded that these alternative approaches are as scientifically valid as those employed by the Staff and Applicants and are acceptable under Appendix A.
Mr. Farrar concluded that the Staff's approach to establishing the maximum vibratory ground motion violates the requirements of Appendix A. The Board majority disagreed and concluded that the Staff's approach is acceptable.
PREDICTED MAXIMUM EARTHQUAKE INTENSITY
- 1. The Appeal Board Erred In Finding That Dr. Chinnery's Testimony was Invalid And Entitled To No Evidentiary Weight The Appeal Board assigned two reasons for its conclusion that Dr. Chinnery's testimony is invalid. First, the Board majority argued that Dr. Chinnery's reliance on earthquake information from utner parts of the country to support estimates of return times for earthquakes of varying inten-sities in New England is totally unwarranted without (1) an exploration o' the geologies in the three regions examined; and (2) some explanation of why any discerned differences are irrelevant to the constant relationship between earthquake frequency and size that Dr. Chinnery posits. Second, the Appeal Bord argued that Dr. Chinnery's straight-line extra-polation from MMI VIII to MMI IX is invalid because it re-quires a conclusion that there is no upper limit to earth-quakes in New England.
Neither reason assigned by the Appeal Board is sufficient to conclude that Dr. Chinnery's testimony is invalid and therefore entitled to no evidentiary weight.
First, as Mr. Farrar pointed out, Dr. Chinnery did not seek to demonstrate that the two areas he examined - the southeastern United States and central Mississippi - were similar to southern New England. The validity of his approach does not depend upon the existence of such simil-arities. Dr. Chinnery's point is that existing earthquake data for geographical areas away from tectonic plate bound-aries suggest that the relationship among the frequency of
i varying sizes of earthquakes is constant. Because of this constant relationship, Dr. Chinnery is able to use the historical record of earthquakes near Seabrook to derive a rough prediction of the occurrence of an earthquake of MMI IX intensity in that region.
As Mr. Farrar correctly perceived, Dr. Chinnery's failure to provide a " geological explanation" for the re-lationship between earthquake frequency and size is not fatal. After all, experts have yet to provide a coherent geological explanation for intraplat^ earthquakes, like those in southern New England. Absent such an explanation, Dr. Chinnery's methods provide a valid approach to earth-quake risk assessment. Furthermore, neither the Staff nor the Applicant introduced testimony to demonstrate that the geology in the three regions discussed was relevant to Dr.
Chinnery's conclusions.
Second, the Board majority incorrectly reasoned that the straight-line extrapolation can be ruled out because it is precised on the assumption that there is no upper limit to earthquakes in New England. There is no evidence in the record that the ascumption is invalid, nor is there evidence establishing an upper limit to earthquakes in New England.
Accordingly, the Appeal Board's reason for ruling out Dr. Chinnery's method is without foundation in the record.-1/
1/ A study recently completed by Dr. Chinnery under NRC contract examines this question more fully. His more recent work bolsters.his earlier position that there is no basis for placing an upper limit on earthquakes in New England and ruling out the straight-line extrap-olation. Infra, at 10.
- 2. The Appeal Board Erred In Failing To Remand The Proceedings For Further Hearings Regarding The Validity Of Dr. Chinnery's Analysis Even if Dr. Chinnery's testimony was not sufficient to carry the day, as Mr. Farrar states, it was unquestionably of sufficient weight to require the Licensing Board to in-sist that the matter be further investigated, perhaps by an independent staff analysis. At a minimum, the Appeal Board should have remanded the proceedings for further hearings on the issue.
- 3. The Appeal Board Erred In Finding That Dr. Chinnery's Approach Is Not Permitted By Appendix A The Appeal Board majority concluded that Appendix A rules out the kind of probabilistic analysis prepared by Dr. Chinnery. The majority's view was that Dr. Chinnery's approach, which compared earthquake records in three regions, does not meet the requirements of Appendix A because it lacked an exploration of the geology of the regions to ascer-tain similarities and differences, or some explanation about why any discerned differences might be totally irrelevant.
Mr. Farrar reached the conclusion that Appendix A per-mits the use of Dr. Chinnery's analysis. He found that the majority's interpretation of Appendix A excluded scientific approaches which aid the effort to establish the earthquake risk at nuclear power plant sites. Mr. Farrar correctly
pointed out that the majority's reading of the Appendix cannot be squared with the language of the original regula-tions and its accompanying statement of consideration.
Furthermore, any doubt about the correct interpretation of the Appendix has been resolved by the January 5, 1977 amend-ment. This amendment makes two principles clear. First, as Mr. Farrar stated in his dissent, "it reemphasizes that, owing to the expert's inability to supply definitive judgments in this field, regulatory decisions have to be even more conservative than usual. Second, it teaches that where selection of a governing intensity standard is concerned, the presence of one approach in the regulations is not meant to exclude other types of analyses that might aid our predictive efforts."
ALAB-561, 10 NRC , Slip Op. at 35-6.
From this, Mr. Farrar correctly concluded that Dr. Chinnery's theory cannot be excluded as inconsistent with the regula-tions.
- 4. The Appeal Board Erred In Assigning No Weight To Evidence That The Montreal Earthquake (MMI IX) Governs Selection Of The Safe Shutdown Earthquake For The Seabrook Site NECNP introduced evidence in the Seabrook proceedings to establish that the Boston-Ottawa seismic belt was the functional equivalent of a tectonic province. This approach dictates that the Montreal earthquake (MMI IX) be chosen as the Safe Shutdown Earthquake for the Seabrook site.
The Appeal Board majority discounted this approach without addressing the evidence, much of it in Supplement I
to the SER, which supports it. Rather, the Board majority was content to note that Montreal and Seabrook are separated by " seismically inactive structures" and to point out geo-logical differences between the area surrounding Montreal and the area surrounding Seabrook.
The Board's response is not adequate. All of the work cited by the Staff in its Supplement to the SER suggeets that the Boston to Ottawa seismic belt is an area of uniform seismic risk. None of it requires the conclusion that the risk at Montreal is different from that at Seabrook.
Second, as Mr. Farrar stated, although the Appeal Board majority pointed to some differences between the Montreal and Seabrook areas, they alluded to nothing in the record which makes those differences significant.
"For example, the record does not suggest that a difference in the time and placement of sim-ilar structures by similar forces is likely to result in substantially different present-day tectonism. And the remaining significant features are quite similar. The rock type, the manner and timing of their creation and emplacement, and the general level of current seismic activity are relatively the same in both areas." ALAB-561, 10 NRC , Slip. OE. at 44.
- 5. The Appeal Board Erred In Finding That Appendix A Excludes Considera-tion Of The Montreal Earthquake In Selecting The SSE For The Seabrook Site The Appeal Board concluded that Montreal and Seabrook are in two distinguishable tectonic provinces. That being the case, they reasoned, Appendix A does not permit one to
9-transfer the Montreal earthquake (MMI IX) to the Seabrook site. Even assuming that Montreal and Seabrook are found to lie in different tectonic provinces, the Appeal Board misread Appendix A to require exclusion of the Montreal earthquake.
NECNP is in agreement with Mr. Farrar that if substantial similarity exists between the Seabrook and Montreal areas, then Appendix A does not automatically preclude locating the Montreal earthquake at Seabrook for purposes of establish-ing the SSE for the site.
PREDICTING MAXIMUM ACCELERATION
- 6. The Appeal Board Erred In Finding That The Staff Approach To Selecting The Maxinun Acceleration Complied Appendix A _
The Appeal Board found that the NRC Staff had properly assigned a value of 0.25g to the maximum acceleration which would result from an earthquake of MMI VIII. The Board con-cluded that the Staff methodology was technically sound and consistent with the requirements of Appendix A.
Mr. Farrar disagreed. He noted correctly that the Commission's regulations flatly require 3 nuclear power plant to be designed to take account of the maximum vibratory accelerations that might result from the SSE. Instead of looking for this maximum value the Staff selected a mean value arguing that the mean value, when coupled with a number of other procedures, provided a basis for designing a safe plant. However persuasive this logic may be, however,
10 -
it does not alter the fact that the Staff has substituted its own methods for the clear requirements of the regula-tions. The Appeal Board's failure to call the Staff on this point is yet another error in the decision.
B. New Information Supports Review Of The Appeal Board's Decision Subsequent to his appearance as a witness for the Coalition in the Seabrook proceedings, Dr. Chinnery was re-tained by the Nuclear Regulatory Commission to investigate the seismological input to the safety of nuclear plants in New England. The major emphasis of this study was to evaluate the possibility of estimating the maxinum intensity earthquake that mignt be expected within a given region. Dr. Chinnery concluded that there is no empirical or physical basis for assigning an upper limit to the maximum possible earth-quake in New England. This conclusion supports the straight-line extrapolation which Dr. Chinnery used in his testimony to estimate a return time for a MMI IX earthquake at the Seabrook site.
A second report, also funded by the NRC, entitled "A Comparison of the Seismicity of Three Regions of the Eastern U.S.", published oy Dr. Chinnery, June, 1979, answers the criticism of the Appeal Board majority. In this study Dr. Chinnery again demonstrates that comparison of earthquake data from the Southeastern U.S., Central Mississippi Valley and Southern New England show a constant relationship
between frequency and intensity of earthquakes. This re-lationship permits him to estimate the probabilities for the occurrence of large earthquakes in southern New England.
The NRC Staff has apparently had this information for some time.-3/However, Counsel for NECNP became aware of it in mid-September, 1979, after forwarding to Dr. Chinnery the dissent and supplemental opinion of the Appeal Board. Dr.
Chinnery provided copies of the studies to Counsel for NECNP. There are attached to this memorandum.
C. The Commission Should Review ALAB-442 and ALAB-561 Commission review of the Appeal Board decision is appropriate for several reasons. First, as a matter of policy, the Commission should review the decision because of the significant split in the Appeal Board. The disagreement between Mr. Rosenthal and Dr. Buck, on the one hand, and Mr.
Farrar, on the other, is not trivial. The majority and minority opinions reflect a strong disagreement about the evidence in the record and the meaning of the Commission's regulations on a matter critical to the safety of the Seabrook plant.
Second, assuring that a nuclear power plant can with-stand earthquakes is critical to the safety findings which this Commission is required to make. The earthquake issue is particularly important for Seabrook because, as the Staff notes, the plant is located in a zone of usually
-3/ It is regrettable that the Staff should fail to notify the Appeal Board of this information so that it could be taken into account in the recent opinions.
b c
high earthquake activity as compared to the rest of New England. -4/
Another reason for reviewing the Appeal Board's de-cision is that it raises important legal questions regarding the meaning of 10 CFR 100, App. A, the Commission's regula-tions on seismic design. The Appeal Board majority read the regulations to exclude the analysis offered by NECNP's expert witness, Dr. Michael Chinnery, and to exclude con-sideration of a large earthquake in :'bntreal, Canada (MMI IX) in selecting the SSE for the Seabrook site. In addition, the Appeal Board majority concluded that the regulations sanction the Staff approach to selecting 0.25g as the maximum acceleration.
Mr. Farrar disagreed with the majority on hpth points.
He read Appendix A as permitting both Dr. Chinnery's prob-abilistic analysis as well as the consideration of the Montreal earthquake in selecting the SSE for Seabrook. In addition, Mr. Farrar concluded that the Staff approach for selecting 0.25g maximum acceleration violated the Commission's regulations. Resolution of the dispute over the meaning of Appendix A is important to the seismic issue in this proceeding and in others as well. The Commission 4/ Supplement 1 to the Safety Evaluation Report (Section 2.5.3.1, pp. 2 */ through 2-9). Seabrook is located in a zone of high earthquake activity known as the Boston-Ottawa seismic belt. This zone is ranked as one of the three most seismically active regions in the eastern United States. The other two are New Madrid, Missouri and Charleston, South Carolina.
9 is considering revamping Appendix A.-5/ However, until new regulations are adopted, the ones currently in force govern the seismic design of nuclear plants. It is important that the Commission review the Appeal Board's decision and resolve the disputes over the meaning of Appendix A.
III. CONCLUSION For the foregoing reasons, NECNP urges the Commission to review the Appeal Board's decision on the issue of the appropriate seismic design for the Seabrook plant and to find the Board majority's conclusions in error.
Respectfully submitted,
'A '
W. . _ l KARIN P. SHELDON, ESQ.
Sheldon, Harmon, Roisman & Weiss 1725 Eye Street, N.W.
Suite 506 Washington, D.C. 20006 (202)833-9070 COUNSEL FOR NEW ENGLAND COALITION ON NUCLEAR POLLUTION Of Counsel:
David S. Fleischaker, Esq.
1735 Eye Street, N.W.
Suite 709 Washington, D.C. 20006 (202)638-6070 5_/ SECY-79-300.
(. ( ,
UN LTED STATI S Ol' IsMERICA NUCLEAR RI:GULATOiO' CO:.MISS f 0:4
- . . . -)
In t he Ma t.te r of )
)
- . > U; : , f C SIMVIC G CO: d' A'J' OF ) Doc::e t I or. . 50-443
. ia H . .'S ".C: ., .e t. _a l . .
) 50 ',
) (
( S c ab r o o'- Stabioa, Un it s 1 )
and 2) )
)
_. . . _ . . . . . . . - - - . _ - . __.. . _ _ _ _ .)
- Ci'P c'I:' iCIJ.':. OF S F n.VIC E I hereby certify that a copy of NECNP Supplemental Memorandum in Support of Petition for Review and attachments were mailed, postage prepaid this 26th day of September, 1979, to the following:
?DDR ORG M.
- Vic tor Gilins::y , Co:n:eissionar U.S. Nuclear Regulat.ory Co.u";ission Washim.j ton, D.C. 20055
- D i ch a'rd T . Kennedy, Com :'. i.r,sione r U.S. Muclear Reg ula to cy Cor:.'212 0 i.cn Ua c h.i ng to n , D.C. 20555
- Pcter Dradford, Cottn i.c rei one r U.S. Nuclear Regulatory Commission Washington, D.C. ,
20555
- Stephen P. Eilperin, Esq.
St.cphen S. Outrach, Ecq.
Office of the General Counsel U.S. Nuclear Regulatory Comminnion Weshington, D.C. 20555 Alan S. Ro.s e n th a l , Chairman A to:nic S0.f e ty and I.icenning Appeal Boa rd U.S. Nuclear Regula tory Co:u:aicc ion Washington, D.C. 20SSS
(.
/
g Dr. John !!. Ducli At.caic Safety and Liccusing Appeal Board U.S. Nuclear negulatory Co:amicuion 0;ashington, D.C. 20555 f.:i.chae l C . Furrar, T.ng.
At.o.'.ic Saf~:ty and Licensing Appeal Board .
U.S. 1:ncl ear Rc gulu tory Co.r t. nice F:a sh ing ' o n , D.C. 20555 /
Ivan Smith Ato.1i.c SD f e ty and Licen nin;; ' Doa-d Pa.n el gg U.S. I ucl .r R +n:la tory Comriission 20555 h l3I 'g kla >hing ton , D.C.
d'{ Uk Dr. Er.est O. Salo Profe.3Cor of Fisheries. E> search Instituto Colleg ' of Fir.h::rior.
Unive r - ity c " Watahing ton Seattle, Un ui;im_l P.o:1 98195 D 2* . Mdr 'in I1. II;t n "t Atomi.c Sa fety and Licensing Boa t-d Panel U.S. : ucl:_a r Regula tory Cc:: micsion Washi.., un, D.C. 20555 Larry Brenner, Esquire Rob 2rh A. B a cla .c , Esq.
Marcia Ilulkey, Esq. O'Neill Backu.s Spic.1n.an Li t tle Office of the Canaral Counsel 116 Lottell Street U.S. I!ucleat Regulatory Commi.r.aion Manchester, 1:cw Hampshire 03101 Wa shing ton , D.C. 20555 Lau in I:urt, Eng.
Acn '. 3 Lan t A t torne.y General One A.Mur ton Place tr.> c t o n , Macuachusetts 02103 Thor .. C. Dignan, Jr., Ecq.
Ropes & Gray 225 Franklin S tre :t Assistunl. Attorney General Boston, lias cachuc e P ts 02110 L'nvironmental Protection Division Office of the Attorney General Sta te IIous Anne::, Room 20S Concord, ?;ew If ampchire 03301
- Docke ting and Service Section U.S. !!uclear Re.3ulatory Commission Washington, D.C. 20555 Y . n l f 4 %w Karin P. Sheldon Hand Delivered
ass e=amb mb see ' , m., - 4 .,...a Bui!etan of the Sewnoicg'ca! Swiety of Amenca. Vol. 69. No 3. pp 757.*?2. June 19~9 A COh!PARISON OF THE SEIS$11 CITY OF THREE REGIONS OF THE EASTERN U.S.*
i By Nf!CHAEL A. CnissEnv ,
ABSTRACT l Frequency-intensity data from the Southeastern U.S., C-ntral Mississippi Valley, and Southern New England are compared. They 7:; all uite parallel to one another and consistent with a stape of about 0.57. Tnere is no evidence for the existence of upper bounds to rr.aximum epicentral intensity ... se data '
sets. Linear extrapolation of the frequency-intensity data to intensities of X feads
' to expected probabilities for the occurrence of large carthquakes. The largest
' events which have occurred in these three regions are consistent with these ,
probabilities. ,
I INTnont cTros Recently there have been rather detailed analyses of the seismicity of three l
- sections of the Central and Eastern U.S. Bollinger (1973) has described an extensive ,
j set of data for the Southeastern U.S., which includes the seismically active zones of ;
, Niaryland, Virginia, West Virginia, North and South Carolina, Georgia, Alabama,
' and Tennessee, for the period 1754 to 1970. Nuttli (1974) has listed the known events !
I in the central hiississippi Valley seismic region for the penod 1633 to 1972. And Chinnery and Rodgers (1973) have analyzed the data of Smith i1962,1966) for the Southern New England region for the period 1534 to 1959. The purpose of this paper is to compare these three studies, and to bring out the similanties between them.
The discussion of seismic risk inevitably involves plotting frequency-intensity (i.e., .
marimum epicentral intensity) diagrams. In what follows we use this t.ge of plot,
, since magnitude data are not available for all three regions. This raises a difficult !
j point, since within each of these regions, the seismic activity is not uniform. The j g selection of the boundaries of the area to be studied is much skin to the problem of ,
j the definition of a tectonic province (which is required, for example, by the Nuclear j Regulatory Commission Rules and Regulations, Part 100, Appendix A).
- For the moment, we shall make the followmg assumptions
- First, we assurne that ,
~
- all subregions withm a given region have a linear frequency intensity relation of the
, form i
l log N, = a. - bl ,
where N, is the cumulative number of events in ith subregion with intensities greater than or equal to I, and a, is a parameter desenbing the level of seismic activity of the ith subregion. We assume that the slope b is common to all subregions.
- Second, we assume that the manmum possible intensity in each subregion, if one exists which is lower than the nominal maximt 1 of XII, is larger than the largest event recorded within that subregion during the period of the earthquake record.
These assumptions sound verv irastic, yet they are really unplicit whenever we plot a frequency magnitude or frequency-intensity curve. Fur hermore, at least in
- The ' news and conclumons contu.ed :n :his doeurnent ue those ai the contr2acr and e.ould not be
- terpreted as necessanly represen:tr.g 9.e detal pohetes, either exprewed ;r mplied. of *he Uruted States Gosernment.
757
-.J h -
758 mCHAEt. A. CIUNNERY principle, they are testable. It is easy to plot frequency. intensity diagrams for '
portions of a region and examine both the lineanty of the results and the constancy
- of the slope b. In practice, of course, scatter in the data often makes such a test i inconclusive. Ilowever, a substantial breakdown of any of the above assumptions .
should be apparem in the data for the region as a whole, either by the appearance of nonlinearity in the frequency-intensity statistics, or by variations in estimates of b using different data sets. As we examine and compare the setsmicity of the three areas under consideration, we shall look for infonnation related to these assump-tions.
Perhaps the most important question which we shall address is as follows: Each of these areas has had one moderately large earthquake in its recorded history (the 1755 Cape Anne, 1811-1812 New Madrid, and 1886 Charleston events). Are these ,
large events consistent with the record of smaller earthquakes that have occurred more recently? Clearly, this question has a direct bearmg on the very fundamental problem of how to extrapolate from a short record of seismicity to the occurrence of low probability events, which is particularly important in the assessment of the potential seismic hazard to critical structures such as nuclear power plants.
We shall disregard questions of the lack of stationarity of the earthquake process in these three areas, in spite of their potential importance (Shakal and Toksoz, 1977). It is very difScult to document this nonstationanty within time periods of 100 to 150 years, because of the small number of events concerned. {
i Tne DAra !
Southeastern L*.S. Bollinger (1973) describes the seismicity of four seismic zones I in the Southeastern U.S. for the period 1754 to 1970 isee Figure lh In this study we shall restrict ourselves to the two southernmost zones, the Southern Appalachian seismic zone and the South Carolina. Georgia seismic zone. The combined area of these two zones is given by Bollinger to be 307,000 km 2 Since we would like to exclude the 1886 Charleston earthquake from consideration, we have analyzed events during the period 1900 to 1969. Even this period is probably too long for the adequate recording of intensity III events, so these have been accumulated for the period 1900 to 1969 only. Total events listed by Bollinger (1973) are shown in Table 1.
These data are easily converted into a cumulative frequency intensity plot, and this is shown in Figure 2. The usual interpretation of such a diagram is to St the data pointa with a straight line, recognizing that the data at the lower intensities is i
ilkely to be incomplete. Such a fit is shown as the solid line in Figure 2. This line t
, corresponds to the equation log N, - 2.31 - 0.46I. (1)
The slope of this line is low compared to other simdar regions, as we shall see below.
The occurrence of three intensity VIII eventa during this 70 year penod seems high, and in fact one of them has been shown to be an explosion (G. A. Bollinger. personal communication). Certainly a La such as the dashed line in Figure 2, which has the equation log N, - 2.38 - 0.55I 12) cannot be ruled out. The slope of 0.55 in this equation ts very close to the slope 0.56
_ __ q
_..~
......._ . , - - - . . . ~ . . . - - . . . . . . , .. -
y
~
i a
.. a a ,, ' ' '
j
.. .. ., .. .o r. ;. .
..r. .. ' 7. J .
. . .i .o
~" @.;N..;:..."~ '7 i #2-SflSMIC ! ZONE
- i .
- L. ~ &. . . <' ,,
= 3,
' .e,,,}.. :. ; : .,, m i
..f: i: .
N w vg-- - f
- r. .
=
- o. .
- o. .
3.
e
, _SO. dPPALCHIAN " S >'
SEISMIC ZONE! ~*" " :$' '4 o' {. l ?
j'g i a @i s o,
^
5
n 3
t 0
, k ' Pg N .C *
}
f s
~^- -
-- p $, __,- ,o g h M ' '
_ 3S Q
, j 4 p 7. -
W 64..'g"@ W i LEGEND , p
.. 9? :: r.ii nepo i. o =
g' g p. . : -: e. .,~y .] o q ;~ : e *r:- ,
o @.j. n _in
,y , y o
>a g
B o 9 ;gy>:.l !!S.C :j: A vivn O $
- %ss , % ; ~54i ;.. x.M*, , g vin-ix @ %
eGaJ j; S, ga.o,
, . ~ ~ .:: -
s, -cm.
d i . 1. - . . x 1 . ;;;
i
.:'. ..:J..( o !
Two or more
>2 M
g SO. CAROLIN A - .o iN *a t
/ -- GEORGIA ,o . i.
M m
. , _ ~
, SEISMIC ZONE s_gm , '
x
-s,: - . 3. ...i.. c w -
~ 1,.
. >o a a ., ,,
[a i
.. .o
,, ,. ,, r. " ,. ,,
c,.
I l'au l. Set. urn ity in the Southca.,tcan U.S for the pern el 1770 to 19fo l(cprmlurcal, with pein.6, ion, froni lluthnger (1973).
- I
. - . _ _ _ - 1 1
t 760 MIClfAEL A. C1f!NNERY
- 0.08 found by Bollinger (1973) for the whole Southeastern U.S. For the moment, we will retain both equations (1) and (2) as possible interpretations of the data.
Central Mimssippi Valley. Nuttli (1974) has given a list of events in the central Mississippi Valley for the period 1833 to 1972. The epicenters of these events are shown in Figure 3. The total area of this zone is given by Nuttli to be 250,000 km .2 Since he lists few events before 1540, we have restricted ourselves to the penod 1840 to 1969. Table 2 lists the events dunng this period as a function of intensit.y. As j
t TABLE 1
' EVENTS IN Sof;THERN APeALACHIAN AND SOUTH CAROLINA GEORGIA SEl$MIC ZOSES tm e, e.,wa N. .t v..co
} I!!
- IV V
1930-1969 1900-1969 10 49 } j=_
190G.1969 46 VI 1900-1969 l' VII 1900-1969 5 VIII 1900-1969 3-
{ oS g
! . \ UTH CAROLP4A-GECRGlA
}
- AND '
' \ SOUTH APPALACH'AN .
j 0 ~
\ SEISM!C ZONES 1900 -1969 i
i
\
' e
- 3. -oS e
z" -i o -
) 2 \
! \
i s
I Log Ne = 2. M - OM I
-2o- Log Ng
- 2 88 - O 55 I j ! t I ' I i I *
?
2 2 = 2::t I
- NTENSITY i
Fic. 2. Cumulative interpretation.s are showTL'requency inter.sity plot for the w as Table 1. Two possible unight line i
' before, smaller events are only counted for the more recent portion of this time period. Since many events are listed with intensities intermediate between two values tsuch as 111 to IV), where this occurs one half event has been accumulated into each value. This accounts for the fractional events listed in Table 2.
Figure 4 shows a cumulative frequency intensity p ot for the data in Table 2. A
~
reasonable linearity is obtained, corresponding to the equation log .V, = 2.77 - 0.557. @
.w- - . - . . . - . . . -.
l j sEIS5t! CITY C05tPARISON-Tif REE REGIONS OF Tif E EASTERN l'.S. 761 l
. Southern .Vew England. The seismicity of Southern New England has been
{ discus, sed by Chinnery and Rodgers (1973), using data of Smith (1962,1966) for the
, period 1534 to 1959. The region defined as Southern New England is shown by the i solid line in Figure 5, which also shows the epicenters in Smith's listing. Following 4
Chinnery and Rodgers (1973), we note that many of the listed epicenters are
- 4. '
}~ s
. 3 0. =,i3 9 "*
, 4 o , ,,,4 g .,,
i j, x So em,.5 9 es.
X 60em 69 ST* i
., l * '
'~ '
i i.
.l ie sr*-......-3.'...
W<
...I.......;....' .....- '- --. -*
' i
/s '*t .
( .
y /' 2. .'U . x..: . d j ', s h
.< .' 'a$. a' s
, .. so, , .\ X *
, /
3 r . . . . .. ) . '. ........,t.. . . -
1 n > .:' * ,. . ,N . ; ' q 1- ,
j- 1.,e *
' +
W d'
3,,----...,2 l 2A ,,,.3
%eq 4Y '= /,,
..........J..r '
.I. '
1
- ' =, =y 3
s ,}I a.. ,~s e
/
'i
?
,,..P h a $ ,
"ioes*C 37
/
I
/
/ r. > ;
2/
-...3.:. ' . . . . 6
/
34 . . ,/ l
.x ..
b'
- l , , 'd ' F, l e
i s . / / s'. .a =
! ' x 5 .; .; '
i a.^4.*
l ARK. '
l yta * ; TENN i
u . 'm'hl -y l '. ,l , . C**, .
j j, s e$ - - - . . '. . . . _ A ' / uo . ..
9
+
I
[,
l
/ ~li w33, ;
/
7, . TENN l'
- i. '.
p f.. \/..*
a 7 '
- j 2 gs*
I Frc. 3. Epicer.ters in the central >!isheippi Val'ey region, for he period 1%3 to 1972. Reprocueed.
with permassion, frocs Nuttii (1974).
TABLE 2 "
Ev::.sts tw Czsnr:. N!!ssisstrer Vrttty
, :m wry P ms .No. .s E. ..
1 II 1930-1969 22.5 III 1900-1969 94.5 IV 1670-1969 143.5 V 1870 1969 63.0 VI 1840-1969 31.5 l
VII 14 0-1969 10.5 I
Vl!! IMO-1M9 1.0 IX 1340 1M9 1.0 1
4 chistered in a region extending from Boston through central New Harr.pshire. We j have outlined this area in Figure 5, and refer to it as the Boston New Hampshire i seismic zone. The areas of the two zones in Figure 5 are approximately '.00,000 km 2 ISouthem New England) and 27,000 k:n (Boston-New Hampshire zone). 5ince we wish to exclude the 1755 Cape Anne earthquake from the data set, events have been t
t I
w= :m.pw - - - _ _ _ -- - - - - - - . ~ . - . - - - - - - - - . .-
f ,
I 762 M!cHAEL A. CHINNERY j, i
accumulated in both the Southern New England regien and the Boston New Hampshire zone for the period 1600 to 1959. These are listed in Tables 3 and 4, respectively. As before, small events are only accumulated for the most recent portion of the record.
The cumulative frequency intensity plot for Southern New England is shown in Figure 6. The straight line through the data has the form Log N, = 2.36 - 0.59I. (4)
In spite of the rather low numbt rs of events, this line is a reasonable fit to the data.
In the case of the Boston New Hampshire zone, however, the number of events j 10 ,
MIS $1SSiPPI VALLEY 1840-1969 o5 -
- o -
i ,
g i
- W cr - o. S -
N.
E' o
I
, a 7 ;
/
/
'" ~ Log N = 2.
c - 0.55 I
-20 -
3 2 2 2::t I NTENSITY Fic. 4. Cumulative frequency-intecn:y pict for :he data in Table 2.
becomes low enough that it becomes difficult to formulate a linear fit with any certainty. A straight line through the upper four data points has a shallow slope (about 0.50), which is significantly different from the other areas studied, and which leads to high estimates of risk for large events. We prefer to interpret these data with a line such as the one shown, which has the equation log N, = 2.15 - 0.591. 15)
With this interpretation, the number of intensity VII earthquakes is anomalously high, due either to poor data or a statistical ductuation. At least equation (5) should lead to reasonably conservative estimates for risx at high mtensuy levels.
- n. ~ -
SE!S5tICITY CO5tPARISON-THREE REGIONS OE THE EASTERN U.S. 763
) o
\ \
, Dr / /
a o
\p . o\
9
]'
p' \, \
a /
oi \ r
.Esb od q = \ \
\ of ,
. \ \
0
')
V \
,\
\g * .wmh.
o o
\ O @\a f)
\o .\
0 s'
\0 O Q
% 1
\g?g,\ ?'1 ,
, ~1 c s s.r.
, q[~,1m n
0 00 d o
~,\ "
o . M.39,4) '-R
- n L=~
p c} d. M 4
\ \ ,58
- i' _i
- t L O 1
) I .4 I:
' f \
,s f. o$ lk
- ^ gy* % g
\ a ' {= ,
p 3%,'3 3oi j-
/ =3,
)/ '
o, 0 0 ' #
' U ci._
' ?. [ \
E,i d N e c
\
= 4!- 5 j--
/ .3:eCL. 8
'a di * =
0 '\ , 4 CO O -\
E-
- 0 \C a
_h' O\ \ :z--3
=
g k,'= P , )\ *
= -l- >
(c0 :r -
h \ \
\ 1I [ 2 k'v'~~a\
4,,
\
j
. v.c c ~ r a t e.,,- . ..e, . : m
- r
/
- o EpicE N1m E :vasens ne es t : 20)
'y Q 0 3hvt C*iCENT8CS
\ '\ Q SEISMCGR A ** STA?iGN
\
>>@Q(
g g
o
-w to ico
, t \ $ calf N MdLES Fic. 3. Epicenters m Neu England. from Scuth < !%3). The whd '.ine x.thnes the .egion eiled Southern New Eng!and m :ha 5:udy. The brocen 'ine .r.dicates the Bos:un.New Hampehae xae see Channerv ind Rodgers,193.
CO5tPARISON OF FREQUENCY-INTENSITY DATA The sequency intensity data shown in Figures 2,4,6, and are showm :oge:her in Fizure L In this case we have cam:ed :he individual interpreta: ton usme 5tted straigh: lines, and show the data a. lone. His emphasizes the .erv sundar character of the four recurrence curves. There is acme scatter, but each of -he curves :s 1
J.
Ifr% S!!Cil AEL A. Cil!NNERY consistent with a s!cpe somewhere in the range 0 55 to 0.60. and we show a slope of 0.57 which seems to be a reasonable average.
In view of the rather inferior quality of much historical intensity data. It is '
surprising how consistent the slopes of cumula:ive frequency-inten.sity data appear TABLE 1 Evtsis 1.s SOLTHFM.s NEW E.NGLA.N D l m .w . Pen..a w a n.nu
(( 192% 1959 12.5 111 192% 1959 X5 IV l'Ko-I!f 59 43 0 V 1 % b1959 24 0 VI 19'O-1959 60 VII 1600-1959 30 TABLE 4 EvrNTs IN Boston.New HAMPSHIRE ZON E
- nt..w.t v P.r o w . >:..nu
!! 192%1953 16u 111 192% 1959 IJ S IV 1900-1959 17 5 V 1660-1959 110 VI 1N 4-1959 15 VII 16tAk1F29 30
'O, I
SOUTHERN NC4' IOND l l '800 '959 CSh .
l . ;
- ; i Lg Pt. = 2 % - 3 ' 9
4 . ;
' ! i j 2
9 I i a o-i i
I !
i
-I SH i
! l
- i
-z b l I ,
2 3 2 =1 I hTE%TY Frc i Cumulauve hequency.mten.sny pict Mr :he data m Table J.
to be. Both Connell and Met: 11975) and Veneziano (1975) have surveyed a number of estimates of this slope, and many of these are censutent with the present data.
The mean of the : estimates quoted by Veneziano :.s 0.53, but his list contains some low values which are probably not realistie. Of particular interest are :h+
I
SEISMICITY COMPARISON-THREE REGIONS OF Tl!E : ASTERN c.s. 765 values 0.59 for the whole U.S. (Connell ar.d Merz,1975) and 0.54 for California
( Algermissen.1969). A recent estimate for the area around the Ramapo fault in New York and New Jersey is 0.55 0.02 i Aggarwal and Sykes,1978).
It is interesting to compare a slope of 0.57 with the value that one would predict from known magnitude intensity relationships. A selection of these relationships have been given by Veneziano (1975), in the form Jf = ai + a:I. (6)
Values of the constant a2 have been estimated as 0.67 (Gutenberg and Richter, 1956), 0.69 ' Algermissen,1969), and 0.60 iChinnery and Rodgers,1973: Howell, os g SCSTCN -NEW HAMPSHRE I 18C0 - 1959 l
r .
! . t:g cs 2.m - om
, - 0 S '-- ,/ l
<t I
Y 5
1 i
- i o '---
z"o i a
? ,
- 15 r i
I i !
- 2 o -- ,
i !
t t l
l
! ! i i
21 I lfCENSiTY Fic. 7. Cumulatise 'requer.cy. intensity plot for the data m Table 4.
1973). The latter estimates of 0.60 were obtained from data in the Eastern U.S., and
~l may be the best estimates for our present purposes.
There is an abdunance of frequency. magnitude data, which is usually represen:ed by the form log N,= a - b3f 47) where the s! ope b often ues between 0.9 and 1.0 isee, for example, Chmnerv and North,1975 >. Combirung this expression with equation 4 6), with a = 0.60 xould lead to a slope of the frequency-interuity relation 'cetweca 0.54 and 0.60. Cleariv the
. . . . . . . . . . . . . ~ --, .-- -. .. - . . . . . . - .
- 766 M'C1tAEL A. CHINNERY o %55 53. Pot <a .E' a W NEas'EaN J $
e SCtt-EaN sE4 g e so 05- a SCSTON - NE W =avFS" 5E s- \
t \
I \
T , \
S LOPE Oe7
. \\ ,
2 \
,$ . i o 'r
\
l \
t
\
\
.,2- \
2 2 7. *1 I NTENSIT
- Frc. 3. Comparuon of the frequency. intensity data ' rom Fwres 2. 4. md ?
i o WSSISS.m saJ ) 5
,3 , a SCUTHEaSTE*N J S e SCUTw(GN NE 4 [*,CJN Ll{jk j*(
a SCSTCS - NE a -4.vPS- m 20-
"a f
a ts-f \
5 e 3:-
s 2
T
- ss-40 -
as-
- : 1 = 1 NTENS.*'
Fic 9 The same :ata used .n Fcre 3. but norm 62eo : he uvu of the vr.ous :or.es.
4 5EISStICIYY COS!PARISON-THREE REGIONS OF THE EASTERN l'.3. 767 0.57 value shown in Figure 9 is eminently reasonable and consistent with other informanon.
The similarity between the four sets of data shown in Figure 8 can be ft.rther emphasized by normalizing for the areas of the seismic regions. After this normali-zation, Figure 9, the recurrence curves are found to lie almost on top of one another (we have chosen to normalize to 1,000 km2 , but this choice is completely arbitrary).
l The apparent similarity in seismic activity per unit area is entirely fortuitous, and 1 is simply due to the particular regions chosen for each study. The true levels of 1
activity in the three regions differ markedly (see, for example, the return penods calculated in Table 5). However, one is tempted to note that the activity per unit area in the Boston New Hampshire zone is slightly larger than that in the South-eastern U.S. Is there really any good reason why an event the size of the Charleston j earthquake could not occur in the Boston New Hampshire zone?
It is interesting to search these data sets for evidence that there may be an upper bound intensity in some of these areas. Cornell and Merz (1975), for example, have proposed a frequency-intensity curve for a site in the Boston area that curves downward and becomes vertical (parallel to the ordinate axis) close to intensity VII.
Since this calculation is for a single site, it is crucially dependent on our ability to predict the location of large events near Boston. Certainly, if large events could occur anywhere within the Boston New Hampshire zone. the present data show no indications of an upper bound. Given our present knowledge concerning the mech-anisms of large events in regions like the Boston New Hampshire zone, it does not seem reasonable to propose such an upper bound.
RANDOSINEs3 OF THE CATALOGS Before attempting to calculate the risk of large events in the three areas under consideration, we should briefly address the nature of the statistical model to be used. It is usual to assume that catalogs such as these are random,i.e., described by the simple Poissonian distribution.
This problem has received ample treatment in the literature (see, for example, Lomnitz. '966). In some cases the Poisson distnbution has been shown to be a good description for large events, Epstein and Lomnitz t1966), and Gardner and Knopoff (1974) have shown that the Southern California catalog, with aftershoc'.
(S), to calculate probabilities.
In passing, Figures 10 and 11 make another point. It is easy to use the quantity l
y mean return period of earthquakes in a sequence as ifit has a deterministic meaning. -
These figures are a remmder that the mean return period is entirely a statistical w I
'l wss:ssa cEv
'9CO '972 J i
- S - 84 EVENTS & . 2 2 RETURN PERCC ]
- O 37 YEARS f
20 W r 0
i -
W $~ -
5
- y l
- f/ *, ,
"C = I t
3 i l , e -, m
- 1 2 3 4 $ 4 ATEROCO'#RENCE 9 VE . rms )
Fic t0. Interoccurrence :unes unng Nuttlis (1970 data for he central hseppi Wiley. The exponential cune would be expected for a Pmwn dumbuticn.
quantity, and that its only real meaning is as one of the parameters desenbing the probebility distnbution that corresponds to the catalog under consideration.
THE PaosABILIrv or Lance EVENTS
[
With the above model it is now possible to address the question posed in the introduction. In each of the three areas under consideration a large earthquake {
occurred shortly before the periods of data that we hase analyzed. Are these large earthquakes consistent with the later record of smaller events?
Our procedure :s simple. We take the linear relations 5tted to the frequency in ensity data, enrapolate them to larger intensities and make estimates of the j mean return periods of these larger intensities. We then use equation @ to estimate '
the probability that at least one of these larger events w111 occur in any 200-year p!
j period, and specifically relate this to the 200-year per:cd ending at the present time ia 3C'0-year period was chosen for New Eng!and, since the largest event occurred m j
the 1700's).
- i l
1 h
h
.aee i 4 m, SEISMICITY COMPARI5ON-THREE REClONS OF THE EASTERN l'.5. 769 The results are shown in tabular form in Table 5. We do not pretend that these :
i numbers are very accurate. In fact, because of the subjectivity that has to be used {
in obtaining the linear relations [ equations (1) to (5)], there is no way to make a ;
realistic assessment of errors. We therefore view the numbers in Table 5 as being a t qualitative indication of risk, rather than quantitative. The results for the individual l areas are discussed below.
t l I .
'o' SOUTHERN NEW ENGLAND k '
1860 -1959 32 EVENTS WITH I27 ,
,1 RETURN :ERIOD To = 3.13 YEARS i , ,
1 i
s D sh b z :
ne 4 3 g' 3
! k
- s i
2 ~- q
! l o
?,
5 li 'O '$ 20 INTERCCC'.,RRENCE TIME ( years )
Fic.11. Interoccurrence times for Sou:hern New England from :he data of Srmth 1960 ;%dt TABLE 5 PRUB ABILITY OF LARGE EVF.>r$ ;N FotlR REctoss or THE EA5TER.N C S L
Feraten L **d s.,..r.,e, w ....n.
1.rne Be* ire P"*"*""".n*r"'*"#'
aPn i
{ <e. em ewne r v,an e:l! ::x xx p u: ::x xx Southe.utern t
- S . ;900- 1 23 68 195 200 99 45 64 IM9 2 33 117 417 XO M -2 M Misstw.ppi Valley. 'MO- 3 43 :51 537 200 99 *] 31
'.N9 Southern New England. 4 ?29 s91 3467 Xo 73 29 +
l 1300-1959 Boston-New Hampshire, 5 371 1445 5603 30 0 35 :9 3 1900 '.359 The earthquake catalog for the Southeastern U.S. desenbed by Bollinger (1973) is approximately 200 years long. Table 5 shows that, on the basis of the most recent 70 years of this catalog twhich may !cgically be expected to be the mest complete at lower mtensitiesi, there is a substantial probability of the order of 50 per cent that at least one earthquake of intensity X 3r greater will occur in a 200-year penod. We ,
i conclude. therefore, that the Charleston ear-hquake of 1568 iintensit . X. Bctlinger !'
! 1977+ :s entirely conststent with the 19M to 1969 data.
l t
i
.m.,,. .. . .
. _ _ . . . _ . _ _ _ _ . _ _ _ . _ . ... . _ .. .. . . . . __ . j
- 770 t MICHAEL A. CHINNERY Without any question the largest earthquakes during the past 200 years in t' e central Mississippi Valley were the 1811 to 1912 New Madnd events. Nuttii (19/J) lists the maximum observed intensity during this sequence as X to XI, at New Madrid, Mi.uouri, Gupta and Nuttli (1976) have recently revised this upward to XI to XII. Some question perhaps remains as to the validity of this value as a true epicentralintensity, since some ampli6 cation by the allusium in the area might be expected. Table 5 lists the probability of an event of intensity X or greater durmg a 200-year period as being about one. third. The New Madrid events were therefore reasonably consistent with the data for 1840 to 1969. Ifit could be shown that these j
were the largest events in the last 300 years in this area (which is not unlikely), or j that the true epicentral intensity was somewhat less than X, it would be easy to increase the calculated probability to 50 per cent or more. l The record of earthquakes for Southern New England is about 300 years long l (Smith, 1962, 1966). During the penod 1800 to 1959. Smith lists 3 events with intensity Vil, and there are none any larger. Table 5 shows that there is a respectably {4 high probability sabout 75 per cent) that an earthquake ofintensity VIII will occur somewhere in Southern New England in a 3009 ear penod. The probability of such '
an event in the Boston New Hampshire zone is about 50 per cent. The epicentral intensity of the 1755 Cape Anne earthquake :s not well denned. Snuth i1962) !!sts this event as intensity IX, which is probably somewnat high. The Earthqucke History of the Umted States (NOAA publication 41 1, 1973) lists this event as intensity Vill. Other unpublished studies have deduced intensities close to VII.
Whicheser is correct, it cannot be said that this event is inconsistent with the subsequent seistmc record.
An equally important result for the Southern New England region is that the !
probability of intensity IX and X es ents occurnng withm a 300-year penod is quW g low. The absence of these events in the historical record is therefore again consistent
]
with the 1500 to 1959 data. Notice, too, that the return period for intensity VIII :s I 229 years, which is consistent with the absence of such an event during the pened 1800 to 1959.
CoNct.t szos We can make several conclusions from this study ,
- 1. The four frequency. intensity plots that we ha.e considered show a remarkable uniformity. All show a pronounced linearity, and have slopes which are consistent f with a value of about 0.57. This, in turn, corresponds to a magnitude b-s alue in the !
range 0.9 to 1.0. This uniformity, and the fact that 0.57 is very close to slopes observed in other areas of both Eastern and Westem U.S., suggests that frecuency-j f
mtensity data can usefully be applied in seismic nsk analysis. In areas wbri data are poor or sparse, it would appear possible to combine data from as little as one intensity value with the apparently universal slope of about 0.57 to construct a local frequency-intensity relationship. Such a procedure may be more reliable than some of those in current use.
- 2. The unifornuty of the shape of the frequency mtensity relation over regions ranging fr m ;he Boston-New Hampshire :one and the Ramapo fault zone ( Aggarw al and Sykes,1975) to the whole of the continental U.S. suggests that the problem or' nonunifomuty ciseism: city within a region is no impediment to the use of frequency. j intensity statistics. The assumptions outlined in the increduction to this paper seem to be useful work:ng hypotheses. (
p f
(
I
SEISN!! CITY CONIPARISON-THREE REGIONS OF Ti!E EASTERN l'.5. III
- 3. The question of the existence of upper bounds to maximum earthquake intensity tiess than the scale maximum of XII) remains unanswered. There is no reason within the data themselves to suggest that the three large events that we have considered are the largest that could occur m these regions. Simdarly. :here are no statistical arguments that a very large event could not occur in other areat tsuch as Southern New England outside of the Boston New Hampshire zonel that have not recorded such an event. A rational, conservative approach to the estimation of the seismic risk at a site would include the pos.sibility of events with intensity X l
or more anywhere in the Eastern U.S. This topic will be discussed more fully elsewhere.
- 4. The validity of linear extrapolation of the frequency intensity data has been tested by predicting the probability of occurrence of large earthquakes in the l historical record and comparing this probability with the known occurrence oflarge earthquakes in each of the three areas. The Charleston and Cape Anne earthquakes are both consistent with more recent data from small events f calculated probabilities I of these events are 50 per cent ore more). The New Madrid sequence is only shghtly anomalous. The chance that such an event would occur during the past 200 years is about 30 per cent, but the chance that it would occur in a 300 year record approaches 50 percent. Thus, it appears that linear extrapolation of frequency intensity data to intensities of IX and X :s a valid procedure in these areas.
ACKNOWLEDGMENT This retearch sas supper:ed by the Nuclear Regu!atory Commtuien. The author appreciates :he heiptui ecmmenta on 'his paper recened from 0 W. Nutth and G. A. Boihnger REFERENCES Agrarwal. Y. P and L R. Sykes (1978L Earthquakes, fault.s. tnd nue: car power planu n Southern New York and Northem New Jer ey. Science 200, 425-429.
Akt. K. 619.%L Some problems m statistical seismology. Ztstn 4,205-223 A:germnssen. 5. T t1969L Seumse Ruk Studies in the L*ntrea States. Proc. M*orld Conf Earthquake Eng 4th. Santiago BoUmger. G. A. 1973L Sets:nicay of the Southeas:ern (*mted States, Bull. Seism. Soc Am. 63, '74
- 308 Bodmger. G. A. s1977t Remterpretation of
- he intensity data for the IM6 Charleston. South Caroima, ear hquake. In S:udies Related to the Charleston. South Carolina, Earthquake :l:+66-A ? elim.
inarv Report. L'S Geol. Su?t e, P oiess. P:per 1023.17-32.
Chtnner3. M. A. and R. G. North i1975L 'he frequency of verv !arge ear,a ,akes. Sc:ence 190, .W7-1198.
Chmnery. M. A. and D. A. Rodgers $ 1973) Earhquake stat.4 tics ;n Southem New Ergland. Ea-thquake Notes 44,99-103.
Cornett. C. A. and H. A. Merz (1975L Setsrrac r.sk analysis of Bo4 ten. ,I Struct. Du. ASCE 101 no. 5T10, 2027-2943.
Epstetn. B. and C. Lomrutz *19%. A model for the xcurrence of targe einhquakes. Narare 211. 354 3.%.
Carsner. J K. and L Knopoff 41974L 13 the sequene; of ear-hquakes a Southem Caiifornia. anh aftershocks re r.os ed. P:tssoman?. Bull Scum. Soc. Am. 64,1J63-1 M7 Gupta. L N. and O. W. Nuttii 19 6L 5patial a::enuanen of m:ensules for rentrall' S rar:nquakes. B il.
Scum. Soc Am. 66, 743-751.
Guter.cerg S and C F Richter ilm Eannquake magmtade, mtensuy and accelersacn Sall. Se: 3m.
Sec Am 46.105-145 Howeil. 3 F . Jr.1973L Earthquake hazard m the Eas:em L'ruted States. Eartat M re 2i Scs. 42. 41-45. i Kropof'. L. ilo64L The stat:stics af eanhquues m Southem California, Ball Scum. Soc. Am. 54, 471- f
' d 3.
Lomm::. C it9% 5tatat. cal predicuen of eachquakes. Ret. Geophys. 4. 377 393 Nutd.O W. 1973t The Mastssippi Valley ear.hquues of im and *.312. .ntenaattes. inunc mo ;cn and matmtudes. Sull. Seum. Soc. Am. 63. .Y-:43.
?00R ORBlM_
_ . . . .. __. . . _ . _ _ . _ _ _ . . . _ _ . . . . . . . . . . . . . . . . . . . . . l
~72 MICHAEL A. CHINNERY Nutth, O. W. i19741. Magrutude recurrense relat:cn for central Muut.uigi Valley euthquakes. But!
Seum. Soc. Am t*., 119-1267 Shakai, A. F. and A! N. Toksor < !977). Earthquake huud in New England. Sc:ence 194. 171-171 ShLen. S and St. N. Tok.u;r i!979L A clustents model for eu*.hquake xcurrences. Buil Susm. Soc. ,
Am. 60,1762,-1767. I smith. W E. T. 41962t Ear.hquakes c." Eastern Canada and ad;acent areas.15'A-1927, Pabt. Dom. Obs Ottaua 26, 71-JJ1 Struth. W E. T. (1966L Earthquakes of Ea. stern Canada and ad% cent areas. 19'.% 1959 Pabl. Dom. Obs.
Ottaa a 32,97-121.
Venetnano. D (1975) Probabilutw and S:atatical Models !c:r Scumic Ruk Ana;>su, 51 i T. Dept. of ,
Civil Er.g., Pubbcaticn 5t:5 'A.
APPLIED St.IAMOLOGY Gnot:P LasccLN LAnonArcay. SI 1.T.
42 CranTow Sintet i C4xanttics Mrs4Acut sr:Ts 72142 Manuscnpt recen ed Octct.er 17 1979 i
.