Dec 10, 1980 - shortening and uplift of the Chugach-St. Elias .... line from about Yakutat Bay to Prince William. 1966 .... westerly trending pattern of historical and ... YAK, Yakataga district; IB, Icy Bay; YB, Yakutat Bay; CS, Cross Sound).
JOURNAL OFGEOPHYSICAL RESEARCH, VOL.85, NO. B12, PAGES 7132-7150, DECEMBER 10, 1980
TECTONIC MODEL ANDSEISMIC POTENTIAL OFTHEEASTERN GULF
Lamont-Doherty Geological Observatory of Columbia University Palisades,
Based on 13 new fault
solutions and published seismological• geological• and geophysical data• we interpret the deformation
central and eastern Alaska are probably also capable of producing large earthquakes• but their
along the Pacific-North American plate margin in the eastern Gulf of Alaska. Three major tectonic units can be distinguished: (1) the North American plate• (2) the Pacific plate• and (3) a
recurrence times cannot be accurately determined. They may be of the order of hundreds or thousands of years.
belt of mobile borderland terranes. The Pacific plate moves in a NNW direction at rates of about
6 cm/yr in relation to the North Americanplate.
The eastern Gulf of Alaska is tectonically one
That motion results in mostly right-lateral strike slip at the Queen Charlotte-Fairweather fault system• a well-known observation. A new finding• however• is that a small component(•1 cm/yr) of convergence may also be present which results in minor subduction of the oceanic plate beneath
of the most interesting but least understood segments of the entire Pacific-North American plate margin. It appears that plate convergence is associated there with an unusual degree of crustal shortening and mountain building. Many of the region's tectonic features resemble those of
portions of the continental margin. Heretofore the Queen Charlotte-Fairweather fault zone and associated continental margin was interpreted as a
continental collision zones rather than oceanic subduction zones. Mountain peaks along the Gulf coast reach heights up to 6 km above sea level•
classical• pure transform boundary. The Yakutat block• a borderland terrane about 400 km long and 100 to 200 km wide• is carried passively by the Pacific plate except that the block slowly over-
while offshore basins nearby have subsided several kilometers since the late Tertiary. Whyare these orogenic processes more pronounced and accelerated here than at other segmentsof the
rides this plate at about 1 cm/yr. This motion is taken up by almost pure thrust faulting in a southwesterly direction along a 400-km long SE striking shelf edge structure. At its NWedge the
same plate boundary? Does Pacific oceanic lithosphere subduer beneath the eastern Gulf of Alaska continental margin despite the absence of any well-developed dipping zone of intermediate-depth
Yakutat block is in turn being thrust beneath the North American plate along the Pamplonazone-Icy Bay lineament. The underthrusting of the Yakutat
earthquakes? Howis the transition from a transform plate boundary in SE Alaska to a consuming plate margin in the Aleutian trench system
shortening and uplift of the Chugach-St. Elias range. The effects of this collision may extend as far as 500 km inland and cause somedeformation at the Denali fault in the central Alaska Range. Subduction of the Pacific plate beneath the colliding margin appears responsible for development of an active
arc up to 300 km inland
the unusual crustal
tions in the overriding North American plate related to the transition tectonics? Do simple plate tectonic principles, i.e., rigid behavior of plates or platelets• apply to such an apparently complex region? How difficult is the assessment of seismic potential and seismic hazards in this tectonic
which trends SE from the Wrangell Mountains to
questions we try
Yukon Territory• Canada, and perhaps to Mr. Edgecumbevolcano in southeast Alaska. The tectonic model proposed implies a high seismic hazard for the Queen Charlotte, Fairweather, and Chugach-St. Elias fault systems. At these fault
study. Consequently, the plan for this paper is (1) summarize the salient features of previous models for the tectonics of this region, (2) analyze the teleseismic data to derive fault plane solutions•
zones we estimate recurrence times for great events of about 100 years, but they may vary between 50 and 200 years. A temporarily very high potential for a great earthquake has been deter,mined for the 'Yakataga seismic gap' located between Icy Bay and Kayak Island. Large or great thrust earthquakes on the detachment fault underlying the entire Yakutat wedge also appear possible but may only occur infrequently. Their recurrence
and determine the dominant patterns of deformation associated with seismicity, (3) compare the seismically determined slip vectors with those computed from global plate models, (4) combine the slip vector data with recent geologic findings concerning the existence of displaced continental borderland terranes, and synthesize these new findings into a new and realistic tectonic model for the region, and (5) discuss some of the impli-
1980 by the American Geophysical Union.
Paper number 80B1228. 0148-0227/80/080B1228501. O0
The Gulf of Alaska region (Figure 1) includes important segment of the active boundary
1Alsowith the Department of Geological Sciences,
to address in the following
'" AREA OF STUDY
62 ø PRINCE
WILLIAM COPPER SOUND RIVER
60 ø )NTAGUE
GULF OF ALASKA
6.3crh/y 5.9cm/y PACIFIC
Nap showinglocation of major faults with late Cenozoicdisplacement(solid
lines) in southeast Alaska and adjacent portions of Canada [King, 1967; Plafker et al., 1978]. Dashed lines denote faults with possible late Cenozoic motion. Shaded areas denote
zones of active
suggested by patterns
of shallow seismicity.
motion of Pacific plate with respect to North American plate at different [Chase, 1978; Minster and Jordan, 1978];
Long arrows represent
rates are given below arrows.
localities Stars denote
volcanoes of Quaternary age. The Gulf of Alaska Tertiary province is interpreted as being formed by a series of sedimentary wedgessubductedby Pacific Oceanfloor. One of these crustal
bathymetric shelf edge structure. Charlotte
edge forms a prominent
stretches from the northern tip of the Queen trench.
Pacific plates. It is a zone of transition between the transform fault segment offshore British Columbia to the southeast [Tobin and
At its northern end• near Yakutat Bay• the Fairweather fault appears to connect with the Chugach-St. Elias fault system. This system
Sykes, 1968], and a region of plate subduction with predominant thrust motion on shallowly dipping faults along the Aleutian arc system to the west [Stauder, 1968; Stauder and Bollinger, 1966; Plafker, 1965]. Between Yakutat Bay and Cross Sound the most important active fault exposed onland along the
consists of a series of seismically active, E-W striking thrust and reverse faults which bend around the gulf approximately following the coastline from about Yakutat Bay to Prince William
continental margin of Alaska is the Fairweather fault (Figure 1). This fault was the site of a
Plafker, 1967; Bruns, 1979] resulting in N-S crustal shortening and a strong uplift of the St.
M = 8.2
On the basis of geological
Sound. Complex fold and imbricate thrust systems
deform the Cretaceous and younger sedimentary strata of the Gulf of Alaska [Stonely, 1967; Range.
The Chugach-St.Elias and Fairweather fault
evidence [Plafker et al., 1978] the Fairweather fault has been shown to have moved in a rightlateral sense at an average rate for the last 1000
systems separate a series of mostly Mesozoic metasediments and crystalline rocks with a few younger intrusives to the north and northeast• from a
years of at. least 4.8 cm/yr and probably closer to 5.8 cm/yr. Thus it is believed to be taking up the major portion of the Pacific-North American plate movementin this eastern portion of the
sequenceof younger, mostly Cenozoic sediments deposited along the continental margin. Offshore, the Cenozoic sedimentary province is boundedby a NWtrending shelf edge, a prominent structural and
northern part of the QueenCharlotte fault, •ear
suggestedby the patterns of shallow seismicity
Cross Sound, to the easternmost continuation of the Aleutian trench SW of Kayak Island. This shelf edge structure delimits the SW front of the wedge-shapedYakutat block (Figure 1). Plafker et al. , yon Huene et al. , and Bruns
(Figure 6)near Kayak Island, the Pamplonaridge, and a lineament SE of Yakutat Bay. The convergence of the Pacific plate with southern Alaska has continued at least since the Miocene [Atwater, 1970]. If the continental North
 propose that this linear
feature is only
Americanplate is being underthrust by the oceanic
weakly active and essentially a remnant structure at the edge of the Yakutat block. No definitive data have been published for interpreting the
Pacific lithosphere, the question arises: where, how far inland, and how far down into the mantle has Pacific lithosphere been subducted? A volcanic arc, active during the late Tertiary and the Quarternary, extends SE from the Wrangell volcanic mountains (Figure 1) into the AlaskaYukon Territory border region [Souther, 1977] and possibly farther into southeast Alaska. The presence of this basaltic to andesitic volcanic arc - however discontinuous - suggests that oceanic Pacific lithosphere may have been subducted to depths of 100 km or more directly beneath these volcanoes. This speculation on a subducting lithosphere is debatable, however, since no coherent zone of intermediate-depth
geologic evolution of this Yakutat block. Several investigators [i.e., yon Hueneet al., 1979; Lahr et al., 1979; Plafker et al., 1978] favor a continental origin of this 'microplate'. Plafker et al.  believe that the microplate is strongly coupled to the Pacific plate and is thrusting beneath or being obducted onto the North American plate. In this paper we refine an interpretation of Stoneley  that the Yakutat block is an allochthonous wedge of late Cenozoic
underlain by late Mesozoic/early to
mid-Cenozoic sedimentary strata
Both the younger turbidires and older sediments were probably derived from the adjacent, rising North American continent. We modify earlier models by proposing that the Yakutat block is allochthonous and may have originated in a more southerly position as a fore arc terrane or accretionary prism during subduction of the Farallon plate [Atwater, 1970]. This borderland block appears only partially coupled to the Pacific plate, which is slowly underthrusting it. On the basis of focal mechanisms and slip directions (see next section) we suggest that the shelf edge of the Yakutat block is associated with a zone of thrust faults that are moderately active and shallowly dipping. We infer that this thrust zone may underlie the entire wedge-shapedYakutat block. Marine geologic evidence of this thrust faulting at the leading edge of the Yakutat block is either not strong [von Huene et al., 1979] or
North of Kayak Island,
fault systembends first
(>70 kin) earthquakeshas been shownto dip beneath the volcanic arc.
Fault Plane Solutions in the Eastern Gulf of Alaska
Data and Method of Analysis
A total of 19 fault plane solutions are presented in this study, of which 13 solutions are new. The remainder of the solutions are taken from the literature as they appear to represent reliable and well-constrained mechanisms. Commentson the quality of the solutions are given in the appendix. The new solutions are shown in Figure 7. Discussion of the techniques used are given by Stauder , Sykes , and Isacks et al. . Figure 2 and Table 1 summarizewhat we consider
to be the most carefully the Chugach-St. Elias
to a NWtrend and then to
Alaska region approximately
a SW trend on the landward side of Prince William
134ø-149•W and for the period 1927 to 1979
Seaward, a complex pattern of branching and fold
Characteristics of Solutions
between Montague and Kayak
[Bruns, 1979]. and offshore
young sediments of the Copper River delta and the entire
screened set of new and
This complex zone of
(Figure 2) consists of two distinct
easternmostlimit of the rupture zone of the great (M = 9.2) Alaska earthquake of 1964 (Figure 4).
strike slip solutions (numbers1 through 4), and 15 dip slip solutions (numbers5 through19). The
Alaska, sometimes referred to as 'Tertiary Province' [Plafker, 1967] is bound in the west, near Kayak Island, by the zone of thrusting which was active during the 1964 earthquake, in the east by the zone of mostly transcurrent motion along the Fairweather fault as outlined by the 1958 earthquake rupture zone, and in the SWby the less active shelf edge structure (Figure 1). Structures of the near-shore portions of the Tertiary Province are characterized by disharmonic folding, i.e., long-wavelength synclines and short-wavelength anticlines [Bruns, 1979],
distinct: the strike slip solutions are confined to the QueenCharlotte-Fairweather fault system in the southeast, whereas the dip slip solutions are associated with zones characterized by mappedor inferred thrust faults. The choice of rupture plane for each of the 19 solutions can be made readily. For the strike slip solutions the northwesterly trending nodal plane is inferred to be the rupture plane, since it is consistent with the northwest trend of mapped throughgoing faults, the northwesterly trending pattern of historical and
ThuWs a distinct portion of the eastern Gulf of
structures of highly imbricate foreland fold and thrust belts. The offshore portion of this province may be transversely cut at crustal depth by a series of NE striking features (Figure 1)
two subsets are
also geographically quite
and the elongation
aftershock zones of individual earthquakes. Examples of this elongation can be seen in Figure 4, for event number4, the 1958 earthquake along the Fairweather fault [Sykes, 1971], and
Focal mechanisms and major faults in southeast Alaska and in adjacent parts of
Canada. The number for each solution is indicated near its epicenter (solid dots). Shaded areas in the focal mechanisms (lower hemisphere projections) represent compressional first motion; P and T denote inferred axes of maximumand minimumprincipal stress. Triangles and stippled area indicate epicenter and rupture zone [Lahr et al., 1979] of. the MS -- 7.2 St. Elias earthquake of February 28, 1979. Pamplona fault zone, . PZ, is Indicated by area between small arrows. Yakutat block, Y, is boundedby Pamplona zone, Fairweather fault (F-F) and the off-shore fault zone of the shelf edge structure approximately between events 5 and 13 (PWS, Prince William Sound; KIs, Kayak Island;
YAK, Yakataga district;
IB, Icy Bay; YB, Yakutat Bay; CS, Cross Sound). Stars denote
Wrangell volcanoes, WV, and Mt.
Edgecumbevolcano near Sitka,
age. Faults are D-F, Denali; T-F, Totschunda; CH-STE-F, Chugach-St. Elias; CH-ST-F, ChathamStrait; QCH-F, QueenCharlotte. Heavy arrow represents motion of Pacific plate with respect to North Americanplate [Chase, 1978; Minster and Jordan, 1978]. Submarine contours are in meters.
1 through 4 the northwesterly
plane is chosenas rupture plane; for all other solutions the shallow dipping plane is assumed to represent
event number2, the 1972 Sitka earthquake[Page, 1973]. The choice of the NWtrending fault plane implies a right-lateral sense of strike slip
1. Analysis of the static field of crustal deformation[Savageand Hastie, 1966] and several
along onshoresectionsof the Fairweatherfault
Great 1964Alaskaearthquake(solution 18).
vertical dip of the inferred rupture planes is also in agreementwith field evidence along the Fairweather fault [Stoneley, 1967].
Chugach,St. Elias andFairweatherranges persistently show overthrusting of older landward formations over younger seaward formations
For dip slip solutions in general it is more difficult to choose between the two nodal planes.
[Stoneley, 1967; Plafker, 1967]. The same sense of motion is inferred for the September 10, 1899,
Nevertheless, the combinedweight of geologic, geodetic, and seismologic evidence suggests shallow angle thrusting as opposedto steeply dipping reverse faulting for all dip slip
great earthquakein Yakutat Bay [Tarr and Martin, 1912]. Thechoiceof the steepanglenodalplanes of recent earthquakemechanisms as rupture planes, however, is inconsistent with the geological
seaward,southernblock which is opposite to the
motion which is consistent with field
[Plafker et al.,
based on the
other lines of geologicevidence[Plafker, 1965] favor the shallow angle thrust solution for the
Geologically observedstructures along the
for it would indicate
I I I
00'•::) U'% 0'•0'•0
0 U'% 00,1,,.00-.1" 0'•,'-I O0 U'% O'•(:X:) O'•(:X:)000'•0'•,,.0 C• C',• C%1 •--Ic•
I I I
C%1 • C%1 ß
C'• C•O C•I '•C:) '•C:) '•C:) 0
C•O I.('• 0
C•I '•C:) C•I I.('• I.('•
ZZZZZZZZZZZZZZZZ 0 0 0 0 0 0 C) O')(:X:) 0,1-.1'-.1'-.1'
0 0 U'% 0
0 0 0 0 0 0 0 C',I O'•U'% %00,1U'b(:X:)
,--• r,,- O• -.1' ,.00•
o,I r'-. o') -.1' 000
sense of motion observed. Finally, the choice of the shallow-dipping plane as the fault plane is
model: viz., RM2 of Minster and Jordan . This model predicts (open arrows, Figure 3)
the expression of surficially
which curve with
as far north as Cross Sound with a
rate of about 5.5 to 5.7 cm/yr as well as normal subduction in the eastern Aleutian arc. Figure 3
3. Only the choice of the shallow plane can efficiently accommodatethe kinematics of defor-
also showsslip vectors (solid arrows)derived for the 19 fault plane solutions discussed earlier,
marion along this plate boundary where convergence
of the Pacific and North American plates occurs at several centimeters per year. Dominant shallow angle thrusting, however, does not preclude occasional steep angle reverse faulting to emerge at the surface, as is
slip vectors (numbers 20 through 24) obtained from individual or clusters of fault plane solutions [Stauder and Bollinger, 1966] for aftershocks of the 1964 great Alaska event. Most of the azimuths of vectors are well determined, since they are
geologically observed. On the contrary, it is reasonable to expect that stress concentrations and thus moderate size shallow earthquakes will occur along the steeper part of the imbricate faults. It is also feasible that these faults would be activated by the occurrence nearby of moderate-to-large events on the shallow angle thrust planes. Motion during the great 1964 Alaska event involved secondary high-angle reverse faulting on the Patton Bay fault on Montague Island [Plafker, 1965, 1969]. Similarly,
entirely controlled by the strike and dip of the auxiliary plane which for all dip slip solutions is the better determined plane of the two nodal planes. The general agreement of directions of seismic slip with those predicted by inversion of global plate tectonic data is rather striking. There exist, however, some deviations from this general pattern of the seismic slip directions which appear to correlate with certain tectonic units. Several solutions from events located distinctly
the surface deformation reported in Yakutat Bay for the great earthquake of September 10, 1899
seaward of the Chugach-St. Elias-Fairweather fault system, i.e., solutions 5 to 7 and 13 to 15, yield seismic slip vectors which are rotated clockwise by several degrees • from the calculated directions. We suggest that these events reflect
can be interpreted
motion along the steeper part of imbricate faults that accompanied shallow dip thrusting at depth. An interesting
of Alaska is
in the eastern
fault plane solutions display (Figure 2).
west near Kodiak Island,
motion between allochthonous
Pacific-North Americanplate motion proper.
P axes have rather uniform trends and plunges. The T axes and B axes are fairly consistent in each subgroup and essentially interchange between the two groups, with nearly vertical B axes along the strike slip faults and nearly horizontal B axes for the dip slip faults. The strike of the B axes for dip slip solutions has a tendency to be
contrast, all solutions associated with thrusts or strike slip faults which most likely involve plate interaction inland show a much better agreement with the calculated directions of plate motions. The northeasterly directed seismic slip vectors along or near the shelf edge structure (i.e., for solutions 5 to 7 and 13) indicate that
aligned with the average trend of faults,
the Yakutat wedgeis overriding the Pacific plate
in a SW directed
A few fault
solutions for aftershocks of the 1964 great earthquake in the Prince William Sound region were given by Stauder and Bollinger  and appear to show locally more complexdeformation (e.g., solution 14). These authors relate the few focal mechanismanomalies to the change in trend of geological
from the Aleutian
motion which contributes
to the relaxation of Pacific-North Americanplate motion. The motion along the shelf edge structure is almost pure thrusting and almost perpendicular to the strike of the shelf edge. We do not directly know the rates of relative motion that are associated with this northeasterly directed convergence, but estimates
of such rates
trend (longitudes W of 148øW)to an E-W trend in the eastern gulf (east of 144øW).
derived from simple vector diagrams involving three mobile but rigid units: (1) the Pacific
plate, (2) a border land terrane (i.e., the wedgeshaped Yakutat block), and (3) the North American plate
A primary objective of this paper is to improve our understanding of how the relative motion
(approximately N30OE) of the thrust solutions 5 to 7 at the shelf edge, that of the predominant
between the Pacific and North American plates is accommodatedin the eastern Gulf of Alaska. In
strike slip solutions 3 and 4 (approximately N20oW)associated with the Fairweather fault, and
section we compare the slip
from focal mechanismsof individual earthquakes in the Gulf of Alaska with those predicted by global plate motion analysis. This approach will allow us to discuss discrepancies bewteen the observed and predicted directions of slip in terms of local and regional tectonic structures. Two recent studies [Chase, 1978; Minster and Jordan, 1978] •oncerned with the patterns of global plate motions yielded nearly identical results for the pole and rate of relative motion between the Pacific
we use here as reference only one
motion of the
Pacific versus the North American plate predicted by model RM2 one obtains a rate of about 1 cm/yr of a N30"E directed convergence between the Yakutat block and Pacific plate. The computed rate of convergence between the Yakutat block and the Pacific plate would be raised to almost 2 cm/yr if for the Yakutat block-North American plate boundary we assume pure strike slip motion along the mapped trace of the Fairweather fault (N30"W) instead of the observed slip direction (N20øW) of solutions
3 and 4.
convergence may slightly
Also the rate
vary along the shelf edge
Perez and Jacob'
Gulf of Alaska
A 58 ø
km 152 ø
Slip vectors obtained from fault plane solutions of shallow earthquakes in
mechanisms in Figure 2; solutions 20 to 24 are from Stauder and Bollinger . Open arrows represent calculated directions of Pacific-North America plate localities as earthquakes 1 to 24 (after global model RM2of Minster Stars represent andesitic volcanoes of Quaternary age. Queries occurrences. Note the agreement between direction of slip vectors
plane solutions and the calculated directions mechanisms 5 to 7 and 13 and 14.
motion at the same and Jordan ). indicate uncertain derived
of plate movement. Exceptions are
Heavy broken lines AA' and BB' show the locations
sections displayed in Figures 5a and 5b, respectively.
boundaryif the pole of relative rotation between
The strike slip portion of this boundary(Queen
very distant. Therefore'ourestimate for the rate of convergence(•1 cm/yr) is somewhat uncertain, and values between 0.5 and 2.0 cm/yr seem
ruptured along almost its entire length by a few large earthquakes. Proceedingfromsouth to north their years of occurrenceand magnitudesare 1949
plate and the Yakutat block is not
permissible. Independent theexact this 19•8 (M =(•8 1)--' 8.2). 1972(M•heir = 7.3), 1927(Ms = 7.1), and result indicates that theof eastern Gulfrate, of Alaska inferred rupture zones requiresa tectonicmodelwith moreelements than
(Figure•) forma verynarrowbelt andnearlyabut
two rigid plates.
without overlapwith the possibleexceptionthat a
quences for tectonic
This complicationhas conseinterpretations
ment of seismic hazards.
between the 1949 and 1972 aftershock
zones [Sykes, 1971; McCann et al., The major
includes the recent St. al.,
The spatial distribution of moderate-to-great,
shallowfocusearthquakes in the easternGulf of
qualitatively supports the
notion that high rates of slip
occur on the
Elias earthquake [Lahr et
February28, 1979 (M = 7.2),
and two great
earthquakes (M> 8) insSeptember 1899. Thespace time patterns o• this seismicity [McCann et al.,
1980b; Tarr and Martin, 1912] will be reviewed in
Fairweather and Chugach-St. Elias fault systems and that low rates occur along the shelf edge boundary of the Yakutat block. Figure 4 shows instrumentally located earthquakesfor the period 1899 to 1979 with magnitudesM > 6.0. The data
a later section when evaluating the seismic potential of the Yakatagaseismic gap [McCannet al., 1980b] located between Icy Bay and Kayak Island. The point we emphasizehere is that in the region near and west of Yakutat Bay large
are taken from NOAAcatalogues. The zone of most
earthquakes have occurred historically.
dimensions measure at least a few tens of
quakes (Ms > 7.8) andmostlargeevents(Ms > 7.0) kilometersalongthe plungein the N-Sdirection is confine•to the Queen Charlotte,Fairweather, andat least as much or substantiallymorealong
systems of faults.
the E-W direction.
58 ø 6.7
1949 8.1 QCH
Earthquakesfrom 1899 to February 1979 with magnitudeslarger than 6.0 in the
eastern Gulf of Alaska region.
Diagonal hatching patterns indicate known rupture zones
of the larger earthquakes in the region, whosedates and magnitudesare also given [Sykes, 1971; Page, 1973, 1975; Lahr et al., 1979]. Hexagonsdenote epicenters of earthquakes of M > 7.8; triangles, shocks of 7.0 6.0rob
5.O- 5.9 4.0- 4.9
o+ ½';;'.:•iii',?.1 b•
148 ø Fig. 6. catalogs
Shallow earthquakes located in southeast Alaska as reported in the ISC and PDE from January 1964 to December 1976 and January 1977 to February 1979, respec-
tively (modified after McCann et al. ). The band of dark shading indicates the southern leading edge of the Pacific-North American plate boundary along which we believe most of the Pacific-North American plate motion is taken up by relatively frequent large earthquakes Star and stippled area near 60 5øN 141øW denote epicenter and rupture zone of the Ms = 7.2 St. Elias earthquake of February 28, 1979. Region 1 indicates the approximate area of the Yakataga seismic gap. It has been nearly quiescent for events greater than magnitude 4.0 since 1964. Many of the events inside this region occurred within a few months after the 1964 great Alaska event. Note the ring of modest activity surrounding the gap region. A northwest trend of seismic activity within this zone (short dashes on land) appears to divide the gap into two parts (la and lb). The subregions are interpreted to rupture either together in a single great earthquake or in two separate future earthquakes. Region 2, which also appears to be divided in two subregions (2a and 2b) by a northeast trend of seismicity, is a zone underlying the entire Yakutat block (Figures 1, 5a and 5b) that may have a finite potential for large or great thrust earthquakes. Such earthquakes may have, however, a low probability to occur within the next few decades because of inferred low rates
of convergence between the Yakutat
block and the Pacific
gap, and the potential slip that may have accumulated in the region from 1899 to the present.
and hence without significant release energy. A more probable interpretation gently dipping thrust zone underlies
We have labeled a region 2 (Figure 6) which corresponds to the offshore tectonic province previously described as the Yakutat block. We have subdivided this region into two subregions on
Yakutat block and relieves strain by large or great but rare thrust earthquakes. The
beneath the entire
Yakutat wedge may not break in
one single •iant
rupture but may break in
of seismic is that a the entire
into regions 2a and 2b schematically
points out the possibility
that the thrust zone a
subregions 2a and 2b). The estimated rate of SW-NE convergence between the Yakutat block and the Pacific plate (approximately 1 cm/yr) is
sequence of separate events. From the inferred rate of relative motion (:1 cm/yr) between the Yakutat block and the Pacific plate, repeat times
of the order
potential of this offshore region in the eastern Gulf of Alaska. It is possible that some portion of this motion may be taken up by aseismic slip
hundreds of years
be expected for this offshore region. We do not know the seismic history for such an extended period, and despite the few large and moderate
wide thrust zone. Yet if at
discussed (Figure 4),
possess any record there
such events did occur time
beneath this exceed
margin seems to
magmaswhich form a poorly developed,
was several hundred years ago then the region could have attained again a high potential for future great earthquakes.
dipping slab. The volcanic arc is best developed in the Wrangell Mountains. Northwest of the Wrangell volcanoes the arc abruptly terminates for
Since we do not have any additional new data on zones of apparently slow recent deformation along the Denali, Totschunda, Dalton, Duke River, and Chatham Strait faults, we follow other investi-
yet unexplained reasons. In the eastern Gulf of Alaska no intermediate-depth seismicity is associated with the obliquely subducting portion of Pacific lithosphere.
gator's conclusions [i.e., Woodward-Clyde Consultants, 1978] that these faults should be considered as still active and capable of producing moderate to large events. The magnitudes of their associated maximumcredible earthquakes and their occurrence times must be defined from the faults' geologic records and
A major result of this study is that the transition from the transform plate boundary along British Columbia to the convergent plate boundary at the eastern Aleutian trench is much more gradual than heretofore recognized. The geological importance of a small component of convergence at a predominant transform fault along
than from an almost nonexisting
Recurrence times for
such events may measure many hundreds or
thousands of years.
millions first, Conclusions
If a substantial componentof convetfor
apparently can lead to,
sufficient amount of subduction of ocean floor that causes a volcanic arc and 'other tectonic
Despite the apparent complexity of the tectonics of the eastern Gulf of Alaska, we have attempted to construct a relatively simple tectonic model. This model is consistent with globally computed rates and directions of motion between the Pacific and North American plates but
features not unsirailar to those of regular consuming plate margins. During such overall oblique motion, large terranes can be transformed large distances along the continental margin. The displaced terranes may undergo substantial rotations during their voyage. The Yakutat block
may be the geologically youngest example of such
one other microplate,
The model indicates north-northwesterly motion of the Pacific plate along the Queen CharlotteFairweather fault system at rates increasing from about 5.5 cm/yr in the south to about 6 cm/yr in the north. The sense of motion is predominantly right-lateral strike slip but contains a component of convergencewhich south of Dixon Entrance and north of Cross Soundmay amount to about 1 cm/yr.
eastern Gulf of Alaska, in this case along the Fairweather fault. The Yakutat block may be a former accretionary prism or fore arc terrane, largely formed prior to Miocene when the Farallon plate was still subducting beneath the margin of western North America [Atwater, 1970]. Therefore the Yakutat terrane may be only a rootless thrust slice rather than a fully developed continental
million yearscauses a significant amount of subduction of Pacific oceanic lithosphere beneath the western continental margin of British Columbia and southeast Alaska; transform faulting predominates, however, and results in probably more
Despite these uncertainties with respect to the origin and crustal structure of the Yakutat block, it is apparent that its collision with the leading edge of the North American plate causes a major orogenic event. The Chugach-St. Elias range is
than 500 km of right-lateral displacement during the last 10 m.y. The depth to which the leading
the surface expression of a deeply penetrative crustal shortening process in the North American place caused by the collision of the Yakutat block with the preexisting continental margin. Molnar
Fig. plane The
(Opposite) Compilation of the 13 fault
newly derived are
those of Table 1 and as used throughout this paper. Lower hemisphere equal-area projections are shown. Solid dots are for compressional, open dots for dilatational, and crosses for nodal P arrivals. Most polarities are from reliable readings of long-period records (large
symbols); a few short-period readings (small symbols) are used but are considered unreliable. Short lines indicate S wave polarizations. Solid nodal planes are well donstrained, broken ones poorly constrained. P = inferred axis of maximum principal stress, T = axis of minimum principal stress. Except for solutions 1 and 2, which show mostly strike slip, all other solutions show mostly dip slip with faulting inferred to occur on the shallow dipping plane.
style of deformation to the relative youthand inferred buoyancy - of Pacific lithosphere in the eastern Gulf of Alaska. We propose as an alternative explanation that the Cordil letart-type orogeny of this subduction zone is caused by the arrival of slivers of buoyant borderland terranes, rather than by the buoyancy of young Pacific lithosphere itself. The slivers of buoyant borderland
plate (acting as a conveyor belt) to the of formerly accreted terranes that make up
the North American continental plate. Because of their lower crustal density the arriving terranes resist subduction to any substantial depth. Instead they detach from the Pacific plate, collide with and deform the existing continental margin, while subduction of the Pacific (oceanic) 1 ithosphere beneath the continental margin continues. The Yakutat block may be in the early
stages of both collision with North America and detachment from the Pacific plate. A major seismic gap has been identified to extend along strike of the thrust belt from about Icy Bay (west of Yakutat Bay) to about Kayak Island. The region of this Yakataga seismic gap
is constrained by only a few P wave readings and S wave polarizations. The distribution of stations is, however, good enough to properly select both nodal planes. Page  reports a northwesterly trending aftershock sequence associated with these events coinciding with the strike of one of
energy since it last ruptured in 1899 to release an average slip of several meters over an area of about 150 by 100 km. This strain would be
sufficientfor at least onegreatearthquake. The timing
such a great
sequence of several large events is uncertain and could be either imminent or a few decades away. This region is certainly a prime target for testing earthquake prediction efforts.
to be the
only its auxiliary (almostvertical) %lanewell constrained.
plane (the inferred fault plane) is determined from S wave polarizations and their convergence toward the T axis. The solution is almost identical to the one given by Lahr et al. . Solution
which we interpret
fault plane. Solution 8. The solution of the February 28, 1979, St. Elias event of M = 7.2 has
earthquake that occurredin 1965. Both s nodal planes are poorly determined.
The purpose of this appendix is to give the reader a clear idea of the quality and reliability of each focal mechanismpresented in this paper (Figure 7). Our solutions are primarily based on long-period first motions of P and PKPphases and, where necessary and possible, on polarization of S phases. These readings were taken by inspecting the records of the World-Wide Standardized
at seismic stations
motions with low signal-to-noise
(WWSSN), the Canadian Seismograph Network, and
long-period seismic stations
of the auxiliary (almost vertical) plane toward the vertical would lead to a normal faulting solutiou, such as the ones given by Chandra . We favor the thrust faulting solution for its consistency with solutions 8 to 19. It does not stand, however, on its ownmerits. We point out the possibility that misleading readings of ratios
near the subvertical
for such a small
belongingto the WWSSN, e.g., HUA(Huancayo, earthquake (mb = 5.7) cauleadto whatweinfer to Peru), HIG (Institute
of Geophysics,Hawaii), RDJ
(Rio de Janeiro, Brazil),
PAL (Palisades, U.S.A.).
Wewill comment also on agreement or disagreement
be misinterpretation of the focal mechanism(i.e., normal versus thrust faulting).
Solution 10. This solution is for a mb = 5.8 ß
with mechanismsolutions given by other authors
earthquake which occurred in 1971 in the VlClnlty
for some of the same events.
of the St. Elias mouutains.
Earthquake and focal
mechanismparameters are summarizedin Table 1. Solution
is for an aftershock
Only one nodal plane
plane) is defined by P wave first
The second plane
of the 1972 Sitka earthquake of Ms = 7.6. Both nodal planes are poorly constrained. The solution, however, is consistent with that of the
first motions iu conjunction with the inferred position of the T axis towards which the S wave polarizations converge.
main shock. Solution
Solutions 11 to 19. The same comments apply for these solutions as for solution number 10. In
earthquake (M = 7.6)
is very well determined.
general, the auxiliary (subvertical) plane is well
Chandra[1974• gives a different solution, with
determinedfrom both first motionsand S wave
the northwest trending plane (Figure 7) dipping to the southwest and the northeast trending plane dipping to the northwest. His solution, however, has over 15 polarities which are inconsistent with
polarizations (an exception is solution number 17). The near-horizontal nodal plane (chosen to be the fault plane) was determined with the help of S wave polarizations. Solutions 13, 14, and 19 are aftershocks of the 1964 great Alaska event, which is denoted as number 18. These four solutions are after Stauder and Bollinger [ 1966].
the nodal planes he chooses. Our solution and choice of nodal planes (northwest trending, dipping to the northeast) does not leave any polarity
agreement with the geologic and tectonic of the region discussed in the text.
Solution 3. This solution of the 1927 southeast Alaska earthquake of M = 7.1 is replotted
Acknowledgments. The seismotectonic portions of this study were supported by the Department of
after Stauder . RegaSrdless of the date,
Energy,Division of BasicEnergySciences,under
this solution is a very good one, although its location may be poor. Stauder constrained both nodal planes by means of first motions as well as S wave polarization data. Solution 4. The same comments for the solution of the 1958 great Fairweather event (M = 8.2)
contract DE AS002 ERO76-3134. The seismic hazards part was supported by the National Oceanic and Atmospheric Administration's Alaska Outer Continental Shelf Environmental Assessment Project (OCSEAP) under contract NOAA-03-5-022-70, Research Unit 16. John Davies, Walter Pitman,
except that its location is well constrained. Solutions 5 to 7. They represent the main shock (M = 6.7) and two aftershocks of the 1973
critically read the manuscript and made numerous suggestions for its improvement. Lynn Zappa and Mary Anne Avins typed many versions of this paper;
shelf-edle sequenceoff shore,of CrossSound.
Kazuko Nagaodrafted the figures. M. Barazangi,
5 and 6 are well
gives almost identical solutions. 7, a weak aftershock of the 1973 event,
G. Plafker, J.P. Foulas, G. Thrasher, and R. yon Huene offered many valuable comments for
which we are grateful.
Facts and speculations on the role of
terrigenoussediments,in Island Arcs; DeepSea Trenches and Back-Arc Basins, Maurice. Ewin• Set., vol. 1, edited by M. Talwani and W. C.
Atwater, T., Implications of plate tectonics for the Cenozoictectonic evolution of western North America, Geol. Soc. Am. Bull., 81, 3513-3536, 1970.
Bouguergravity map of Alaska
D.C., 1977. Jones, D.L., and N.J. Silberling, Mesozoic stratigraphy, the key to tectonic analysis of southern and central Alaska, Geol. Surv. Open
File Rep. U.S. 79-1200, 37, 1979.
1:2,500,000, Geophys. Invest. Map MAP-GP-913, U.S. Geol. Surv., Res:on, Va., 1977.
Jones, D. L., N.J. Silberling, and J. Hillhouse, Wrangellia -- A displaced terrane in north-
Berg, H. C., D. L. Jones, and P. J. Coney,Map
westernNorth America,Can. J. Earth Sci., 14,
showingthe pre-Cenozoic tectono-stratigraphic terranes of southeasternAlaska and adjacent
areas, Geol. Surv. OpenFile Rep.U.S. 78-1085, 1978.
Boa:wright,J., Preliminarybodywaveanalysisof the St. Elias, Alaskaearthquake of February 28, 1979, Bull. Seismol. Soc. Am., 70, 419-436, 1980.
Late Cenozoicstructure of the
Proceedingsof the 6th Alaska Geolosical Society Symposium, edited by A. Sisson, pp. II-I30, Anchorage,Alaska, 1979. Chandra, U., Seismicity, earthquake mechanisms,
and tectonics along the westerncoast of North
America, from 42øNto 61øN, Bull ß Seismolß Socß Am., 64, 1529-1549, 1974. Chase, C. G.,
2565-2577, 1977. Kanamori,H., The energyrelease in great earthquakes,J. Geophys. Res., 8_•2, 2981-2988,1977. Kelleher, J.A., Space-time seismicity of the
Alaska-Aleutianseismiczone, J. Geophys. Res., 75, 5745-5766,1970.
Kelleher, J. A., and J. Savino, Distribution of seismicity before large strike slip and thrust-
type earthquakes, J. Geophys. Res., 80, 260-271,
Kelleher, J.A., L.R. Sykes, and J. Oliver, Possible criteria for predicting earthquake locations and their applications to major plate boundaries of the Pacific and Caribbean, J.
Geoph•s.Res., 78, 2547-2585,1973.
map of North America King, P.B. , Tectonic 1.'5,000,000, U.S. Geol. Surv., Res:on, Va.,
East Africa, and the rest of the world, Earth Planet. Sci. Let:., 37, 355-368, 1978.
Lahr, J. C., G. Plafker, C. D. Stephens, K. A. Fogleman,andM. E. Blackford, Interim report on
Chase,R. L., and D. L. Tiffin, QueenCharlotte fault-zone, British Columbia, In:. Geol.
the St. Elias earthquakeof 28 February1979, Geol. Surv. OpenFile Rep.U.S. 79-670, 35 pp.,
Churkin, M., and G. D. Eberlein, Ancient borderland terranes of the North Cordillera: Correlation and microplate tectonics, Geol. Soc. Am. Bull., 88, 796-786, 1977. Davies, J. N., and L. House, Aleutian subduction zone seismicity, volcano-trench separation, and their relation to great thrust-type earthquakes, J. Geoph•s. Res., 84, 4583-4591, 1979. Dehlinger, P., R. W. Couch,D. A. McManus,and M.
Lahr, J. C., C. D. Stephens, H. S. Hasegawa,and J. Boa:wright, Alaskan seismic gap only partially filled by 28 February 1979 earthquake, Science, 207, 1351-1353, 1980. McCann,W. R., S. P. Nishenko, L. R. Sykes, and J. Krause, Seismic gaps and plate tectonics: Seismic potential for major plate boundaries, Pure Appl. Geophys., 117, 1082-1147, 1980a. McCann, W.R., O.J. Perez, and L.R. Sykes,
Congr., Rep. 24th Sess., 17-24, 1972.
Gemperle, Northeast Pacific
Sea, vol. 4_, (2), edited by A. E. Maxwell, pp. 133-189, Interscience, New York, 1970. Fedotov, S. A., Regularities of the distribution
of strong earthquakesin Kamchatka,the Kurile Islands and northeastern Japan (in Russian), Tr.
Yakataga seismic gap, southern Alaska:
and earthquake potential, 1314, 1980b. Minster, J.B., and T.H.
Science, 207, 1309-
plate motions,J. Geophys. Res., 83, 5331-5354, 1978.
Zemli Akad. Nauk SSSR, 36, 66-93,
Mogi, K., Somefeatures of recent seismic activity
Fitch, T.J., Plate convergence, transcurrent faults and internal deformation adjacent to
great earthquakes, Bull. garthquake Res. Inst. Univ. Tokyo, 47, 395-417, 1969.
southeast Asia and the western Pacific, J. Geoph•s. Res., 77, 4432-4460, 1972. Hillhouse, J. W., Paleomagnetism of the Nikolai
Molnar, P., and T. Atwater, Interarc spreading and
in and near Japan, 2, Activity
J. Earth Sci.,
Irving, E., Paleopoles and paleolatitudes America and speculations about
14, 2578-2592, 1977.
of North displaced
Isacks, B., L. R. Sykes, and J. Oliver, Focal mechanisms of deep and shallow earthquakes in the Tonga-Kermadec region and the tectonics of
Geol. Soc. Am. Bull.,
Jacob, K.H., Trench-volcano
gap along the Alaskan-Aleutian
the age of subducted oceanic lithosphere,
41, 330-340, 1979.
Monger, J. W. H., Upper Paleozoic rocks of the western Canadian Cordillera and their bearing on
1832-1859, 1977. Oh: ake, M., T.
Can. J. Earth Sci.,
gap near Oaxaca, southern Mexico as a precursor
to a large
Appl. Geoph•s., 115, 375-385, 1977. Page, R.A., Crustal deformation on the Denali fault
K. Nakamura, and J.N.
before and after
77, 1528-1533, 1972. Page, R. A., The Sitka,
• Alaska, earthquake of 1972
--An expected visitor, Earthquake Inform. Bull. 5, pp. 4-9, U.S. Geol. Surv., Reston, Va., 1973.
Page, R. A., Evaluation of seismicity and earthquake shaking at off-shore sites, Proc. Annu. Offshore Techno1. Conf., 3, 179-190, 1975. Perez, O. J., and K. H. Jacob, St. Elias, Alaska earthquake of February 28, 1979: Tectonic setting and precursory seismic pattern, Bull.
Seismol. Soc. Am., 70, 1595-1606, 1980. Plafker,
of October 24, 1927, Geofis.
Pura App1., 44,
135, 1959. Stauder, W. S., The Alaska earthquake of July 10,' 1958: Seismic studies, Bull. Seismol. Soc. Am.,
50, 243-322, 1960. Stauder,
quakes, Adv. Geophys., 9, 1-76, 1962. Stauder, W. S., Tensional character of earthquake foci beneath the Aleutian trench with relation
to seafloor spreading,J. Geophys.Res., 73(24),
G., Tectonic deformation associated with
the 1964 Alaskan earthquake, 148(3678), 1675-1687, 1965.
Plafker, G., Geologic map of the Gulf of Alaska Tertiary province, 1:500,000, Misc. Geol.
7693-7701, 1968. Stauder, W. S., and G. A. Bollinger, The focal mechansmof the Alaska earthquake of March 28,
1964, and of its aftershock sequence, J. Geophys. Res., 71, 5283-5296, 1966.
Invest. Map 1-484, U.S. Geol. Surv., Reston,
Stone, D. B., and D. R. Packer, Paleomagneticdata
Va., 1967. Plafker, G., Tectonics of the March 27, 1964 Alaska earthquake, Geol. Surv. Prof. Pap. U.S. 543-1, 74 pp., 1969.
from the Alaska Peninsula, Geol. Soc. Am. Bull., 90, 545-560, 1979. Stoneley, R., The structural development of the Gulf of Alaska sedimentary province in southern
Plafker, G., T. Hudson, and D.H. Richter, Preliminary observations on Late Cenozoic displacements along the Totschunda and Denali fault system, Geol.Surv. Circ. U.S. 733, 67-69, 1977.
Plafker, G., T. Hudson, T. Bruns, and M. Rubin, Late Quaternary offsets along the Fairweather fault and crustal plate interactions in southern Alaska, Can. J. Earth Sci., 15, 805-816, 1978. Richter, C. F., Elementary Seismology, 768 pp., W. H. Freeman, San Francisco, Calif., 1958. Richter, D. H., and N. A. Matson, Jr., Quaternary faulting in the eastern Alaska Range, Geol. Soc.
Am. Bull:, 82(6), 1529-1540, 1971. Alaska, J. Geophys.Res., 80, 3786-3790, 1975. Savage, J. and L. M. Hastie, Surface deformation faulting,
Res., 71, 4897-4904, 1966.
Savage, J. C., M. Lisowski, and W. H. Prescott, Strain accumulation across the Denali fault in the Delta river canyon, Alaska, J. Geophys.
Res., in press, 1980. in
Soc. London, 123, 25-57,
Geophys. Res., 72, 2131-2153, 1967.
Sykes, L. R., Aftershockzones of great earthquakes,
prediction for Alaska and the Aleutians, Geophys. Res., 76, 8021-8041, 1971.
Surv. Prof. Pap. U.S. 69, 1912. Thatcher, W., and G. Plafker, The 1899 Yakutat Bay, Alaska earthquakes, paper presented at
Geophys., Durham, England, 1977. Tobin, D. G., and L. R. Sykes, Seismicity
tectonics of the northeast Pacific Ocean, J__. Geophys.Res., 73, 3821-3845, 1968. yon Huene, R., G. G. Shot, Jr.,
and J. Wageman,
Continental margins of the eastern Gulf of
Alaska and boundaries of tectonic plates, Mem. Am. Assoc. Petrol. Geol., 29, 273-290, 1979. Woodward-Clyde Consultants, Offshore Alaska
look, Volcanic Regimes in Canada, Spec. Pap. Geol. Assoc. Can., 16, 3-24, 1977. Stauder, W. S., A mechanismstudy: The earthquake
Tart, R. S., and L. Martin, The earthquake at Yakutat Bay, Alaska, in September 1899, Geol.
Seismic Exposure Study (OASES), 5 vols.,
Souther, J. G., Volcanism and tectonic environmerits
IASPEI/IAVCEI Assembly, Int.
Savage, J. C., Further analysis of the geodetic strain measurements on the Denali fault in
associated with dip slip
Sykes, L. R., Mechanism of earthquakes and nature of faulting on the midoceanic ridges, J.
(Received October 5, 1979; revised July 21, 1980; accepted August 22, 1980.)