THE CRUSTAL STRUCTURE OF KILAUEA AND MAUNA LOA
VOLCANOES, HAWAII, FROM
SEISMIC REFRACTION AND GRAVITY DATA
A DISSERTATION
SUBMITTED TO THE DEPARTMENT OF GEOPHYSICS
AND THE COMMITTEE ON GRADUATE STUDIES
OF STANFORD UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
By
John Justin Zucca
June
1981
certify that I
I
have read this thesis and that in my opinion it
fully
adequate,
is
in scope and quality, as a dissertation for
degree
the
of Doctor of Philosophy.
(Robert L. Kovach, Principal Advisor)
certify that I
I
have read this thesis and that in my opinion it
is fully adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
certify that I
I
have read this thesis and that in my opinion it
is fully adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
(Robert J. Geller)
certify that I
I
have read this thesis and that in my opinion it
is fully adequate, in scope and quality, as a dissertation for
the degree of Doctor of Philosophy.
Approved for the University Committee on Graduate Studies:
II
ACKNOWLEDGEMENTS
I
wish to express my appreciation and gratitude to my principal
advisors: Dr. Kovach of Stanford University and Dr. Hill of the U.S.
Without their help and guidance, this work would
Geological Survey.
not have been completed.
Much of the credit for the success of this work is due to the
people who helped with the support work.
John Coakley and Ed Criley
did an excellent job of deploying the temporary seismographs.
I
am
also indebted to Bob Koyanagi and Dr. Fred Klein of the Hawaiian
Volcano Observatory for making available the tapes containing the data
from the permanent seismograph stations.
Dr. Fred Duennebier of the
Hawaii Institute of Geophysics gratiously provided the Ocean Bottom
Seismograph data.
Elsie Hircher speedily typed the rough drafts of
the thesis.
The U.S. Geological Survey supported this work by financing the
field operations and providing me with employment during the bulk of
my graduate career.
Finally, I
would like to thank Anne Garvey for her patience,
encouragement, and critical readings of rough drafts of this
manuscript during the months of writing.
111
TABLE OF CONTENTS
ACKNOWLEDGEMENTS
iii
LIST OF TABLES
vi
LIST OF ILLUSTRATIONS
vii
ABSTRACT
1
CHAPTER
I
II
INTRODUCTION
3
Geologic Setting
Previous Work
6
9
DATA COLLECTION
13
USGS Permanent Stations
USGS Temporary Stations
HIG Ocean Bottom Stations
25
25
31
40
111 ANALYSIS OF SEISMIC REFRACTION DATA
Kilauea Profile
Method of Analysis
Velocity Structure: S.E. Flank
Velocity Structure: Rift Zones and Summits
Mauna Loa Profile
Method of Analysis
Velocity Stucture
Teleseismic Data
Amplitude Modeling
IV
42
.
43
49
55
62
79
82
GRAVITY DATA AND INTERPRETATION
91
Data
92
Computational Methods
Velocity-DensityRelationships
Density Structure
Kilauea Profile
Mauna Loa Profile
96
98
IV
102
103
V
DISCUSSION
105
High Velocity and Density Regions in the Crust
Kona Coast Anomaly
Forceful Injection of Magma into Rift
Zones and the Kalapana Earthquake
Crustal Thickness
Loss of High Frequency Energy at Mauna Loa
Conclusion
Recomendations for future work
REFERENCES CITED
.
...
110
112
112
114
115
117
119
121
V
LIST OF TABLES
Table
2.1
Miniranger positions, Kilauea Profile
18
2.2
Miniranger positions, Mauna Loa Profile
18
2.3
Shot locations, Kilauea Profile
22
2.4
Shot locations, Mauna Loa Profile
23
2.5
Locations of permanent seismographs
30
2.6
Locations of temporary seismographs
30
vi
LIST OF ILLUSTRATIONS
Figure
1.1
Map of the northwest Pacific Ocean
5
1.2
Map of Hawaii island volcanoes
10
2.1
Locations map for Kilauea Profile
15
2.2
Locations map for Mauna Loa Profile
16
2.3
Bathymetry along Kilauea Profile
20
2.4
Bathymetry along Mauna Loa Profile
21
2.5
Record section at station HLP from the
Kilauea Profile.
26
Record section at station CAC from the
Mauna Loa Profile
27
2.7
Record section at station XII from the
Mauna Loa Profile
28
2.8
Record section at station HSS from the
Mauna Loa Profile.
29
Record section at station BSC from the
Mauna Loa Profile.
32
Record section at station ISB from the :
Mauna Loa Profile.
33
Record section at station PUO from the
Mauna Loa Profile.
3^
Record section at station PWA from the
Mauna Loa Profile
35
c
2.6
2.9
2.10
2.11
2.12
.
2.13 Record section at station SMR from the
2.14 Record section at station WST from the
2.15
Mauna Loa Profile
37
Record section at OBS RTI from the
Kilauea Profile
38
VII
2.16 Record section at OBS TOK from the
39
Kilauea Profile
.
44
3.1
Composite traveltime curve for the Kilauea Profile.
3.2
Calculated velocity structures for the
south flank of Kilauea
45
3.3
Map of the Hawaiian Islands showing
orientation of the Molokai fracture zone
48
3.4
Seismic and structure profile across
the east rift of Kilauea
51
3.5
Seismic and structure profile across
the southwest rift of Kilauea
53
3.6
'Semi-reversed* traveltime curve from
the Mauna Loa profile
56
3.7
Upper crustal structure beneath Kona coast
58
3.8
Example velocity structure
59
3.9
Example ray diagram
61
3.10
Ray diagram for station BSC
63
diagram for station CAC
64
3.11 Ray
3.12
Ray diagram for station HSS.
65
3.13
Ray diagram for station ISB
66
3.14
Ray diagram for station XII
67
3.15
Ray diagram for station PUO
68
diagram for station SMR
70
Ray diagram for station WST
71
3.17 Ray
3.18
3.19 Calculated velocity structure for
the Mauna Loa Profile
3-20 Perspective view of Mauna Loa velocity
structure
VIII
73
74
3.21 Ray diagram for teleseismic rays
80
3.22 Results of teleseismic modeling
81
3.23 True
amplitude record section for station PWA
83
3.24 True
amplitude record section for station PUO
84
3.25 Observed
3.26
85
amplitudes at PWA
87
Ray diagram for modified velocity structure
3.27 Moho and upper mantle velocity structure
88
3.28
89
Synthetic seimogram for station PWA
3.29 Calculated
amplitudes at PWA
90
4.1
Bouguer gravity of Hawaii.
93
4.2
Density model for Kilauea Profile
95
4.3
Density model for Mauna Loa Profile
97
4.4
Velocity-density curve for all types of rocks
99
4.5
Velocity-density curve for Hawaiian rocks
100
4.6
Velocity-density relations for the
Bay-of-Islands ophiolite.
...
5.1
Vp and Vs from Bay-of-Islands opiolirte.
5.2
Seismograms from stations DAN & SWR
IX
.—
......
100
108
116
1
ABSTRACT
In November 1976, and October 1978, the U.S. Geological Survey
established two 100-km-long, on-shore/off-shore seismic refraction
profiles: one located perpendicular to the west (Kona) coast of Hawaii
Island and the other located perpendicular to the southeast (Kau-Puna)
coast of Hawaii Island.
The
1976
profile was established with the
cooperation of the Hawaii Institute of Geophysics.
These profiles lie
across Hawaii's two most active volcanoes, Mauna Loa and Kilauea.
Combined analysis of these profiles along with the gravity data and
teleseismic P-wave residuals for the profiles suggest a complicated
crustal and upper mantle structure for the island.
Beneath the
Kau-Puna coast, the oceanic crust dips about 2° toward the island
compared to a dip of about
coast.
3-3°
increasing to
8.5° under the Kona
The total vertical displacement of the Moho is about 6km
beneath a point near the summit of Mauna Loa.
velocity is
The unreversed Pn
7.9 km/sec beneath the Kau-Puna coast and 8.2 km/sec
.
beneath the Kona coast
Analysis of the profiles suggests that the volcanic rift zones
of Mauna Loa and Kilauea are cored with high velocity, high density
rocks (Vp about
6.9 km/sec,
density about 2.9 g/cc), which reach to
within a few kilometers of the surface of the island.
The rift zones
widen with depth and coalesce such that a large part of the volcanic
pile is composed of high velocity and high density rift zone rock.
2
High velocity and high density rocks are also observed in
regions of the crust with no clear surface expression as rift zones.
The data suggest that an old, buried rift zone, possibly related to
Haulalai volcano, lies just onshore, parallel to the Kona coast.
also suggest that the north flank of Mauna Loa is a rift zone.
They
3
CHAPTER
I
INTRODUCTION
4
Hawaii Island is located in the middle of the Pacific Ocean on
the southeastern end of the linear chain of islands comprising the
Hawaiian Ridge, a subset of the Hawaiian-Emperor Seamount Chain
(figure 1.1).
These islands and seamounts are composed chiefly of
basaltic rock which is increasingly younger in age towards the
southeast end of the chain (McDougall, 1964).
The southern end of
Hawaii Island, and its offshore extension, Loihi Seamount, is the
youngest part of the chain and is the site of active volcanism.
Kilauea and Mauna Loa volcanoes, which compose the south end of Hawaii
Island, have been active during historic times.
The isolated position of these volcanoes, away from the
complications of continental settings, provides an ideal setting for
the study of active volcanoes.
This thesis examines geophysical data
collected on Kilauea and Mauna Loa volcanoes in an attempt to better
understand the structure of the volcanoes and their relationship to
the surrounding crust.
The main scientific problems to be addressed
1) What is the nature of the transition from the oceanic crust
onto the crust beneath the island?
Does this transitional
structure play any role in controlling the large earthquakes
that have been observed around the island of Hawaii?
2) What part does the presence of high velocity and high density
material in the volcanic rift zones play in the structure of
the island?
5
* >^ni
o
1
)
m"
v
■*\
1
"
\
5V|
i
jS]A-»
$
"
4
/ F* :Sf4r
.\
IS
fi
o
ft
i##/
//
ZZ^^b
o
MP£RO*
RO#
£EMP£
<-3
S£ AMCU«TS
0
O
Ay?
O
w
If)
—
co z
o
m
o
O
-
Map showing generalized bathymetry of the northwest
Figure 1.1
Pacific Ocean. Subaerial land masses are shaded black.
Submarine contours are at 3 and 5 km. (after Chase et al. ,
1971; reproduced from Ellsworth, 1977)
6
3) How much thicker is the crust under the interior of the island
compared with that under the coastlines?
4) Using a detailed model of the crustal structure of the island,
what can we reasonably infer about the growth process of the
island?
To address these questions, this thesis presents the data and
interpretation of two seismic refraction profiles: one located
perpendicular to the Kona coast of Hawaii and the other located
perpendicular to the Kau-Puna coast.
Most studies of the velocity
structure of Hawaii have dealt with the island proper; this study
provides substantial new data on the structure of the island and its
relation to the surrounding ocean crust.
In addition to the seismic
refraction data, the gravity field over the south end of the island is
analyzed along with other pertinent data.
GEOLOGIC SETTING
The Hawaiian-Emperor seamount chain is a major bathymetric
feature of the central Pacific ocean.
The chain extends south from
the Aluetian trench and covers a distance of about
3500 km in an
almost straight line, broken only by a sharp eastward bend in the
middle, which separates the Emperor chain in the north from the
Hawaiian chain in the south.
The islands and seamounts in the chain
form an age-ordered sequence. The oldest seamounts in the north have
ages greater than 50 my while the youngest volcanoes on Hawaii are
active (Dalrymple, 1973).
The chain rests on lithosphere of
7
Cretaceous age (about 100 my old) which increases in thickness from
about 80 km under the southernmost part of the chain to about 90 km
near the Hawaiian-Emperor bend (Yoshii and others, 1976; Forsyth,
1977).
The weight of the islands and seamounts on top of the seafloor
has resulted in the downward bending of the lithospere in response to
the load.
This action has depressed the seafloor immediately seaward
of the chain to form the Hawaiian Deep. Surrounding the Deep is a
gentle upwarp of the crust of about 500 m in amplitude that extends
out to about 500 km, which is called the Hawaiian Arch (Dietz and
Menard, 1953).
Walcott (1970) has explained both of these features
using a line load on an elastic plate which is broken along the length
of the load.
The arch is particulary well developed around the
younger islands and tends to disappear toward the older parts of the
chain.
This observation led Detrick and Crough (1978) to propose that
the Arch is formed by lithospheric thinning and thermal expansion as
the lithosphere moves over the proposed Hawaiian hot spot.
After the
lithosphere moves away from the hot spot, it cools and thickens, and
Many theories have been suggested to explain the origin of the
Hawaiian-Emperor chain. Although it is beyond the scope of this thesis
to discuss these theories fully, it is worthwhile mentioning the three
main groups of hypotheses: 1) Thermal instability (Shaw, 1973; Shaw
and Jackson, 1973); 2) Propagating rift (McDougall, 1971); and 3) Hot
spot or mantle plume (Wilson, 1963; Morgan, 1971).
The propagating
8
rift model was the first theory concerning the origin of the chain.
The work of McDougall is one of the more recent papers dealing with
this hypothesis. In my opinion, the hot spot model seems the most
reasonable, although significant objections can be raised to all three
hypotheses.
Hawaiian volcanoes generally evolve through four stages of
evolution (Macdonald and Abbott, 1970).
The first stage is called the
'youthful' or 'shield building' stage. Eruptions of tholeiitic basalt
spread out over the ocean floor to gradually build up the edifice of
the volcano.
As the shield reaches the surface of the sea, the basalt
becomes more vesicular and pyroclastic steam eruptions occur.
The
volcano continues to build above the surface of the water until the
top collapses to form a caldera.
At this time the volcano has entered
the second stage or 'mature' or 'caldera' stage. Active volcanism
continues as the caldera repeatedly collapses and refills itself.
Eruptions occur along the flanks of the edifice to form the volcanic
'rift zones'. In the third stage or 'old age' stage, the activity
declines and the lava becomes more viscous and lower in silica
content.
Then the lava forms a steep cap over the caldera, completely
hiding it. This is likely to be the end of life for the volcano.
However, some do go through a fourth stage called the 'posterosional'
stage.
This stage begins after the volcano has lain dormant for a few
million years.
The volcano can then come to life briefly and
violently, erupting silica-poor alkalic olivine basalts,
nephelintites, and basanites.
After this final activity the volcano
is apparently permanently dormant.
9
Volcanoes in the mature stage tend to erupt from the central
summit or along the roughly linear volcanic rift zones that extend
radially from the summit (Macdonald and Abbott, 1970).
The morphology
of the rift zones is characterized by the presence of pit craters,
cinder cones, spatter cones, fissures, and grabens.
Exposure of rift
zones by erosion show that they consist of numerous vertical and
sub- vertical dikes of dense basalt. On volcanoes in the old age
stage, the eruptions chiefly issue from vents in the summit to form an
alkalic cap.
Volcanoes in the 'posterosional' stage erupt principally
on the flanks to form cinder cones such as Diamond Head on Oahu island
Hawaii Island incorporates five separate volcanoes in several
stages of evolution (figure 1.2).
From north to south (also oldest to
youngest) they are: 1) Kohala Mountain (elevation 1680 m), 2) Mauna
Kea (elevation 4200 m ), 3) Hualalai (elevation 2500 m), 4) Mauna Loa
(elevation 4160), and 5) Kilauea (elevation 1240).
Kohala, Mauna Kea,
and Hualalai have all reached the old age stage of evolution even
though Hualalai has been active in historic time.
Mauna Loa and
Kilauea are currently in the mature stage of evolution.
PREVIOUS WORK
The first crustal velocity model for Hawaii Island was developed
by Eaton (1962) using earthquake traveltime data.
His preferred
velocity structure, an average for the southern part of the island,
consists of a three layer crust with the Moho at a depth of
10
Figure 1.2
- Map of Hawaii island showing the five volcanoes
that
solid lines: rift zones, dashed lines:
contacts between lavas from adjoining volcanoes, (reproduced
from Ellsworth, 1977)
comprise the island,
11
approximately 15 km, which is about 3 to 4 km deeper than the average
for the Pacific (Raitt, 1963).
summit of Kilauea
Seismic-refraction surveys across the
(Ryall and Bennet, 1968) and around the major
coastlines of the island (Hill, 1969) reveal both an average structure
consistent with Eaton's model and evidence for high velocity cores
within the rift zones and summit areas of the volcanoes.
These high
velocity cores are also inferred to have a high density and to reach
. , 1979).
within 2 km of the surface of the rift zone (Broyles et al
Recent studies have used data from the USGS network of short
period seismic stations, which has some
40 instruments concentrated on
the southern end of the island (Koyanagi et al., 1978).
Ellsworth and
Koyanagi (1977), and Ellsworth (1977) have inverted P-wave traveltimes
for the structure of the crust and upper-mantle under the island.
They find that upper-mantle velocity variations average only
whereas variations within the crust exceed
H%. Crosson and
*\.6%
Koyanagi
(1979) have inverted traveltimes from local earthquakes to determine a
layered model for the crust beneath the net.
They obtained a Moho
depth of about 12 km and evidence for a pronounced low velocity layer
at the base of the crust.
Estill and Odegard (1979) have augmented
the array with short period stations located on the older islands and
performed a Tau inversion using local earthquake traveltimes.
Their
resultant model is an average for the ridge and is in general
agreement with the models computed from the seismic refraction data.
Several regional geophysical studies have been performed along
the Hawaiian Ridge. Furumoto and others (1968) reported the results
12
of a series of refraction profiles scattered along the ridge.
Their
computed velocity structures are complicated, but in general they show
that the Moho is at a depth of 10 to 13 km off the islands and is
depressed by as much as
23 km under some of the islands. Later
studies by Furumoto et al. (1971 and 1973) are largely consistent with
these results. Malahoff and Woollard (1970) have investigated the
gravity, magnetics, and crustal structure data of the Hawaiian ridge
to test the hot spot hypothesis for the origin of the Islands.
Watts
and Cochran (1974) and Watts (1976 and 1978) have analysed the
free-air gravity over the ridge for crustal structure and lithospheric
flexure.
13
CHAPTER
DATA
II
COLLECTION
14
In November 1976, the U.S. Geological Survey (USGS) in
conjunction with the Hawaii Institute of Geophysics (HIG) of the
University of Hawaii, established a profile of marine shots
perpendicular to the Kau-Puna coast of Hawaii.
In October
1978 a
similar profile was established off the Kona coast of Hawaii by the
These two profiles are hereafter referred to as the Kilauea
USGS.
profile and the Mauna Loa profile, respectively. Although these data
have been described (but not interpreted) by Zucca et al. (1979) and
Zucca and Hill (1980), it is important to discuss aspects of the data
that are pertinent to this study.
The Kilauea profile consisted of
roughly 100 km long (fig. 2.1).
43 shots oriented in a line
The shots were of two sizes:
twenty- three 300 lb shots fired at about 5 km intervals, and twenty 5
lb shots interspersed with the large shots.
The Mauna Loa profile
consisted of 30 evenly spaced shots also oriented in a line roughly
100 km long (fig. 2.2). The shots weighed 300 lbs except for the five
shots closest to the coast, which all weighed 180 lbs. In both cases
the charges consisted of the commercial explosive Tovex, and were
deployed from the fantail of the USGS research vessel 'Samuel P.
Lee.
The shots were detonated by a fuse which burned about 80 sec
allowing the charge to sink to a depth of roughly 60 meters before
firing.
Shot locations were obtained using minirangers. This is a system
of accurately located land-based transponders that communicate on
demand with a ship-based device.
Slant ranges accurate to 1 part in
o
-
Figure 2.1
Map of Hawaii Island showing locations for the Kilauea
Profile. Open circles: 5-lb shots. Closed circles: 300-lb
shots. Dashed lines: rift zones. Triangles: seimograph
stations. Solid straight lines: seismic profiles across
volcanic rift zones. Contours in kilometers of elevation.
*
o
in
o
?2
/
5
\
x
)\\
75
/« \
~<r-\
3 S
(m^°
\5
"
3
/
/
f
<
<
§
©
52
-J
p
O
a.
4
a.
(0
o
4
ri
\ o-»
kv
y, p
3 Q </>
—
YVy
J
.J \ \
\
V
\
<
<
2*
w
i< -2
tv
Q
16
**
«0
< <^
<
<-
CO
CO
d
CO
a
to
t C
r_
a>
JC
43
..
CO
r,
CSJ
C O
o o c
"h
co
-P C E CO
(D D S- £
c am a
o o o. CO
rH
bO
C
"H
S
0
C
" "" bo
O
CO CO E
4J Q> CO
OH -H
.C bO CD
.c
co c co
CO
CO
t«
"
CO C
t, O
CO rH
rH O "D O -H
n lo a-p
H -H co E CO
O O CD
"H
H -P (D
"H -D CJ
rH
..CD
CO CD
CO
5 W
>
CO O « CD rA
3C rH CO rH O
U L bO
CD C CO
0-4
T) IB Im
O
C -H <D
D. <D O £_ 4->
CO rH D. 4-> CD
2h n
c
Cm c c o
1 O CO CD rH
U r, a-h
.
< (/)
O
W
$_
-H
v n
l.
<D 4->
I
Z
CO
3
t, CO <D O
O CD CO -P
<h H
C
O 4J O
C
8+
£o
o
E
S
o
CM
CM
3
bO
"H
hi
"
OU 4-> o j*
17
50,000 can be obtained in this fashion. For each profile, four
transponders were used (figures 2.1 and 2.2, and tables 2.1 and 2.2).
Slant ranges to at least two, and many times three, transponders were
taken at the point where the charge was dropped overboard and at the
point where the water wave from the explosion arrived at the ship.
The ship's location at the charge-drop and water-wave arrival
positions was obtained by minimizing residuals of the set miniranger
distances in a least-squares manner using the algorithm in HYPO7I (Lee
and Lahr, 1975).
The distance between the charge-drop and water-wave
arrival positions averaged about 0.4 km.
In both cases the most
distant shot subtends an angle of 300 with the transponder array on
shore.
The shots were located to an accuracy of plus or minus 0.05 km
along the length of the profile, but the locations are poorly
constrained perpendicular to the profile and may be in error by
several kilometers.
The water-wave from the explosion was used to estimate the
detonation time of the shot.
Its arrival at the ship was detected by
a hydrophone mounted in the hull. For the Kilauea profile, the signal
from the hydrophone was transmitted to the Hawaiian Volcano
Observatory (HVO) on the rim of Kilauea volcano, where it was recorded
on the same time base as the telemetered seismic network.
For the
Mauna Loa profile, the signal from the hydrophone was recorded on the
ship against an IRIG-C time code on a strip- chart recorder.
time was obtained by referring the IRIG-C code to WWVH.
Absolute
In both cases
the firing time was estimated from the water-wave arrival time by
18
GER
LATITUDE
001
002
003
005
19
19
19
19
Table 2.1
155
155
155
155
02.53
23.273
18.61
35.35
ELEVATION(m)
201 .2
2030.6
683.1
659.3
- Miniranger positions used in the Kilauea Profile.
NIGER
001
002
003
004
005
Table 2.2
21.34
29.735
17.85
07.55
LONGITUDE
LAT ITUDE
19
19
19
19
19
30.58
25.38
18.70
44.51
10.61
LONG ITUDE
155
155
155
155
155
55.23
53.08
52.67
57.37
45.53
ELEVATICDN
478.7
271.3
399.7
982.1
3661 .6
- Miniranger positions used in the Mauna Loa Profile
19
making a correction based on the distance between the point where the
shot was deployed and the point where the water wave arrived at the
ship, the velocity of sound in water, and the charge depth.
Relative
timing between shots to plus or minus 0.05 sec was obtained through
this method.
However, the absolute shot times may be systematically
off by as much as 0.2 sec due to errors in calculation of the depth of
detonation and location of the ship.
Tables 2.3 and 2.4 list the shot
locations, sizes, and detonation times for both profiles. Figures 2.3
and 2.4 show the bathymetry beneath the profiles.
The energy from the explosions was recorded on a variety of
seismic systems:
1. USGS Permanent Hawaii Seismic Station Network (both profiles)
2. USGS Portable 5-Day Recorders (Mauna Loa profile only)
3. HIG Ocean Bottom
Seismographs (Kilauea profile only).
Although five sonobouys were also deployed along the Mauna Loa profile
in an attempt to better define the structure directly under the shots,
all five failed to operate properly and no data were obtained from
The data from each system are presented in record sections in
fixed-receiver format.
In this manner, the seismogram written from
each shot at a particular station is plotted in one record section.
The horizontal axis represents distance from the shot to the station.
The distance from shot to receiver was calculated using the
short-distance formula from Richter (1958).
The vertical axis is
time-reduced to 8 km/sec using the relationship
w
00
o
ID
/
IO /
/
1
/
/
/
o
/
/
/
/
/
r'
CO
CD 4->
r^ O
"H SZ
r. CO
o ""
a co
r,
CD
CO <H
CD O
D U
CO >H
rH O
"H
M
jc
C -P
CD CO
JO 4J
00
CD
rH x:
"h a
Cm CO
o t-
/
D. O
E
0 CO
"H -H
j_ CD
4J CO
CD ""
E
>>
o
I
CO
JC <D
4-> i-t
CO
CO
1 -H
U
CO
CD
U
3:
"H
X
lAl»'H±d3C]
W)
CQ C
CM
CL
-J
oo
c
o
co
CD -H
1/
o:
cr
"
o
4->
''
i
rA
21
n
o
x:
CO
o
O)
rH
"H
C«
O
td
CO
o
~3
CO
c
CO
CO
c
£**
CD
-H
"H
E
-_.
o
IT)
Cm
o
SCv
o
"H CO
S 00
-P CO
CO <D
CQ rH
O
"H
.=r
C\l
_
"H
o
ro
WX'HJLdBQ
U>
o
LAT]
ITUDE
LONGITUDE
18N
18N
18N
18N
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
154W
155W
155W
155W
155W
155W
155W
155W
155W
155W
155W
155W
155W
155W
155W
155W
155W
155W
18N
18N
18N
18N
18N
16N
18N
32.33
33.19
33.53
3".17
34.76
35.37
36.01
37.63
39.29
Ml. ol
42.51
18N 44.61
18N 46.13
18N 48.36
18N 49.98
18N 50.52
18N 51.02
18N 51.57
18N 52.67
18N 53.22
18N 54.08
18N 5^.69
18N 55.28
18N 55.89
18N 56.44
18N 57.56
18N 57.62
18N 58.19
18N 58.41
18N 58.99
18N 59.97
19N 0.58
19N 1.16
19N 1.80
19N 2.04
19N 2.63
19N 3.61
19N 4.91
19N 6.66
19N 8.52
19N 9.79
19N 11.07
19N 12.29
Table 2.3
36.09
36.23
37.36
37.84
38.33
38.81
39.33
40.83
42.42
43.93
45.43
47.43
48.91
51.11
52.73
53.31
53.84
54.38
55.52
56.08
56.65
57.26
57.87
58.48
59.04
59.69
0.18
0.81
1.31
1.85
2.49
3.05
3.60
4.18
4.71
5.28
5.94
7.17
8.62
10.41
11.51
12.60
13.69
- Locations of
OTIME(HMS)
19
9
9
9
9
9
9
9
11 29.27
15 40.88
20 39.90
26
26.30
30 44.36
35 42.35
41 30.74
56 28.70
10 11 24.86
10 26 28.20
10 41 27.50
11 1 26.81
11 16 24.24
11 36 23.97
11 51 25.20
11 55 38.52
12
12
12
12
12
12
12
12
12
12
12
0
41.38
6 28.04
15 38.74
21 30.18
25 41.80
30 41.76
35 36.89
41 25.43
45 42.07
50 40.51
56 28.80
13 0 42.27
13 5 40.29
13 10 39.84
13 16 30.86
13 20 39.19
13 25 40.84
13 30 41.69
13 36 23.19
13 40 40.00
13 45 41.37
13 56 26.44
14 11 31.71
14 26 28.53
14 36 29.03
14 46 29.23
14 56 24.71
22
large
small
small
large
small
small
large
large
large
large
large
large
large
large
large
small
small
large
smal 1
large
small
small
small
large
small
small
large
small
small
small
large
small
small
small
large
small
small
large
large
large
large
large
large
shots in the Kilauea Profile. Origin time
Large shot = 3001bs,
(OTIME) in hours, minutes, seconds.
small shot
SIZE
= 5-lbs.
23
SH. No
LATITUDE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Table 2 4
▼. « c
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
10.26
10.44
10.97
11.19
11.57
11.82
12.28
12.72
13.20
13.63
13.90
14.53
14.91
15.67
16.10
17.01
17.50
17.93
18.45
18.89
19.43
19.89
20.45
20.92
21.27
21.69
22.03
22.41
22.64
- Locations
LONGITUDE
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
156
155
51.60
48.96
48.04
46.25
44.45
40.97
39.23
37.57
35.88
34.26
32.60
30.86
29.21
27.59
25.87
24.20
21 .31
19.62
17.92
16.21
14.41
12.70
10.98
8.95
7.17
5.52
3.82
2.03
0.25
58.12
OTI ME
19
19
19
20
20
20
34 38.70
44
54
4
14
33
20 43
20 53
21 3
21.13
21
21
21
21
22
22
22
22
22
23
23
23
23
23
23
0
0
0
0
0
(HM S )
23
33
43
53
3
13
30
40
50
0
10
20
30
42
52
2
12
22
32
42
7.79
7. 35
8.21
2.01
10.96
19.35
24.06
17.00
1 5.04
19.82
23.30
19.05
17.36
19.84
19.39
19.84
21.97
18.98
20.65
18.47
19.34
20.83
20.27
16.15
20.56
17.90
17.60
33.73
16.34
of the shots in the Mauna Loa Profile,
(OTIME)
hours, minutes, seconds.
in
time
Origin
24
T
= t-x/8.0
where
= reduced time
t = total travel time
T
x = distance.
The horizontal axis is in kilometers and the vertical axis is in
seconds.
depth.
No corrections have been made for station elevation or water
The record sections are plotted with the amplitude of each
trace independantly normalized.
However, selected record sections
from the Mauna Loa profile are plotted in a true, relative amplitude
format.
In constructing these true-amplitude record sections, it is
assumed that shots of equal weight and fuse burn-time will release
equal amounts of seismic energy. For the smaller shots, the amplitude
of the seismogram can be approximately corrected for weight using the
empirical relation (John Orcutt, personal commun., 1979):
(charge
v0.65
weight
\
I
normalizing weighty
In addition, the amplitude of each seismogram is multiplied by the
distance (x) from the station to the shot and the amplitudes scaled by
a constant for plotting. In the rest of this chapter the record
sections from each seismic system are presented and described
separately.
25
USGS PERMANENT STATIONS
The USGS operates and maintains a network of some 40
seismographic stations on the island of Hawaii designed to study the
seismicity of the island and its relation to volcanic processes
(Koyanagi et al., 1978). Figures 2.1 and 2.2 and table 2.5 give the
locations of the stations that were used in this study.
The stations
all have vertical seismometers with 1 sec free period. The data are
telemetered from the field site to HVO on the rim of Kilauea caldera.
Here it is recorded both on 1-inch analog magnetic tape and a
Develocorder system.
The magnetic tapes containing the data were
taken to Menlo Park where they were played back, passed through a
high-cut aliasing filter, a low-cut filter to remove long-period
noise, and digitized at 100 samples/sec. Figure 2.5 shows the record
section written at station HLP which is used in the Kilauea analysis.
Figures
2.6 to 2.8 show the record sections written at stations CAC,
XII, and HSS which are used in the Mauna Loa profile analysis.
USGS TEMPORARY STATIONS
In addition to the permanent stations, seven 5-day seismic
., 1970)
recording units (Eaton et al
were deployed during the Mauna
Loa experiment. Each recorded a vertical and two horizontal
components with a 1-second free period.
Only the vertical component
is plotted in the record sections presented here.
The stations were
deployed in a line across the north flank of Mauna Loa, perpendicular
to the Kona coast (Fig. 2.2).
Table 2.6 is a listing of the
26
o
o
o
o
o
CD
iH
"H
Ih
O
/
£
o
o
Ou
CD
3
CO
iH
"H
o
CD
o
f-
E
O
£-,
Cm
o
UJ
z
<
6
V
O
-J
X
c
o
"h
4->
CO
CO
4J
CO
O
O
o
"H
4-5
o
co
/
/
T3
o
o
/
W\Y/
\aa vw
wJ
o
o
fyJ\f\fiiv\NNi
CM*
3
-H
a.
o
/
o
/
/
/
o
/
o
o
OD
D3S '09/X-l
o
o
o
27
ri
h'^V<;vV^
3
*-j^**V*!^
*
O
i
,
o
C
■
CO
rd
O
,
..
I
"■»
"
i
i
-_
s1
£
V.'''."\'-^>*r~»t\<'*\'^W^
c_
!
L
_:
]
.. . .
*.._.
. ..
is-- A^y'^'V^v^A^
,
<J3
O
-h
-P
to
'W v"***"^'.' *„ "'-""VM',< 'Vi, V V Yy\^l'^\^^Y^-^*rJ^J^M JlV^^
hr^-VS/V"*/ "-VM/V-' V^wi^/YnAiU^^*^ *wvi
V^yf
O
■
c
o
"H
4-5
o
CD
CO
■o
d
,
r
f *>>*AY.-'"V -H^/vV'-"A'*"A A A/,a^I*v*-~1*v*-~*''Y'Wv*'w^^
,
i
-■
-
J
[^-vV^^Vn/^^^wVH^/^/^^
T3
o
o
CD
PS
I
'8
/
\j-\Y*V~~~^Y^ \Ji
a
o
]yw^-^j^j\^^\Mjy/y^A^^\^
I
—
'
/S j/iAftA-~ws^^/\^>^^v^y\^^-ys/\/\/Vyv/
-ywvJV\
ro
Io
.o
C
C
6
cc
CZ
C 9 OS
Ct-
32S '8/X-I
CC
CI
0 I
CO
bO
28
CD
iH
"H
U-,
O
SL,
c_
CO
o
_")
CO
c
3
CO
z
<D
JZ
■P
B
O
U
29
a
S
\i
t\
"
'"~~w^
i
I ?_
Cm
v
cv
CO
o
o
a
co
c
3
'VvrVJV*^MA/VW^
rZ
CD
rZ
a
a
it,
Cm
CO
CO
joj o
c
o
"H
to
CO
4^
CO
CO
r"3
~
4-5
CO
c
o
"H
4J
o
yW\^v\^AAAA"^w^vv\,
A
USjv\/\/v*a/'«*v~vA/WA^^^
i
CD
f^V -^/v^V^^/v^^**^^
(X
>
v^A
O
1°
ra
CO
J
CM
!
CD
I
Ls
bO
i
Eh
"rs
i
/\^^^vv^^^-^^aA^VV^'V^^-^\
v
I
j
v^wv\A^
i
I
0
o*ll
o*ol
o*6
o*B
0
-
I=
!i d
/
o*9
33S '8/X-l
o*s
o*v
o*£
o*2
"H
30
STATION
AIN
CAC
DAN
DES
HLP
HSS
LUA
XII
Mo<
MTV
SWR
WHA
.
LATITUDE
LONGITUDE
19
19
19
19
19
19
19
19
19
19
19
19
155
155
155
155
155
27.60
55.00
40.00
23.30
155
29.10
4.20
45.90
35.90
3.70
36.30
2.90
22.50
29.29
21.42
20.20
17.96
36.31
24.55
30.56
29.28
30.25
27.26
19.90
155
155
155
155
155
155
ON ( m )
E L EVAT I
18.60
1524
323
30C3
815
707
2445
622
IS4I
4000
409
4048
29
-
Table 2.5
Locations of the permanent seismograph stations of the
USGS Hawaii Seismograph Station Network used in this study.
TATION
BSC
ISB
KUM
PUO
PUA
SMR
UST
Table 2.6
LATITUDE
19
19
19
19
19
19
19
35.47
35.91
39.85
30.45
30.33
33.07
31.95
LONG!ITUDE
155
155
155
155
155
155
19.85
24.47
13.45
42.36
50.76
32.83
35.74
ELEVATJIo\<(n
1696
21 19
851
2512
1183
3186
3494
-Locations of the temporary seismograph stations.
31
locations.
The field tapes were processed in the same manner as the
tape for the permanent stations in Menlo Park.
2.9
Record sections (Figs
to 2.14) were constructed from the digital data.
HIG OCEAN BOTTOM STATIONS
Offshore, the energy from the shots was recorded on University of
Hawaii HIG pop-up Ocean Bottom Seismographs (OBS) during the Kilauea
experiment. A description of the system can be found in Sutton et al.
(1977). Four OBS's were deployed; however, one did not pop up, and
one did not record.
in Figure 2.1.
The locations of the two that operated are shown
Figure 2.3 is a profile of the bathymetry under the
shot line also showing the location of the OBS's. The exact location
of the instrument on the ocean floor cannot be ascertained since the
OBS is deployed from the water surface and must fall several
kilometers to the ocean floor. Each OBS was located by looking for
the two shots with smallest water-wave travel times and then placing
the OBS midway between those two shots.
The depth of the instrument
could then be read off the bathymetric profile. The slant range from
shot to OBS was found through the traveltime of the direct water wave;
and horizontal distance was then determined by solving for the unknown
side of the triangle using the range and the depth for the known
sides.
In the two record sections (Figs. 2.15 and 2.16), the vertical
geophone channel is plotted in a fixed receiver format as before,
except the data are reduced by a velocity of
6.0 km/sec. Some traces
have been omitted due to unresolvable timing errors.
2.3 for recording geometry.
Refer to Figure
32
VVVf^^^7Vv^A^M^^
I °1;o
M
■
a
o
—
— —— -—— ~— — —
„
/W^r\y\jJ\M\J^\\^
/v^w\^/\A/Aj^/Vvy^
Ayv
—^
S/-"V">ww\/^^^
CD
<D
x:
4J
—^^_>^
—^
—v^-J
-v.
S-
E
O
o
U
Cm
.—.
?
>s^\/\fi<yjryJ\nM)fKf<*^^
CO
co
o
J
-J
CO
c
3
CO
CO
rZ
3£
"*■"
CD
i—l
"H
"H
Cm
<M
O
£-,
a,
0
O
CO
CO
/>v^^aAW%^
I
— —— —
wVA/\jWv,"V^
r*J\f\f\J\/\fo\/\^'\f^^
i
"
— — ~L^w^A/lrwv^'w'vAWwW
A
i/\A^^V^M\/\^^y,'vv^
"//
'-vv
CO
-P
■
£
1
-*
"■*■
—
CO
c
o
"H
4J
a
CD
<D
CO
CO
*-
-o
"a
r>
"
in
a
"
"
o
o
o
CD
PS
~_—
v
"
ri
CO
—
r^^-\i\P\i^\N^^
c
o
"H
4J
I
__~
—
■
cr>
o
CO
CD
/Uw^yv^
3
bO
"H
C3
in
/\AAAA-W\A/V
/VVfwvvA^/l/
o
a
in
1C
0
t
o*n
o*ol
o*6
o*B
Q'l
0"9
335 '8/X-l
o*s
O'V
0-£
Q'l
33
in
ie
o
s
iH
"H
<M
o
o
IT
r,
Or
C
3
4-5
.m
S
S-.
o
<M
CO
CO
a
in
C
O
s
"H
4J
v
o
2
cr
m
CO
—
4-5
CO
CO
o
in O
o
CO
C
O
4-5
O
CD
CO
o
T3
£,
O
O
CD
OS
o
I
■
-in
o
CM
CD
o
"in
r,
3
bO
Eh
'
"H
o
in
m
o
0*
on
oot
0
6
0i
OB
33S
09
'8/X-X
o*s
Of
oc
uo
o*
34
CD
.H
"H
Cm
;
O
Ou
CO
o
CO
J °.
Vf
£3
4->
s
o
v
A^^Vv^^/'V'v"VV^
~
/l
"**-
-*WvV-V^Ay^/A/WVwA^
"
□
o
3>
"y
i
C
O
"H
4->
——
hnfi*Aj\r^r«r<^^
CO
y
■
\WV^\W^v
a
Cm
i
:
I
■
4-5
CO
4-5
CO
c
o
"H
4-5
a
o
CD
co
CO
Vv^/^^vvyAmwWW^\^
1 i/i=
~^s^yf\/^^
s/
Jry^s
'
j^.^y^jK^js>r^^J^
/VvvAjA
AA
T3
M
O
o
O
CD
(X
Or
I
i
L5
fS
A
m
3
°*
i
rin
V\tya/\/Vw\^
'^^^
M
"
YVVVV -^Afu
v
Pi
j
o
in
o*ll
o*ol
C*6
o*B
33S
0"!
C*9
'8/X-I
C*S
O'V
o't
o*2
bO
hi
o
d
O
o
o
CD
rH
o
►j
I
i
co
c
h—^t^^^^wvw^^
CO
<yv/i\v>^^^j*As°**fi^^
CD
JY
o
Cm
j
<
o
I
c
o
;o
2_
UJ
4J
CO
i 82
I t—
I
°
!
"*
9
C
O
"H
4->
CD
CO
v
O
O
CD
.
A A
V^\AlM
-J
CM
I
CD
I
Eh
bC
-
rs
1
wvjvy\^^
o
a
d
C
t
o*ll
o*ol
o*6
o*B
33S
O'i
o*9
'8/X-I
o*s
0*»
o*£
o*e
"H
36
37
o
j-
rH
"^
c
o
ir
CL,
o
-1
o
CO
c
3
IP
CO
X
CD
SZ
4-5
o
E
O
o
CO
o
C
_1.
O
"rH
LJ
4-5
—
4^
CO
2
o
cr
f—
■p
4-5
CO
c
"H
4J
O
CD
CO
CNJ
(0
o
in
m
J~
o
on
oot
o'6
o*B
0V
09
'8/X-I
C'S
0*»
C C
07
S3
"H
38
c
"H
wvwwwwwwww^^
i/WVWWW^
VVWWWWIIW^
—
CD T3
iH CD
"H CO
O
<m 3
O
L,
i_
CO
o
o
wwvwvwvwwy^^
CO
CD
3
CO
in
/
bd CD
S
o
o
<r
CD
x: -H
4J
4J rH
CD
/
E >
O
CO
t*
£-,
Cm 4-5
o
R
H CD
—4
*w- » y vvvriVi »tt " ym<y v mi » V**>/vYIV♥vv y v v VAVv yy vVVl! AAV^i
w^jWi^^^
WwVJw/AV^
OS 4-5
CD
CO t.
o
o
e\j
mMW^H^
wmwn^t^
CQ CX
s
o
CO
LU
—
o
o
z
<
S <s>
£
5
w«*'
—
1
CO
*-.
rH .O
"H
/
VVWWWWIA^^
CD
C
.'jyrl*
c
O
"h
-M
o
CD
co
T3
o
i—^_
»
o
/
tCD
c
"H
0
s:
.P
CD
t,
CO
CO
O CD "
O C CO
CD
r,
-H -H
OS H CO
>>
I 4-5 rH
o
g
-i
i
o
o
I
V
<\J
I
A
1
wwwwwwvwwvvwvwvwwww
■
o
o
O
5
|
q
1
o
o
I
o
I
o
00
ID
D3S '09/X-l
I
o
CO
CM t. CD
4-5 SZ
CD CO 4-5
1
1
.c co
in w c
«— " vH CO
I
q
e\i
T5-'
O
o
£■
3
bO
39
c
"H
Q) *0
.H CD
"H OO
<H 3
O
t, 00
O. CD
JC
CO o
CD G
3 CO
CO SrH .O
"H
S
<D «H
JC -P
-P rH
CD
S >
O CO
«M -P
!_5 T3
O <v
IH -P
CD
tO.
t.
<D
4J 4->
CO C
"H
CO
CQ
O
C
O CD
"H .C
-P -P
o
<D CD
GO J-
CO
Sh
O
O
CD
X
00
CD
"
C 00
-H .H
rH >>
CO
I -P H
S2
" CO CO
CO "- CD
■P JC
0) CO -P
CO
vo bO C
t- -H
3
bO
"H
D3S '0 9/X-l
40
CHAPTER III
ANALYSIS
SEISMIC
OF
REFACTION
DATA
41
The interpretation of onshore/offshore seismic profiles presents
a challenge for the interpreter, because
of strong lateral variations
in structure.
The planar layer assumptions of the classical method of
interpretation break down.
can all be expected to be
island.
Dipping, discontinuous, and curved layers
present
in the real structure of the
For example, a major discontinuous dipping structure is the
water layer. It thins irregularly towards the land and pinches out
entirely on land.
Fortunately, the bathymetry is well known along the
shot profiles and can be included in the interpretation without making
misleading water-depth corrections.
Another strong lateral change is
produced by the some four kilometers of topographic relief on Hawaii.
This too can be included in the interpretation without having to make
elevation corrections.
Furthermore, the depression of the Moho
beneath the island is another possibly strong lateral variation.
Although the Mauna Loa and Kilauea data sets are similar, they
were analyzed in a different manner due to differences in the
recording geometry.
The presence of the OBS's in the Kilauea profile
allowed the construction of reversed profiles over the coastline and
submerged south flank of Kilauea.
On the Mauna Loa profile, because
of the failure of the sonobouys, the structure of the underwater part
of Mauna Loa's flank could not be determined with as much accuracy.
Nevertheless, the structure under the subaerial part of the island can
be more precisely defined since the array of temporary recorders and
permanent stations that extended across the north flank of Mauna Loa
provided good coverage.
In this case, analysis proceeded by
42
developing a starting model based
on the results from the Kilauea
profile. Rays were traced through
this model and traveltimes
computed. The model was changed
successively until the computed
traveltimes agreed with the observed traveltimes.
At this point it should be emphasized that the record sections
presented in Chapter II are not plotted in the
normal manner.
The
record sections were constructed by appealing to traveltime
reciprocity between shot points and receivers and plotting the data as
though the seismographs were shot points.
KILAUEA PROFILE
Method of Analysis
The configuration of this experiment permits the construction of
a set of three reversed and overlapping seismic record sections.
The
station pairs used were (fig. 2.3) HLP-RTI, RTI-TOK, and HLP-TOK.
The
record sections along with the traveltime picks for HLP, RTI, and TOK
are shown in figures 2.5, 2.15, and 2.16, respectively.
The
traveltime curves are shown in a composite section in figure 3.1.
Neither elevation nor water-depth corrections have been applied to the
data.
Rather, the bathymetry was approximated beneath the shot
profile landward of station TOK by a slope of
5.5° (fig.
2.3) which
includes the water as a layer of known configuration and velocity in
the calculations for structure.
The calculations were accomplised using the slope-intercept
method to invert for plane, dipping layers assuming a constant
43
velocity for each seismograph pair. The resulting velocity structures
were overlayed and ray tracing was used to check the results.
error in apparent velocities is estimated to be + 0.1 km/sec.
The
The
extreme range in apparent velocities shown in figure 3.1 is due
largely to the steep submarine slope of the volcano.
Seaward of station TOK, the data are unreversed.
Accordingly, in
the absence of additional constraints, the simple assumption was made
that the structure of the oceanic crust seaward of the volcanic pile
can be approximated by horizontal layers.
Velocity Structure:
Southeast Flank
In figure 3-2, the calculated velocity structure for the south
flank of Kilauea, perpendicular to the coast, is plotted together with
the structure determined earlier by Hill (1969) from a refraction line
along the southeast coast of Hawaii.
The figure shows two possible
interpretations of the data for the structure of the lower part of the
crust.
The 3.1 to 3.2 km/sec layer extends from the flank of the
volcanic pile out onto the seafloor.
Landward of the break in slope
between the volcanic pile and the seafloor where station TOK is
located, the
3.1 to 3.2 km/sec
layer probably represents material from
the most recent lavas erupted from the volcano botji as lava flows and
debris from downslope movement.
Seaward of TOK, this is probably not
the case, since it is likely that the lavas pinch out at the break in
slope.
At this point, the thickness of the layer determined between
44
» /
//
'.'';/// .""
»
"
I
/
"
/ "
'I'/ // "
'I»i // :
i//
'
■
"
"
.
"
T
"
""
"
'
X
' //
"
"
/ ;
"
<r>i/
tOlf> /
"'
l
I/
"
"
"
_c
O
CM
"/ -"
'I
I
"
*
ih :
i \
i \
I:
■
»/i/ I J
■
__\
o
.
"
»-
\
\\
I
I
\
4,
t
"
>
r,
CO O
H
3 co
O
B
t, 4^>
A
I
'1
rH
CD
h\ co
"
\\
11
CD
I
r
_n
*i
■
UJ*
co;
CT>
4-5 -H CO bO
"H H CD C
CO
X3-H
O T3 O 4-5
Q. -H C SE -H CO CO
vW
A.v
."
-
5 o S *>
XI n
\/
\
■'
in.
Q.
_J
X
<"f
4-5 <+-! 4-5
-P
t bOO
O C Q
<m -H "
+5
CD
bd
CD CO Cm
CC
r>
4
s
0)0X3
m cd
x:
..
fe
\
\
\
CD
p, co C
CO X 3
CO Q
CD "
CO
3
CO M CD
rH H C
"H OS -H
r-i
US
E X: Cm
"H O
4-5 C bO Ou
H ni CJ
CD S- -H X
>0 +»
S- E
CO
£_
CO O
4J CO 4J S-
\i
"\
-I
C
CO o
0 <D £-
\
\-
CD
CD -H CO
rH <H CD
"h
x:
Cm T3 O
O <D C
v sz co
■
M "'■"
/ wf
/ f
co
C\J
03S 0'9/V-l
45
X
o
CO
CD
3
4-5
O
3
i
i
i
i
I
Is
'
o
(cm
I°
ii
11
[
I j
i
[2
_c
i |
i i
o
ir
<
_c
j
CO
/
i
Q
l
i
i
\-
ii
_J
5<
X
I
I
I
/
I
v
UJ
CO
i
Ieo 1 Is
i
i
i
i
i
\
»
I
I CO
a.
_i
x
-I
+
WX
'
OO
t
i
iri >
CO
a>
U
"ao
_l
_J
hi
N
X
Hld3Q
00
co
N)i
t
_
CO
(D
i
ro
_i
CVJ
CO *H
4-5
O
0 SZ C
U 4-5 "—'
Q.
3
t, O 0
0 CO r,
4J
3
C 0 4-5
"h x: o
-P 3
so
3 Ci_, 4-5
H O CO
i
X
o
z
Cm
rH
O Cm CO
O O
CO
-H
C
4J
O C S"H CO CD
+5 H
CM
i
I
cr
JL
i
i oi
i
i
i
CO
.
I
t
i
CM
Hld3o
4-5
CD CO -^
4-5 -H C
o
cd
H (B -H
3
r>
O
CO
_.
rH
CO CO CD
O CD tO
3 bO
CD CO CO
SI iH X
4-5 -H CD
US
>
lII]
/ « Icm' io- [
|
ro
CM
i
CO I
(^
o
I
I
II .1
<
i i
i
|
I
! I §o
!ill
'
111
' ' is
II'
''
o
i
J
/
i
I -I
'l
II
CM
CM
\
I
cv
0_
1
x
1o
/
f
i
T
I
I
i
v
Hr i
m
* o-
i
T3
f cvj I0> j
a,
I
I
i
/
> x:
; !
I
ro I co
"J
>>
If
i
i
S- T3
4-5 CD
CO S-.
S-.
CD
Cm
4-5
"H CD
O M
O O.
rH
CD CD
0
"H
46
stations RTI and TOK is projected seaward.
Therefore, tight control
of this layer is lost seaward of station TOK.
The 3.1 km/sec velocity
is intermediate to what is normally measured for the velocities of the
"sediment" and "volcanic" layers in the Pacific (Shor et al., 1970).
The 5.1 to 4.6 km/sec layer pinches out seaward in the vicinity
of station RTI. Rays traced along the top of this layer in the updip
direction are in good agreement with the observed data (the 3.5 km/sec
branch extending seaward from HLP in figure 3.1).
Rays traced downdip
are not in good agreement with the observed data (the
branch extending landward from RTI in figure 3.1).
6.9 km/sec
Although the
correct velocity is obtained, the intercept time of the calculated
branch is about 0.8 to 1.0 sec greater than the observed intercept
time when the structure in figure 3.2ais used.
discrepancy.
This is a large
The dip on the refractor, however, is measured only over
a short distance between stations HLP and RTI. Ray tracing suggests
that there is little control on the refractor underneath RTI except
that the 5.1 to 4.6 km/sec layer must not extend far into the section
formed by RTI and TOK.
The discrepancy is reduced to 0.3 sec by
flattening the dip on the refractor as shown in figure
3.2b. The
details of this part of the structure remain poorly resolved.
The crust is about 11 km thick near the volcanic-pile/ocean- floor
contact and about 13 km thick under the southeast coast (Hill, 1969).
This gives a dip of 20 to 3° of the Moho toward the island.
Pn
TOK) is 7.9
measured along the unreversed profile (seaward from
km/sec. Hill measured 8.2 km/sec on the line which ran parallel to
47
the southeast coast.
This discrepancy could be an indication of
upper-mantle velocity anisotropy as
found in the Pacific by
al. (1970).
Shor et
In that study, velocities measured parallel to fracture
zones were found to be about 8.2 km/sec, whereas those measured
parallel to magnetic stripes were about
7.9 km/sec.
The 7.9 km/sec
velocity is consistent with the findings of Shor et al., given that
the Molokai Fracture Zone (figure 3-3) intersects the Hawaiian Ridge
near Molokai roughly parallel to the southeast coast of Hawaii.
Other
workers (Eaton, 1962; Hill, 1969; Ryall and Bennett, 1968), however,
have found Pn around the island to be between 8.0 to
8.3 km/sec.
There are several alternate explanations for the apparent 7.9 km/sec
value that was measured. For example, there may be a higher velocity
upper mantle dipping slightly (about 2.5°) away from the island, or
there may be progressively weaker Pn arrivals such that the
first-arrival picks are in error.
Another possibility is that the low
Pn velocity is somehow associated with the presence of the Loihi
Seamount, located about 20 km southwest of the profile. Recent data
have shown Loihi to be a young volcano forming on the south flank of
Hawaii island (Moore et al., 1979).
It may be that local heat
concentrations in the upper mantle produced by volcanic processes have
Modeling of rays from the Moho incident at station HLP produce a
mismatch which is best corrected by extending the velocities under
station RTI to under station HLP.
This puts 4.6 km/sec where Hill
(1969) has 5.1 km/sec and puts 6.2 km/sec where Hill has 7.1 km/sec.
Ul
o
4
\1
i
vi
<
<
p
<
■
il
3
X
r
u-
i"
—>
Q
o
c
o
"H
4-5
CO
4-5
c
0
"H
L,
O
T
o
0
x:
4-5
bO
c
"H
»O
x:
co
oo
o
.
"o
c 0
co c
r-t o
CO
N
M
0
C U
CO 3
"H 4J
"H o
"
CO CO
3 U
CO <M
X
"H
0 CO
sz _*
4-5 O
rH
O
o£
co
x:
2: 4J
2
0
£.
3
m
UJ
o
"H
_U
"
X
49
Similarly, to model the 5.0 km/sec branch extending seaward from HLP,
the 5.1 km/sec velocity of Hill has to be replaced with the 4.6 km/sec
velocity from this study.
These differences are difficult to
reconcile in detail with the available data, but they serve to
emphasize the level of resolution attained by linear profiles over a
strongly varying three-dimensional structure.
The velocity structures plotted in figure 3-2, a and b, show two
interpretations that can be made for the
6.9 to 6.2 to 7.1 km/sec
layer, depending on how a set of second arrivals (the
9.4 km/sec
branch extending landward from TOK) is interpreted. If these arrivals
are from the Moho, then this suggests a dip of 8o on the Moho toward
the island (figure 3-2).
unreasonably steep dip.
I
feel that this interpretation gives an
If, however, the 9.4 km/sec branch is
interpreted as a phase which bottomed in layer three, the layer can be
divided into a top layer with a velocity of 6.2 km/sec and a bottom
layer with a velocity of 7.3 km/sec (figure 3.2b).
With this
interpretation, the lower crust has a more uniform velocity, namely,
7.1 to 7.3 to
6.9 km/sec, and the
anomalously low value of
6.2 km/sec
is confined to the upper part of layer three between stations TOK and
TTT T**)
riLr
.
Velocity Structure:
Rift Zones and Summits
Several researchers have noted markedly higher P-wave velocities
under the rift zones and summit areas of the volcanoes than under the
surrounding shield areas.
Crosson and Koyanagi (1979) and Ellsworth
50
(1977) have observed strong differences in
vertical velocity between
the two areas.
Hill (1969) found anomalously high P-wave velocities
in the shallow crust near rift zones and concluded that they are cored
with high velocity material.
The structure of the rift zone and summit areas was investigated
by modeling the traveltimes from two profiles formed by a particular
shot and several stations. One profile is oriented north-south and
intersects Kilauea's east rift at a right angle. The other profile
runs northwest-southeast, crossing Kilauea' s southwest rift at a right
angle and extending into Mauna Loa's summit area (figure 2.1).
The middle parts of figures
3-4 and 3.5 show the refraction data
(first arrivals) across the two lines.
Also shown on figure 3.5 are
data from a shot on Kahoolawe Island, published by Hill (1969), that
partially reverse this line.
The arrival at the receiver nearest the
source in both figures is late with respect to other stations.
The
average J-B residuals plotted from Ellsworth (1977) show the arrivals
at stations in the rift and summit areas to be early with respect to
the stations in adjacent shield areas.
These observations were modeled for the structure and velocity of
the rift/summit areas.
To do this, a generalized representation of
crustal structure was chosen, namely, the basic structure derived in
the first part of this paper projected under the island.
This
assumption gave a dip on the M discontinuity of 2o with an upper
mantle velocity of 7-9 km/sec.
The lower crust was assigned a
velocity of 7.1 km/sec with constant thickness under the island.
The
51
D
O
CO
c
/";■":
■.V " 4/ rV
O TJ
"H 0
4J >
CO
SJ_
0
0 CO
bO X 3
bO O
.
::/
1/
S
-IT)
i
;:"
">
4-/
■mm&m.
L
m
</
■
<3
X
*
CO ""
X
0 CO
0
rH rH
CO o "
O M
/
.
"
-a,\V.,,v
:""""■"
■JV^v^r
"""": j;>^r--.v.v ■
"yy^
:
"H -H rH
4-5 O 0
JT3
(DO O
0 B
CO
" .^1
""■"■""■.J
KM4my
>m
>
O E
O
C rH O
":
<
X
_!
>M[','"■■
£
r- ■■■■■■.■
1//.
O
<
■ ■
.
"s
I
'■ '.
.'..
oiiyijjjjiyii
:■■-1 X »■:::::::::::
*yj:0:::::::::::::!
O
3 3
t, CO Ou
Q. rH E
"H O
0 UJ O
S-.
3 <m CO
4-5 O 40
CO
O
3 4-5 T3
..
r, CM
4J -H
CO t, CO
0
X) 4-> CO
C CO CO
cfl cO O
t,
0 o
o
"H CO
E co "
CO O Cfl
"H
4->
£<
0 O CO
CO CO "3
■
z
;ir>
o
4
<
cr
fw
1
I
08/X-l
r,
.:.:
3
M
Cm
VM
cr
tf>
■5
0
UJ
X
>r
>■":
U-
UJ
>
"""■1[2::::::::::::::
M>
■.■""■
■"
:MJ
o
<
cc
UJ
D3s
CM
"''
CO
— — or~— \
o
o
UJ
z
m
.'■■Y
: UJ ';*'::.C:*:'.*::.:
:i':v.*:v:.:
:
:-V:V
: .**
::*:
yVyy CO .V':'.-V:':::'."■";:'v*-v-V \ ".":-.':**:'-:;*.:.*-:'
O
CE
O
4MI
4:1
I Mm
_j
x
'(
>
*
>*
O SCm
0 "
.-I
T3
"H (0 <D
Cm 0 -m
■f.-::!
Y.S'Y.YJYY7
V
■".*;
in
O
'
WX Hld3Q
m
52
upper crustal velocity in the areas between rift zones was chosen to
be 5.1 km/sec.
The upper crustal 3.1 km/sec layer was ignored.
The
rift zones were assumed to be laterally discontinuous structures of
high velocity confined to the upper crust above the
7.1 km/sec
layer.
I
assumed that the first-arrival data consisted solely of arrivals
from the M discontinuity.
This assumption is justified because the Pn
crossover is expected to be between 40 to 50 km for this model,
roughly corresponding to the smallest source-receiver distance in both
profiles.
The velocity and boundaries of the rift zones were allowed
to vary along ray paths corresponding to the shots and teleseisms
until a fit of the traveltimes was obtained.
The final models of the structure of the rifts and the ray paths
are shown schematically in figures
3.4 and 3.5. Where a
ray enters
the rift, some increase in velocity along the path is required. The
Pn ray travels through the model at a much lower angle than the near
vertical teleseismic ray.
A change in the model will affect these ray
paths in different ways, providing a tradeoff between the teleseismic
and Pn rays which helps constrain the model better than just the Pn or
teleseimic data alone.
The uncertainty in the velocities is estimated
at +0.1 km/sec, since a change of that order will not significantly
alter the agreement between the observed and computed data.
In the east-rift profile (figure 3.4), the rift zone must spread
out with depth to satisfy the data.
The velocity in the rift averages
7.0 km/sec. However, for station NAG, the data are difficult to
model.
The calculated Pn traveltime agrees with the observed by 0.1
3
O
CO
E::::::::::::::
t--M-/^
"y-4/.
:
""444-44.>
J- J:-.y4 4./
M-:M;4/
" / J
* -.*/: /
/"/
■■■/'.
/
!tyinM| t i'"
";.4-
"
r;Y-f/y.
M/M/
'."/:":■"".
■4/:-:;
'//""4- :::.::::::.:■::!:
<?^j
--.- /
■r...-,
mm
7
-4/y
r
'■
;
*
a.
_j
X
X
'"
":"":"/-.
rr.;--^:
"v
iisifi
-/mm
:
"
r<mm-'
»^S« ■^Y-YYI
S*M4S4- {. iiNiii
:::::::::::::
::.4''::J.:::r':"
'
C ■
O CO
"H O "O
4-5 J 0 CO
4^>
CO
t, CO S* CO
0 T3
0 C
M 3 CO
bO Cfl £-
' "/4;4;^h./
/"""4->.■.■."..■"./:-
**
X
ii^/44^ " "
iiSt!!!!l!ililii
M:4-./444
co
CO
y-
a
Z
<
X
>
:
..
*-v -i;
cfl S O co
CD
X
0 Cm "" CO
O CO CO
0 O
iH
Cfl 4-5 rH £-.
O -H O O
"H E
+0 E-H »
S- 3 C "-»
cr>
0 co
>
C MD
0
a\
0
ox:
an
4->
O m
*C'
4-i
/"::/M-4:1
r::
"..::""."■".■: ■-. -I "M:
UJ
Cl"-;:
VA
f
X
"A-:
■/..■"
f;/>.":..
M-J!
\-'*"■'"
11.
«V:44::.v.K.X:.V.--*--.*.-i
■*!
v "
f:'■""'
"j
r,
r■■
.■ .
'.■
—
y;-4
Mv-4
/l
nV
\y
V^m:.
/m/mm.4
::-.y Ji
vff.■"■■:£
.■:..:
.- .::■"■A:\ *\4Y*y-:Y
MM/
m
4M;;-4y$■■■/
."'."."."jr
".\W_
IM/ :4 :.\'C\Y
' V■■/
■/ :".■■' V
m
■
"
/.
:""*/
*.*/y/-v
';7
£/"'■■■v f::.-.:-:
::i:ir::iji
2
<
o
«
o
4:1::::::::::: CM
X
M
o
5
V.4,;4
M*
\.
V \. V
iM
1.*..*".:
"
'
_
44
!<T>
°
■
,v \
.'- <\
w
_-
1 M■J:::::::::::
:
MMM4\ MM-
-
<
(<
CO
UJ
cr
a
»
=oz
UJ
o
O
<
cr
■
I—
<
cr
v.
v.
UJ
O
z
UJ
cr
<
O
—
■■■:-■■:
. ...\\- ..
3
q
cr —r—
■
:.-. i
.""".".■:Vv*\-:
\i
"'":."::.
.-'.v.
V --.-,,.:,
\4":■*
".;■.." "".".■;. i
"""""■':."--.V
4'X"'.■."
j
"»%v *"":■::*:;■ vv\yy>:-.
:yM;--^.v:\
:^4..4,4l
_j
<
I-
■"
ro
03S*a8/X-l
iCM
\V
mm
wmm
mam
I
'"'.'">"
!
MM.J
#&J
m
■'
mi
ml
I;;;;
'^ m-'l
>m
3Y-M
iiih
*»4;\Xm
yM:
%3! I!
-YY-i:
44-VWJ
*««
i»iii
::
4
s«
Iil
l'Y.:-:\Y
mm
MMM.X
.- o
■
-J
_J
UJ
>
UJ
-I
<
o
::::::::::::
:::::::::::::!
-2
4->
O
SZ
CO
0 rH
S 0
Cfl TD
rH O
o
4-5 0 CO O
co 5 cd x:
x: iH cfl
"a 4J o us
■■'■'■
hm " " l-^OV:
K-y
.co
A A.'..-:;I ■■-:""»4
;s:-v:*y^;y
".H
Cfl rH
4-5 -H
Cfl X
TD s_^
O 3
r, CO T3
O. rH >
0
"H
0 UJ t,
M
0
3 <M CO
4J O X>
fj
O
3 4-5 ""
$-, co
:::::::::::
/V
*
*
0 4J
rH
-H Cfl
Cm 0
:::::::::::
.'.
■■:■■
o
r
i
E
s
o
v
<M
C 3 M
Cfl O -H S
CO O O 73
C (D
O
"H CO TJ
0
SW
CO O CO
"H t, O
0 O rH
$-4 4-5
3
Cfl D,
4-5 B
Cfl O
CO Cfl O "3 O
I
in
0
v
i
"H
54
sec, but the calculated J-B residual is late with respect to the
observed, suggesting that the crust should be thinner than shown under
this station. From Hill's (1969) work, the crust is transitional
under NAG, thinning to the southeast to about 12 km.
This discrepancy
is ignored since the Pn ray must travel behind the rift zone, where
there is little information on structure.
In the southwest rift profile (figure 3-5), as in the Kilauea
east-rift profile, the rift zones are required to spread out with
depth.
The data further suggest that the summit complex of Mauna Loa
and the southwest rift of Kilauea join together beneath station AIN.
The velocity in the rift is found to be
6.5 km/sec,
lower than the velocity in Kilauea 's east rift.
significantly
The 6.5 km/sec
velocity, however, does not satisfy the teleseismic data at stations
MOK and SCA.
The calculated residuals are 0.2 sec late, suggesting
that the summit area of Mauna Loa is underlain by a high velocity
core, perhaps as high as 7.1 km/sec.
There are a few difficulties with the interpretations shown in
figures
3.4 and 3.5. As discussed here, 7.9 km/sec is not
upper-mantle velocities measured around Hawaii.
typical of
Therefore, 8.1 km/sec
was tried as an upper-mantle velocity using the same layered
structure.
The same rift-zone velocities are obtained, but the rift
boundaries do not spread out quite so much.
significantly change the model, however.
The result does not
An alternative to allowing
the rift-zone boundaries to vary is to allow the dip on the M
discontinuity to increase to accommodate the faster upper-mantle
55
velocity.
Simple calculations suggest that the dip of the M
discontinuity would increase from 20 to about 3.2°, which would
not significantly affect the interpretation.
MAUNA LOA PROFILE
Method of Analysis
The lack of data over the ocean part of the profile ruled out the
use of the classical slope-intercept method of refraction profile
interpretation.
The assumption could not be made that a layer would
maintain its attitude, thickness, and velocity across the profile. In
this case, a computer ray- tracing routine was used to analyze the data
by forward modeling.
However, for the shots and receivers closest to the Kona coast, a
limited amount of reversed traveltime data was obtained.
These data
could not be interpreted in the normal manner since the shot and
receiver array do not overlap, as is required in the classical
slope-intercept analysis. Therefore, a modified version of the
classical method was used. Using the reciprocity, the apparent
velocities across the shot and receiver arrays, for distances less
than 40 km, could be computed and compiled in a semireversed
traveltime curve (figure 3.6).
An apparent velocity of 5.6 km/sec
(long line in figure 3.6) was observed across the station array with
an intercept time of 2.8 sec.
The corresponding apparent velocity
across the shot array averages about
3.6).
3.6 km/sec
(short lines in figure
These parameters can then be used in the slope-intercept
56
o
CO
4J
CO
0
rH
rH
Cfl
1
0
sz
4-5
J-
o
Cm
0
>
L
3
O
0
S "
-P CO
o
"H
rH 0
0 O
Z
Cfl
E
111
o
z
<
I-
5
-
0 v
CO 0
t, >
0 -H
>
0
0
1 I"H I
S -P
0 o
co x:
w
►
i
m
0
M
"H
o
"*
CM
(oes) o'B/X-l
O
a
PS 0
o
eg
C
Cfl
U .p
4J CO
"H
T3
57
formulas to invert for structure with the assumption that the
refractor is continuous across the profile. Figure 3-7 shows the
calculated structure for the upper crust beneath the Kona coast.
The
value of 3-2 km/sec was assumed for the upper layer.
The results shown in figure 3.7 and the general velocity
structure determined for Kilauea were combined to form a starting
model for the ray- tracing.
A 2-dimensional, ray- tracing computer
program written by I. Psencik based on theory presented by Cerveny et
al. (1977) was used in the traveltime modeling.
The method used to specify the velocity structure in the computer
program is the following.
Figure
3*B shows a
sample velocity
structure composed of two layers separated by a boundary.
The
boundary is defined by a series of x,z pairs. Boundaries must always
begin at
x=o
and end at x(maximum).
Straight lines or cubic splines
are fit between the points to form the boundary. Boundaries may have
an arbitrary shape under the constraints that they must not double
back on themselves and they must not cross other boundaries.
The velocity is defined immediately above and below the
intersection of the boundaries with vertical grid lines (figure 3-8).
It is also specified at the points where the grid lines intersect the
top and the bottom of the model.
In this way, a box is defined by the
intersection of two boundaries and two grid lines.
A two-dimensional
linear interpolation is performed to define the velocity at any x,z
The source can be defined at any x,z position.
An initial
58
"
CO
CD
3
W)
"H
<M
c
"H
CO
CO
-P
-o
CD
x:
■P
§
v
-P
CO
CO
O
o
CO
c
o
U
CD
-o
c
3
CD
S-.
3
-P
O
3
"P
co
i
E—"
CO
CD
3
ttf)
"H
CX4
(Ui>|) Hld3Q
——
o
CO"
-
IO
a>
id
>
DC
<
O
z
O
€0
o
CD
O
z
z
0
x:
-p
U-
<m
liJ
O
o
5
i
CL
o
C
O
co
LU
z
Z
"H «
4-5 E
CO cfl
_J
M
3 &
Q
CC
O
CD
M
4J bO
CO o
rH a
rH
"H bO
c
0 -H
o
x:
4-5 Cfl
r,
U 40
O I
Cm >,
Cfl
0 v
v
3 0
4J x:
O 40
3
U O
E
o
LU
o
z
<
CM
*
w
— _______ . .
-*
'
-
<r-
<M
CD
40 40
CO
co
>» 0
v
40
"H
O
O
rH
0
>
-P
CD
E
Cfl
U
CO
0
■H 4->
CX 3
E a
co c
CO -H
CO
I
CO
m
>OC
<
O
0
r,
3
bO
"H
IX4
Q
CM
z
D
O
CD
o
IN*
CO
▼"
o
a>
o
(ujm) Hld3Q
o
CM
CD
60
take-off angle is chosen along with an angle increment and the number
of rays. Trial-and-error is used to determine which rays return to
the surface.
Figure 3-9 shows the rays and traveltimes for a
hypothetical source located at (10.0,0.0).
Propagation of the rays is based strictly on ray theory. Head
waves cannot be modeled.
For a down-traveling ray to return to the
surface it must either reflect off a boundary or turn in a velocity
gradient. This is a minor constraint since a very small gradient can
be used to model the head wave traveltime rather well.
Furthermore,
work with synthetic seismograms has shown that head waves have very
small amplitudes compared with reflected and turned phases.
It is
likely that in many cases, traveltime branches interpreted as head
waves are, in fact, turning rays (Cerveny, 1966; Braile and Smith,
1975).
Since head waves cannot be modeled, in the following discussion
the term Pn is used loosely to refer to a group of rays which have
turned in the region of the upper mantle directly beneath the Moho.
The same ray-tracing program, with some modification (McMechan
and Mooney, 1980), can be used to calculate ray- theoretical synthetic
seismograms.
The method is based on the idea that the spreading any
ray tube (i.e., group of rays with equal increment angles) experiences
is directly proportional to the amplitude for the rays.
If reflection
and transmission coefficients are also calculated, an amplitude can be
obtained for the tube.
61
o"
o
oo
0
M
3
bO
"H
<M
00
c
"H
c
O
x:
CO
o"
0
£h
o
3
40
O
CD
3
4-5
CO
0
JS
4-5
o
Cm
o"
E
cfl
o
r,
"3-
"H
CO
OS
I
<J\
en
o■
0
_.
5
o
"H
C\J
CD
"
lo
(D
■
"*-
O
o
ro
CM
"
0-9/13Q-1
O
■
o
o
o
o
o
o
o
62
The approach used to interpret the data was a forward modeling
process in which rays were traced through the starting model and
theoretical traveltimes were calculated and compared with the observed
traveltimes.
The model was then changed and the process began again.
As many as 50 models were computed for each station before agreement
between the observed and computed traveltimes was obtained.
Synthetic
seismograms were also calculated for station PWA (figures 2.2 and
2.12) and compared with the true relative amplitude section.
The
following section describes the results of the calculations.
Velocity
Structure
The calculated velocity structure for the Mauna Loa profile is
similar to that found for the Kilauea profile: an ocean crust which is
thickened and depressed under the island.
This basic structure, with
minor differences discussed below, explains the traveltimes at all the
stations to within the experimental error of 0.1 sec (see Chapter
II). Figures 3.10 through
3-18 show
ray diagrams, calculated
traveltimes, and selected observed traveltimes for all the stations.
It should be noted that in figures concerning the Mauna Loa profile,
the depth axis is somewhat misleading because of a restriction in the
ray tracing program.
at a depth of 4 km.
Sea level, which is normally at zero depth, is
Zero depth was chosen to be at the same level as
The velocity model determined from the data recorded at station
PWA, which is located about 20 km landward from the Kona coast, is
63
"
o
o
CNJ
O
CO
CO
C
O
o
■
o
T3
<D
>
U
0
CO
X)
"
"H O CO
0
4-5
LD
Cfl
«"
S
40 CO -H
CO 0 4-5
rH rH
0 O CD
rH r,
"H -H Cfl
Cm O m
40
O
>
0.
C T3
0
M
Q,
c;
■__!
o"
o
o
_
<D
cfl O
O
cfl
s*SM
%a__i
'rill
"«*&■
*Ji
a
Cfl
C
3
CO
2:
0
SZ
O
rH
"" rH
CM CO
""
c
O
o
""
CO
-H CD
40 4-5 CO
Cfl W
■»■'
_. m
O
O 0
Cm bO
bO
E CO
X
CO
S- CD
bO
Cfl rH
"H CO
-O O
-H
40
rH
_,
CO
"H
Cfl
r,
O
"
CO
CD
E
£
t;
> -P
OS CD
I
o
rH
en
O
"
0
r,
3
bO
o
LO
"H
O
CO
O
O
o
o
o
ro
(TB/13CI-1
o
o
o
o
o
o
■
o
C\J
o
o
ro
o
o
C\J
O CD
<c >
o"
o
O
C
O
LO
U
0
00
XI
"
"H O CO
4-5 "" 0
Cfl
S
40 CO -H
CO 0 4-5
rH rH
_
000
>
rH L,
"H -H Cfl
<m O
40
O
£-. C
Q. 0 -O
C 0
cfl O -P
O
cfl
" rH
3
Cfl rH
0
C "" rH
3 CM cfl
O
cfl
_!
o
o
o
C
""
0 O CO
x: -h 0
40 4-5 CO
Cfl
CO
t* o
v 0
o
v
<m bO o
.
bO
B CO
CO
X CO
SL.
bp
0 0
a
cfl rH >H
"H cfl 4J
O OH
0
>> "H
-P >
CO £-. cfl
<D
OS
_
> 4->
rH
rH
o"
o
CO
0
b
3
LO
W)
"H
CZ4
o
<
o
D
o
o
LO
"3-
o*B/130-1
o
ro
O
o "
o
Csl
o
o
o
■
o
o
o
C\J
o
■
o
ro
65
"
O
o
-O
o
CO
CO
X
C
"
o
o
T3
0
>
t-
0
CO
"
O CO
40
0
cfl
E
LO
LO
O
"H
X)
..
-P OT >H
CO 0 40
rH rH
0 O 0
rH t"H >H Cfl
Cm 0
O
4-5
_
>
0
Q
S
«
ef
I
m
Cfl
O
►J
Cfl
C
3
CO
-O
O
X
O
o
O
o
)b
C
Q. 0 T3
a0
O 4-5
Cfl
" rH
3
rH O
rH
CM cfl
"" 0
"♥
c ""
0 O CO
CD
40 40 CO
cfl CO
x: -H
r, U
O 0
Cm bO
bO
E CO
CO X
S-. 0
bO
CO rH
"H Cfl
ox
x
x
—
O
U
"
CO
0
E
"H
4-5
"3 0 rH
"H CD
40
CO U cfl
OS 0 S4J
I
|(
>>
>
>
CM
rH
"o
o
m
o
0
s-
IO
LO
3
bD
"H
00
00
"
o
■
o"
CD
LO
0-8/13Q-1
o
o
o ■
o
ro
o
o
o
o
o
o
o
o
ro
66
"
B CO
CO
X CO
S-, O
O
bO
Cfl rH
"H Cfl
T3 O
"H
B
-H
-P
-H
<D
>>4J >
cfl s- cfl
o "_,
« > -p
i
CD
O'B/130-l
o
o
CNJ
"
o
o
-o
M CD
M
US S0
C CO
>
O
■
o
O X 5"
LO
"H O CO
4-5
0
CO
".
B
4-5 CO -H
CO 0 4-5
rH rH
0 0 0
rH U
"H »H CO
1
>
<m O U
40
O
r,
C
a (D-d
<d
a 40
co
o
o
co
rZ-
..
" rH
3
CO rH 0
C
rH
3 CM CO
cfl
0
c ""
0 O CO
X
CD
o"
O
O
H
4-5 4J 00
Cfl to
O
O 0 v
bO
E cfl
CO X co
U 0 0
bO
S
Cfl rH "H
"H Cfl 4-5
"O 0 rH
"h a>
$-, t,
<m bO O
.
>> 40 >
_, cfl
co CD
t.
OS >
40
I
rH
o
o
CO
0
v
LO
3
_D
"H
E_>
"— t
I— l
O
O
LO
o*B/130-1
o
ro
o
o "
o
eg
o
o
o
■
o
o
o
OJ
o
■
o
ro
"
o
o
a
o
_3 >
_o
O
0
C CO
O
o
LOo
LO
OX 3
CO
40 CO
co 0
rH
0 0
rH _.
"H «H
<m 0
0
O
£- C
v^
f
O
rl "
CO rH
C ""
3 CM
Cfl
X "
c
0 O
1
_o
cfl
3
iH
0
-H
CO
0
""
CO
x:-h 0
-P 40 CO
CO
_, Cfl
M o
O <D Jm
o
o
X
X
Cfl
S-,
4-5
CO O 40
44!
0
0
>
Q. 0
'4
jy
E
-H
-p
tH
-a
Q, CD
o
0"
o a
0
"
"H O 00
4-> "" 0
(h
bOo
bO
E cfl
cfl X CO
0 0
bO
E
CO rH -H
"H CO -P
O
)$
.
r,
X
X
T3
"H
rH
0
>>40 >
Cfl La Cfl
os a> t,
c
> -p
i
in
rH
O
"o
"
on
0
o
LO
"H
o
ID
Q_
o
o
o
■
o"
CD
LO
tT*
ro
o*B/130-1
o
o
o
o
o
o
o
o
CM
o
*
o
ro
"
o
o
C\J
< ■3
0
>
rt
PU
o"
o
M
0
"
CO
C X)
O
"H O CO
4-5 "" 0
LO
cfl
B
4-5 CO -H
CO 0 40
rH rH
0 0 0
(Hr,H
>
r,
"H H (D
<m 0 t,
40
O
m C
a 0 -3
a cd
Cfl
o
►J
o 40
cfl
" rH
3
""
""
Cfl rH 0
C
rH
3 CM Cfl
0
Cfl
X
O
*
""
c
0 O CO
X2
-H CD
-P 40 10
cfl CO
M S-, O
O 0 J-.
<m bO O
o
o
bO
E cfl
cfl X
£-0
bO
Cfl iH
"H Cfl
"
CO
0
S
-H
40
"3 0 rH
"H CD
>» 4-5
Cfl t, Cfl
OS CD
40
I
>
>
-
vX>
rH
co
o"
o
0
t,
3
bO
LO
"H
<
a_
■
o"
CD
LO
o
o*B/130-1
o
o
■
ro
o
■
C\J
o
o
o
o
o
o
o
o
c\i
o
o
ro
70
"
o
o
"3
<D
PS >
s
co
o"
o
C
t,
0
co
ox»
"
"H O CO
40 "" 0
LO
cfl
E
rH £-,
>
4-5 CO -H
CO 0 -P
rH rH
0 0 0
"H «H Cfl
Cm 0 U,
40
O
S- C
3. CD T3
O. 0
CO O 40
0
0
0
O
rZ
CO
" rH
3
CO rH 0
rH
C
3 CM CO
O
cfl
."
X "" ""
c
0 O CO
o
o
o
x: -h 0
40 40 co
Cfl CO
£- m O
O CD S-,
Cm bO O
bO
E cfl "
CO X co
J- 0 CD
bO
E
>>
>
CO rH -H
"H CO 40
■O OH
"H 0
40
Cfl U Cfl
OS <D U
40
I
>
rH
en
0
o
"
o
v
3
bO
"H
LO
o
■
o
"
CD
IO
o*B/130-1
o
o
o
o
ro
o
o
o
o
o
o
C\J
o
o
ro
71
o
o
CM
E-«
CO
2E
C
O
O
"
"3
0
>
_.
0
CO
X> "
"H O X
40
0
CO
E
o
LO
..
4-5 CO -H
00 0 40
rH t-\
0 0 0
.H t.
H H Cd
<M O M
40
O
>
a, C
0 -3
r,
cd
a. 40
co
rH
cfl o
o
J "
Cfl rH
C ""
3 CM
Cfl
o
o
o
3E
-
3
0
rH
CO
0
""
0 O CO
X-rl <D
4-5 4O CO
Cfl CO
M r, O
O 0 £-.
Cm bO O
bO
B cfl "
cfl X CO
r,
c
0 0
E
bO
CO rH -H
"H CO 40
"O OH
"H 0
>*4-5 >
CO U Cfl
OS CD
> 40
I
OO
rH
o"
en
0
O
LO
r,
3
bO
"H
Cju
o
o
o
r^
CD
LO
0-8/130-1
o
o
o
o
ro
o
o
o
o
o
o
C\J
o
■
o
ro
r,
72
representative of all the velocity models at the
the receiver array.
9 stations comprising
Station PWA will be used as a guide for the
discussion of the structure of the Mauna Loa profile. Figure 3.19
shows the boundaries, grid lines, and velocities used in the model for
station PWA.
Figure 3.20 is a perspective view of this velocity
structure, plotted with distance along the profile as the axis in
perspective.
A noteworthy feature of the model is a high-velocity region
located beneath the coast (about 140 km range in the figures),
Interpretation of the traveltime data alone would not reveal the
presence of this region.
It was included in this model on the basis
of a gravity high that runs parallel to the coast through the profile.
This feature will be discussed in greater detail in the next chapter.
The upper crust of the model is composed of a layer which extends
across the entire length of the profile (figure 3.19). It has a
high-velocity gradient with a median velocity of 3.0 + 0.5 km/sec.
Its thickness is variable.
Under the ocean, on the west end of the
profile, it is about 2 km thick, It gradually thickens landward until
it reaches a maximum thickness of about 3.2 km near the coast.
Further eastward , the layer thins toward the top of Mauna Loa's north
flank and gradually thickens again toward the east end of the profile
to a maximum of 3-5 km.
constrained.
The velocity of the layer is not well
The value of 3-0 km/sec was chosen to be consistent with
what was found for the upper crust in the Kilauea profile.
This
velocity is an intermediate value for the two upper layers in Hill's
(1969) model.
"
o
o
CM
<0*
CM
00
I
o"
o
X
bO
_
C
"H ""
o o
JC -H
CO 40
< cfl
2B CD
*a
r,
10
o
co
hi
N.
C cfl
O X
"H CD
40
Cfl rH
4-5 Cfl
CO o
"H
40
O U
<m 0
_
i
>
rH
\o _
0 "
O CO
id
"3
a
f\
"3
O JbO
0
S- *3
3 C
40 Cfl
Cm -H
I
O
CO
0
3 CO
JL, <D
40 -H
CO 5-
T-
CO
O
X)
X
<
CO
<
O
UJ
CO
UJ
o
-j
"
..
rH
0 0 rH
sz
4-5 .=r
1
T"
LU
10
>>
TJ
4-5 C
"H 3
0 O
10
10
o
"
o
I
0 -H
x: h
+0
>
CM
CO
ON
rH
UJ
«J
ro
z
<
0
s3
bO
"H
hi
I-
2
cr
UJ
Q.
CL
D
<
_2
O
"
o
o"
o
"
o
o"
o
(\1
o"
o
ro
74
VELOCITY (km/sec)
2
Figure
3.20
- Perspective view
4
of the velocity structure is figure
high
Note
the
velocities and gradients beneath the
3.19.
Kona coast.
75
The thickness of the upper crustal layer is well constrained in
the vicinity of the Kona coast.
The ray diagrams for the stations
west of the flank show a ray set which passes directly beneath the
base of the layer and then returns to the surface.
These rays do not
reach the west end of the model and hence there is little control of
the thickness of the layer under the deep ocean.
Its thickness
beneath the deep ocean was arbitrarily chosen as a balance between the
corresponding configuration in the Kilauea profile and the average for
the Pacific from Raitt (1963).
The thickness of the upper crust is
also poorly constrained on the east end of the profile. Its thickness
here is traded off against the thickness of the midcrustal layer.
The midcrustal layer has a moderate velocity gradient with a
median velocity of
5.45
+ 0.5 km/sec. Its thickness is more variable
than the thickness of the uppercrustal layer. The midcrust is
pinched out almost completely on the west end of the profile. At a
range of 65 km, it begins to thicken until it reaches a maximum of
about 3.5 km under the Kona coast.
Toward the top of the flank, the
structure of this layer becomes complicated although this is not
apparent in figure
3.19.
The models at the other stations must be
inspected to appreciate this structure.
The sequence of stations up
the flank is:
CAC (near the coast), PWA, XII, and PUO (2/3 of the way
up the flank).
The midcrustal layer is progressively thickened under
the stations towards the top of Mauna Loa's flank.
The ray diagrams
in figures 3.11, 3.16, 3-14, and 3.15 for stations CAC, PWA, XII, and
PUO, respectively, show the progression of this thickening.
This
76
change is not shown on figure
3.19 since the
change takes place in the
third-dimension; the stations do not all lie on the same line
connecting the shots.
The thickening of the midcrust is taken into
account in the gravity model to be discussed in the next chapter.
At the top of the flank, the midcrust is required to thin again
as shown in the ray diagram for station WST (figure 3.18).
Resolution
of the thickness of this layer at the east end of the profile is poor
because the rays pass through this region only on their way to the
mantle.
Therefore, the thickness of these layers is traded off
against the velocity.
It is significant to note that from stations
BSC to ISB to HSS (figures 3.10, 3.13, and 3-12, respectively) the
upper crustal layers must thicken progressively to a total change of
believe that this is a
about skm at the east end of the profile. I
three dimensional change in structure and that this thickening of the
upper crustal layers reflects the dip of the high velocity core of the
rift away from its surface expression. This is supported by the fact
that the models for stations SMR and ISB (figures 3-17 and 3.13,
respectively) have the same thickness for the upper crustal layers.
This correspondence should be expected since the two stations lie on
the same line with the shot profile.
The third or basal layer of the crust also has a variable
thickness.
Under the ocean at the west end of the profile, it is
about 3 km thick.
Westward it dips down at about 3-3 degrees and
thickens slightly to about
3-8 km. Just seaward of the coast, the
layer thickens dramatically eastward until it is at its maximum
77
thickness of about 22 km beneath the top of Mauna Loa's north flank.
It has a mild velocity gradient with a median velocity of 7.15 + °-5
km/sec.
This layer forms the major part of the crust of the island
and reaches to within a few kilometers of the surface under the north
flank. This same combination of high velocities close to the surface
is also observed in the Kilauea profile beneath the rift zones.
This
suggests that the north flank of Mauna Loa is also a rift zone, even
though it is generally recognized as one. However, a close look at
the geologic data also supports this interpretation. Pit craters and
cinder cones occur along the north flank, although not in the same
abundance as found on Mauna Loa's major rift zones.
Resolution of the thickness of the basal, crustal layer has
varying degrees of precision. At sea, on the west end of the profile,
the thickness of the layer varies with its velocity.
The value of
7.15 km/sec was chosen to be consistent with what was found from the
Kilauea profile and the results of Hill (1969). Under the coast, all
the Pn rays shown in the ray diagrams turn directly beneath the basal
layer. Its thickness, therefore, is well constrained in this region.
Since, by definition, the base of this layer is the Moho, it is also
well constrained in this region. To the east of Mauna Loa's flank,
there is no constraint on the base of the layer from the refraction
data.
However, some control on this boundary is provided by
teleseismic data discussed below.
Strong Moho reflections (PmP) observed at station PWA (figures
2.12 and 3.16) and weak PmP arrivals at station CAC (figures 2.6 and
78
3.11) help constrain the position of the Moho.
The observed tendency
for PmP to disappear abruptly is also predicted by the model as a
structural effect of layer three necking down just to the west of the
coast.
In addition, a strong PmP arrival is also observed at station
PUO (figure 2.11).
was unable to model the traveltime of this phase
I
to better than 0.3 sec, as is shown in the ray diagram (figure 3-15).
The calculated arrivals are in fact early, which suggests that the
crust is somewhat thicker beneath station PUO than the model
predicts.
Or, it may be that the observed branch is not PmP at all,
but a reflection of off some other feature of the crust.
Whatever the
case, all the other arrivals at PUO are well predicted by the model.
The three layers discussed above comprise the crust.
At the west
end of the profile, the crust is about 4.8 km thick with the Moho at a
depth of about
9.5 km below sea level. Westward, the crustal
thickness follows the trend of other layers; it thickens until it
reaches a maximum of 19.5 km under the north flank of Mauna Loa.
The
Moho lies flat at the west end of the profile. At about 60 km from
the coast it dips landward at about 3«3 degrees. Further landward,
the dip increases to about 8.5 degrees. The sharp bend in the Moho
beneath the north flank of Mauna Loa is probably an artifact of the
model inasmuch as analysis of the gravity data suggests that the Moho
flattens out under the flank.
Precise values for the gradients of the
crustal layers are not given because they are not well determined from
traveltime data alone.
79
Teleseismic Data
Ellsworth (1977) has compiled Jeffreys-Bullen teleseismic P-wave
residuals at stations of the USGS permanent Hawaii seismograph
network.
The sources range from epicentral distances of 60° to
90° and are
positioned at all azimuths around Hawaii.
Three of the
stations of the USGS net, which were used in Ellsworth's study, are
part of the Mauna Loa profile. Data from these stations provide an
independent check on the velocity structure calculated from the
refraction data.
To use the data, two assumptions were in the modeling process.
First, it was assumed that the P-wave residuals reflect variations in
crustal structure only. That is, a teleseismic wave incident at the
base of the crust is not disturbed by features in the mantle of a
smaller scale than the aperture of the three recording stations.
Second, it was assumed that the teleseismic waves are plane waves
which make an angle with the vertical of 25 degrees. The curvature of
the waves should not introduce significant error for array apertures
less than 150 km (D. Oppenheimer, personal communication,
198l). For
epicentral distances of 60° to 90°, the angle of incidence at the
base of the crust should range from
the Herrin Tables; Herrin, 1968).
30° to 20°
(calculated from
The value of 25° was chosen as an
Rays were propagated through the model, and a traveltime curve
representing relative teleseismic P-wave residuals was obtained for
the model at each of the
3 stations. Figure 3.21
shows the ray
80
"
o
o
CM
E
O
X
X
r,
Cm
5X
rH
0
3
o
"
o
X
X
o
E
LO
0
r_
40
X
_.
o
X
<M
X
to
>>
X
X
X
CO
M
0
"H
x
x
B
oo
"H
0
co
0
<-i
o"
o
o
o
4-5
v_>*
r,
o
<M
B "
CO CO
CO
bOX
cfl
"H C
*3 O
r,
"H
Cfl Cfl
OS 4-5
CO
I
rH
CM
m
o
o
0
3
-D
"H
hi
U
LO
00
00
>
■
>
o
■
o
»
lo
0-00001/130-1
o
ro
o
CM
O
o
o
o
o
o
o
o
CVJ
o
o
ro
■
TIME (sec)
81
CM
o
CM
CO
CO
X
X
co
0
rH
0 0
o
X
2_
v >
IO
-H
O 40
"H
Cfl
X
"O rH
0 0
CO
o
<
r,
o "3
rH
O CD
40
" 3
Q.
bO B
C O
"H
rH
0
..
0
"o CO
O CD
E co
o
o
E
j_
UJ
o
z
<
00
0 o
"H S*
E O
CO
"H
0 00
CO rH
0 CO
rH
3
0 T3
4-5 -H
ICO
0 0
CO SZ
r,
o
40
c_
<M
*?
|
O-3
"
CO
CO "3 rH
-P 0 cfl
rH
3
3 £- "3
CO CD -h
0 CO CO
OS X) 0
O h
I
>
o
IO
CM
CM
en
0
M
3,
"H
&4
82
diagram for the model at station HSS.
The relative residuals at the
three stations are plotted with the J-B residuals in figure 3.22.
The
relative residual at station HSS was set equal to the J-B residual so
a comparison could be made.
Agreement to within +0.1 sec is observed
between the observed and the calculated data.
These data help confirm
the validity of the velocity model.
Amplitude Modeling
True relative amplitude plots for stations PWA and PUO are shown
in figures 3.23 and 3.24, respectively. These stations were chosen
because they provide the best amplitude data of all the stations.
Inspection of these record sections reveals strong PmP arrivals which
occur over a short distance range. At the small shot-receiver
distances, the amplitude of the crustal arrivals is not quite so large
as PmP and their distance dependence is complicated.
The Pn arrivals
appear over a long distance range and their amplitude tends to taper
off slightly with distance.
Figure 3.25 shows the observed amplitude
behavior of the Pn and PmP branches observed at station PWA.
that PmP rises and falls off rapidly with distance.
Note
Pn follows a
pattern of falling off with distance until about 70 km and then
maintaining a fairly constant amplitude for distances greater than
km.
70
These data were used to model the fine velocity structure of the
Moho and upper mantle.
For two reasons, no attempt was made to model
the amplitude of crustal arrivals:
1) the data available are limited
and only represent energy which bottomed in the midcrustal layer, and
83
o
o
w
o
\
ii
\f.
|/w^
CI
r*»
\j\[\h^^
w
c>
6
v
»ii i»*i" *"*i
v
»
ci
'liif
IT
in
o
"o
S
o
_.
o
"U"
Cm
Cfl "
4-5 X
Cfl
"3 >>
o
"o
IT
0
"kfi
IT
40
_. rH
CO
0
o
Cm
O
—
s
'
VvWv^v wvVW^
i/V^
__
V^w^^^-
"-—
Mv^vv*M/np/vv'^\A/V^
~—
XI
x: -3
0
o
o
" u-
■s r:
_;
CO
C CD
O M
"H CO
40
0 CO
0
0 "3
CO
CE
o t_''
QL
S
0 40
IT
3 rH
4-5 Q.
"H S
iH Cfl
rv Z
3
Q.
S CD
E-4
0
>
"
"H
40 «3J
Cfl IS
co x:
-
W
ci
\r
9
O
rH PL,
0
L. C
O
o
o
"g
.
IT
Bfl
cTt
CX!
C
6
0»
C
I
G9
33S '8/X-I
C*S
C
»
CC
6f
0 -H
r, CO
H 40
I
CO
CM
CO
CO
84
rj
O
r
ir
c
o
"H
u~
CO
4->
CO
Cl
C
M/W/iM/W\/VnAJ|ffy
o
L
Cm
o
IT
4-5
CO
o
6
B
■AA/YVW^^\A'VV'VV^
0
o
IT
o
Cm
c
o
o
"H
o
IC
o
T.
s
O _]'
CL
X
Z
E
I—
40
0
0
CO
0
"3
3
+o
H
.r-l
X
"
CO X
B
I
o
8
a
3
o
ir
o
o
.ci
CD >>
>
X)
"H
40 T3
CO CD
-H rH
0 Cfl
Jh 0
0
»
-J*
CM
_
ro
o
o
n*
O
si
□
?'
C
*'
GO*
B'l
ci
gj:
0 9
33S '8/X-J
I
C'S
c
t-
0E
0
-
CO
3
r, O
H
PL.
0
"H
J■
■
■
c
85
o
o
a.
X
X3
I
■
T3
0
rH
CO
0
■
CO
■
(
■
"D
CD
°*5 _"
P-r
40
CO
£-,
(
O
■
■
o
"H
>
cfl
E x:
0
X)
"I■
UJ
o
z
<
!■
0
T3
3
40
"H
rH
H a
C0 E
co
D -3
■
0
>
_,
0
co
n
o
■/
/
i
in
CM
■
■/
■
E
(
■/
■
/
3anindwv 3Aiivi3a
0
o
to
v
3
bO
"H
86
2) the synthetic seismogram program fails when lateral variations
within a layer exceed about
10%. Since the
high velocity region under
the coast exceeds this figure, the high velocities were removed
artificially before proceeding with the modeling.
This produces a
mismatch between the observed and the computed traveltimes near the
critical distance for Pn (figure 3-26), but Ibelieve that the model
gives a close approximation to the real structure.
The fine structure for the lower crust and upper mantle which
resulted from the modeling is shown in figure 3.27.
A large velocity
step is required at the base of the crust to model the observed
behavior of PmP.
Then a transition zone in the upper mantle of
decreasing gradients is required directly beneath the large step in
velocity. Figure
3.28 shows the
synthetic seismogram for station PWA
plotted at the same scale as the data (figure 2.12).
Note that the
source is a simple two-sided pulse. Figure 3.29 shows the
amplitude-distance behavior of Pn and PmP plotted in the same fashion
as for the observed data.
Reasonable agreement between the observed
and computed has been obtained.
The small distance range over which PmP is observed can be
explained in a different manner.
If the Moho is not a first-order
discontinuity, but a transition zone with a high velocity gradient,
this would produce a short retrograde traveltime branch which could
easily be misinterpreted as a reflection.
Close examination of the
phase changes of the PmP wavelet could resolve the difference;
however, the present data are not good enough to do this.
O
O
CM
CO
c
o
us
0
x:
4-5 T3
o
o
_.
LO
\y^G
c
CO
-
0
■3 "O "3
C 0 CD
3>
m
\y
0
?
"
>
I
i^
m?/
O
>
St
CO
co 0 CD 0
0 CO CO E
"H X2.O -H
40 O O 40
"H
"" rH
0 <M
0
O O CO >
rH
CD Cfl
0 x: rH C_
0 0 4-5
4-5 £-,
X n) -h-o
bO E 3 0
"H CO
4-5
SZ <__ C 3
B <D D.
3. E
x:
40 0 O O
"h
x:
o
S 40
<0
O
CO 00
0
CO
CO
3 40 0
B
CU O «H
_. 2S 40
O " H
o
Cm
"3
O
Li
0U
>
"
<D CO
6 >
S- CO
3
Sh O 40 0
bO S
E
0
3 0 *3 -H
"H U CD 4-5
-O
+5H
40 3 0
CO O.
>>
>
33 E 3
OS O O U
0 040
I
CM
m
o
o
0
_,
3
bO
LO
"H
eJP
3
o
o
o
LO
tT
ro
o*B/130-1
o
o
o
cm
o
o
o
o
o
o
CM
O
O
ro
88
CM
CO
U
CO
"5
■>.
E
>-
r
i>
00
-*
-P
c
3
S
1/
o
o
-J
UJ
>
M
0
a
a
3
"3
3
3
|
O
X
0
40
o
<M
o
CO
2
3
4->
0
3
_.
4-5
CO
0
C
|
r,
r_
0
_.
3
5)
"H
I_
CM
X
«*■
<WM) H±d3d
o
CM
CM
89
o
0 T3
-P C
O 3
as
CO "
" r-K
C
X 3 O
o
9
>> D.-H
4J
£■ TJ O
0 c
"3 T3 3
0 -H Cm
rH CO
3 I 0
000
CO 3 £-
<
in
o
S 0 00
PU rH
3. rH
C B 3
O -H 0
H Xl C
4-5
o
3 3 0
10
4-5
SZ
CO 00 40
"H
occ
UJ
—
et
Cm O 0
"H 00
E 40 0
3 0
r,
c a
m
bO 3 0
O <m t.
B
3 0 O
m
"H 0 4-5
0 M
OO 3 TJ
O 0
0 00 TJ
■H
C
40 0 0
0 x: 40
x: 40 c
+0
a
o>
-H
c
>> 40
3 +0
CO £
40
00
CM
en
o
0
"H
1-4
o
p»
<a
j.l
in
3NU
tf
O
c
9
■
■
■
c
.o
o
X3
0.
3
0
I
■
o
<
1
rH
UJ
I<
3
3
C
O
■
ii
<
"H
4-5
3
40
3
■
4-5
3
_,
O
>
E
__\
w
■
LU
o
z
<
■//
H
C0
Q
■
S-,
0 3
T3 bO
3 -H
40 (m
"H
x:
a. 40
E -H
3 £
T3
3
40
3
3
O
o
IO
CM
OO
0
M
3
bO
"H
Cr,
aarundwv 3Ai±vi3a
£,
D.
3 _■
rH
I
f
0
CO
B
d
3 O
■
/
.
"H LO
CM
3 "
JC CO
0
"O 0
"
X
90
91
CHAPTER
GRAVITY
IV
DATA
AND
INTERPRETATION
92
In this chapter, density models based on the Bouguer and Free-air
gravity data from Hawaii are described for the Mauna Loa and Kilauea
profiles. The aim of the gravity modeling is twofold:
first, to
provide a check on the consistency of the results of the seismic
modeling and second, to extend these results into areas where the
seismic coverage is limited.
The density models were constructed from
the velocity models described in the preceeding chapter by converting
velocities to densities using conventional velocity-density
relationships. Two-dimensional and modified two-dimensional
techniques were used to calculate the theoretical gravity field from
the density models.
Adjustments were made to the models until
satisfactory agreement between observed and calculated gravity fields
was obtained.
DATA
Separate sources provided the gravity data over the land and over
the ocean used in the modeling.
The complete Bouguer anomaly (CBA)
map of Kinoshita et al. (1963, my figure 4.1) was used for the
subaerial part of the profiles.
The Bouguer reducing density was 2.3
g/cc-an average density of the surface lava flows measured around the
island (ibid).
Inspection of the map reveals that topography is
well- correlated with gravity in Hawaii.
Gravity highs are observed
over the summit areas of all the volcanoes, with the exception of
Hualalai.
Elongate gravity highs are also observed over the rift
Figure 4.1. Complete Bouguer Anomaly map for the island of Hawaii.
Triangles: seismograph stations used in this study. Solid lines:
locations of gravity profiles. Circles: gravity stations. Heavy
solid lines: gravity contours, interval 10 mgal, dashed where
uncertain. Lines with short dashes: gravity contour, interval 5
mgal. Thin solid lines: topographic contour, interval 1000
feet. After Kinoshita et al. , 1963.
94
zones of the volcanoes.
This phenomenon is not surprising.
Since the
rift zones are cored with high velocity material, as was described in
the preceeding chapter, one might expect high-density rocks to be
present
in the rift zones also.
At sea, the free-air anomaly map (FAA) of Watts (1975) is used.
This map is a compilation of the data collected by Lamont research
vessels from many cruises around the Hawaiian Ridge and vicinity.
Since the map itself is too large to fit conveniently into a page-size
format, it is not included here as a figure. Instead, it will be
described briefly:
The FAA is well correlated with the bathymetry of
the Hawaiian Ridge.
A band of negative anomalies, about 100 km wide,
surrounds the ridge, corresponding to the Hawaiian Deep. These
negative values average about -20 mgal.
A 200 km wide band of small
positive anomalies of less than 15 mgals, corresponding to the
Hawaiian Arch, surround the negative values.
The difference in the scale of the two maps presents a problem
when one attempts to use them together. The scale of Watts' FAA map
is 4 x 10°: 1, some seven times larger than Kinoshita's CBA map.
In
a study in which the coverage on the CBA map is adequate, the coverage
on the FAA map is insufficient.
The gravity data along the Kilauea
profile illustrate this point.
The top of figure 4.2 represents the
data as dots.
At the coast, there is a mismatch in the data (question
marks in the figure).
This mismatch is probably due to the fact that
the profile lies in the middle between two ship tracks which are
spaced about
70
km apart at the coast.
Since the observed gravity
MGAL
X
h-
o
DEPTHI (KM)
o
o
I
i
L_
o
2
CM
CO
CO
co
Q
a
UJ
>
UJ
X
UJ
3
0.
-H
4-5
w
—-
CO
H
o
co 2
CD O
o o
ll
3
40
O
*
°
-»
2E
"w
X
LU
II
to
<
\
-H 0 ""
0
"3
rH 00
5
"H C
O
r,
0 3 L
OA L
£- E <U 3
O. 3 X
2 E
3
3
0 " C
3
3 3 "3
rH 40 0
"H cfl C
US T3 -H
5
bO
C
-H
40
C
O
-H
Q.
rH |
0 T3 U 3
x: 0 0 5
4-5 40 T3 O
UJ
CO
"
.cc
..
C
O
*h
"
3C Q 0
H D
rH
-H
3
0 " « «"-"
rH
bO O
3 O C U
iH
0 0 0 -H O.
*3
\ rH
O
-3 3 3
B 0
"3 O
""
0»
CM
"w
co
o
X
o
>>
_>
-3
o
JoJ o
X
LU
"
"
c-
CM
CO
""'""/
— «A
voi visinvw
X
o
Z
//
f
w
UJ
_J
H
Z
<
2
CM
3
I
3
C
3
3
S
„ x:
C-P
0 O *H
" -O
0 5
rt
!C
40 "" "3
O
cfl CO
C
*o 0 0 >H
3
-P
Q
"a 3 0 0
0 x: sz 3
40 4-5 CO
CM U 3
.00 E 3
:=T CO
O 4-5
X> 3 S-. C
3.Cm
0 O
-H
r,
(Idld 3N)
w
co C
-
>
r,
CM
c-
X
C
3
3
40
CO
.
o
-J
X
C3
O J
-H «H B
4-5 rH
(OO
"H
3 "3 0 0
CH H 45
0 rH 4-5 0
T3 O -H 0
So
"H
t_
96
field is only sampled along the ship tracks, resolution of the FAA
near the coast is expected to be poor. The tracks angle toward each
other at sea and intersect with each other and the profile at about
110 km from the coast.
Therefore, resolution of the FAA should
improve towards the south end of the profile. In the modeling, the
first FAA data points near the coast were ignored.
In the Mauna Loa profile figure (4.3) there is no problem
matching the two data sets. In the FAA map, a ship track crosses the
profile just a few kilometers from the coast.
Another ship track
crosses the profile at about 100 km offshore.
Since ship tracks cross
the profile at these points, one would expect that the data are better
defined along this profile.
METHOD OF ANALYSIS
Computational Methods
Inspection of the CBA map (figure 4.1) shows that the anomalies
tend to be circular rather than elongate.
This suggests that
three-dimensional modeling would be an appropriate method of
analysis.
However, the complexity of the model that would result is
not justified with the presently available data.
Therefore, the
approach taken was to first develop a two-dimensional model.
Then a
modified two-dimensional method was used to correct for the fact that
some of the structural features are not truly two-dimensional.
Forward 2-dimensional modeling of the data was done with a
computer program that utilizes the standard 'polygon' method of
MGAL
CO
<
UJ
DEPTH I(KM)
o
o
o
o
CM
o
_l
o
CO
o
CM
I
I
97
►
(>!NVId HIHON)
VOl vnhvw
to
CM
CO
CO*
CM
UJ
CM
-J
:>
7
'M
Z
<
.
cm" ci 2
\\
00 CO
2
X
UJ
OL
o
CO
0.
3
\
\
\
CO
s
3
CO
0
rH
"H
<M
O
£-
to
3
w
3
C
3
3
O
\
\
\
\
\
\
0
ISVOO
0
X!
4-5
U
O
Cm
r-\
0
O
-3
s
!
CO
CM
a>
2
CM*
o
I-
IO
CO
3
X
Q
UJ
>
X
UJ
CO
CD
UJ
h3
CL
2
O
O O
II II
"
\
CO
o
O
X
o
o
5
-J
_
3
bO
X
"H
UJ
<
<
"
LU
3
_
0 rH
co
it
bO
3 O
4-5 Cm
3
Q CO
3
" "3
CO
" C
0
"=r bo
0
UJ
CO
0
T3 3
3
o
X
Ul
>>CM"
4-5
"H St
w
3 3
£-,
o
98
Talwani et al. (1959).
This method approximates the structure using
any number of infinitely long prisms which are polygonal in
cross-section.
The density contrast is specified for each prism and
the total attraction of all the prisms can be calculated at an
arbitrary number of field points with arbitrary location along the
profile.
This two-dimensional model was then modified using an
end-corrected two-dimensional computer program adapted from Cady
(1977).
This method allows truncation of the ends of the prism at a
finite distance away from the profile. In doing this, it is assumed
that the material on the other side of the truncated prism has the
same density as was used in the Bouguer reduction.
case, this assumption is probably justified.
The reducing density of
2.3 g/cc used in the reductions corresponds to the
lava flows.
In the present
density of surface
These flows probably surround any feature of finite
extent (e.g. a rift zone) in the crust.
Velocity-Density Relationships
A reasonable estimate of the density structure from the velocity
structure, requires a reliable velocity- density relationship. The
velocity and density of a sample of rock can be measured in the
laboratory and plotted to obtain a velocity-density curve.
Dobrin
(1976) has published a curve (figure 4.4) for all types of rocks which
summarizes the results of many studies. Manghnani and Woollard (1968)
have published a curve (figure 4.5) for Hawaiian rocks measured at
99
Figure 4.4.
Velocity-density curve (reproduced from Dobrin, 1976) for
—
r—
1
i
/
/
NAFE ond
j
//
DRAKE
*"
./
f "/
60
'/" ""/«"
I
">""
L
"
"
."
"
100
."
«r
" **
.
""
j+.
"f:*i
/_
,,f i"
?
t.
,00
**c
Ofc?
?'o/
-
"**»...
/
40
7
%J% J
_M-
"/*
~ //»"."/"/"/"/./
v
"°.
"to
./»
VO
/
/
/
.
.
'.
"
/
/
/
20
I
BULK DCNSITY
Figure 4.5.
S£
20
{gn-, 'cfr 3 )
Velocity-density curve (reproduced from Manghnani and
Wollard, 1968) for Hawaiian rocks.
Figure
relationships for the Bay-of-Islands
(reproduced from Salisbury and Christensen, 1978).
4.6. Velocity-density
ophiolite
101
atmospheric conditions.
Comparison of these curves suggests that
Hawaiian rocks tend to be higher in density for the same velocities
than the average of all rocks.
From these curves, the velocities calculated in the preceeding
chapter can be converted to densities. For oceanic velocity models,
another way to do this is to measure velocities and densities in an
ophiolite. This has been done for the Bay-of-Islands ophiolite by
Salisbury and Christensen (1978).
Figure 4.6 shows their results.
The crustal part of the section shows a small gradient in density. At
the base of the section there is a jump in density to about 3.3 g/cc
which is interpreted as the Moho.
A surprising result from this study
is that the densities do not vary as much as the curves in figures 4.4
and 4.5 indicate.
Since the above curves indicate different densities for the same
velocities, it was necessary to chose densities considering other
information also. Densities were assigned in the following manner:
A density of 2.3 was assigned to the upper crustal layer of the
profiles, based on the Bouguer reducing density used by Kinoshita _et
al. (1963). The density of the midcrustal layer was more difficult to
choose.
The velocity of this layer shows large variations, ranging
from 4.6-5.4 km/sec.
From figures 4.4 and 4.5 it is seen that
densities ranging from 2.6 to 2.7 g/cc are appropriate. The density
of these layers were not allowed to vary in the modeling.
In contrast , the density of the lower crustal layer was allowed
to vary during the modeling. A value of 2.9 g/cc was finally arrived
102
at after repeated adjustments.
This is within the range indicated for
a velocity of 7.0 km/sec in the curves of figures 4.4 and 4.6.
The value for the upper mantle density was also allowed to vary
slightly.
The final value after completion of the modeling turned out
to be 3.25 g/cc.
This value is somewhat lower than the upper mantle
density of 3-35 g/cc used by Strange et al. (1965), and the
3.4
g/cc
value used by Watts and Cochran (1974) in studies of Hawaiian density
structure. However, the results of Salisbury and Christensen (figure
4.6) suggest that upper mantle densities may be around 3-3 g/cc or
lower. In this light, the value of 3.25 g/cc is not unreasonable.
DENSITY STRUCTURE
Kilauea Profile
Figure 4.2 shows the density model for the Kilauea model along
with the observed and computed gravity values.
The model is sensitive
to changes of about +0.05 g/cc in the density values, however the
actual resolution of the densities is probably only +0.2 g/cc, since
density can always be traded-off against structure within the
available constraints.
The boundaries seaward of the coast were taken
directly from the results of the traveltime modeling.
Since the
seismic coverage over the subaerial part of this profile is limited,
the boundaries shown in the density model are only approximately
defined.
For this region, the structure derived from the rift zone
profiles was used as a first estimate, then allowed to vary slightly
during the modeling.
At the north end of the profile, under the
103
down-pointing arrow in figure 4.2, is the the intersection of the
Kilauea profile with the Mauna Loa profile.
The two profiles have the
same density structure at this intersection.
One explanation for the large positive CBA values over Hawaii, is
that a large part of the crust is composed of high density rocks.
This is shown in the model by a very thick section of rocks with
density equal to that of the lower crustal layer.
This result is
consistent with the traveltime modeling from the last chapter.
This
dense layer rises close to the surface under the west rift of Kilauea
and east rift of Mauna Loa, which is also consistent with the results
of the seismic modeling.
The thickened portion of the 2.9 g/cc layer was truncated at
about 40 km on either side of the profile to correct for its finite
extent.
To maintain the agreement between the observed and computed
gravity, the density of this layer had to be raised by
k%
(the
underlined density in figure 4.2).
Mauna Loa Profile
Figure
4.3 shows
the density model for the Mauna Loa profile
along with the observed and computed gravity values.
The estimated
uncertainty in densities is the same as in the Kilauea profile.
The
boundaries in the model were taken directly from the results of the
traveltime modeling. Adjustments to the model were mainly to the
density of the lower crust.
The profile intersects the Kilauea
profile under the down-pointing arrow.
104
High density material (2.9 g/cc) comprises most of the crust of
the island. This is consistent with the observations from the seismic
data.
A small wedge of 2.7 g/cc material is inserted between the
coastline and the top of the north flank to account for the thickening
of layer 2 beneath this region as was discussed in the preceeding
chapter
.
A primary feature of this model is the high density region under
the coast.
This region is necessary to explain the 10 mgal elongate
gravity high over the Kona coast (see figure 4.1). Note that this
gravity high is not correlated with any topographic feature as is
normally the case around Hawaii.
As was shown in the last chapter,
this high density region is consistent with a high velocity region
located in the same place.
The 3-dimensional aspect of the structure is more severe along
this profile than along the Kilauea profile. The high density core of
the north flank probably does not extend north or south of the profile
for more than a few kilometers.
To account for this, the 2.9 g/cc
layer under the island was truncated at 18 km to the north of the
profile and at 20 km south of the profile. The high density region
beneath the coast was also truncated to account for its probable
finite extent.
To the north of the profile the region was truncated
at 18 km, and to the south it was truncated at 12 km.
This had the
effect of raising the density of these bodies by a small amount.
Figure
4.3 shows these as the underlined values.
105
CHAPTER
V
DISCUSSION
106
The velocity and density structure presented in the preceding
chapters can be interpreted in terms of the geology of the volcanoes
The upper crustal layer in both profiles has a velocity of about
3.0 km/sec. As was
briefly discussed in chapter 111, this was not
meant to imply that the lithology is continuous across the layer;
instead, the composition of the layer is thought to change from the
ocean crust into the volcanic pile. In the pile, the upper crustal
material probably represents the most recent material from the
volcano, consisting of lava flows and debris from downslope movement.
In the ocean crust, past the break in slope that separates the main
part of the volcanic pile from the ancient sea floor, the upper
crustal layer probably consists of much older material.
3.0 km/ sec is intermediate to what is
A value of
usually measured for the
velocity of the 'sediment' and 'volcanic' layers in the Pacific
(Raitt, 1963; Shor et al., 1970).
But, these layers are difficult to
differentiate without detailed profiling at small epicentral distances
(ibid.) and sufficiently small sampling intervals were not available
for this study.
The observed value of 3-0 km/sec thus probably
represents an average velocity for the so called oceanic 'sediment'
and 'volcanic' layers
The midcrustal layer is only detected within the volcanic pile,
It has an average velocity of 5.4 km/sec in the Mauna Loa profile and
ranges from
5.1-4.6 km/sec in the Kilauea
profile.
beneath the coast lines and thins out landward.
It is thickest
This layer is
107
believed to consist of older lava flows and pillow basalts from the
initial stages of
growth of the island that were extruded under a
large head of water and, consequently, are probably not as vesicular
as subaerial flows.
Lack of vesicles, together with burial by
subsequent flows, should tend to make the rocks denser and higher in
velocity.
The 7.1 to 6.2 to
6.9 km/sec
layer in the Kilauea profile and the
7.1 km/sec layer in the Mauna Loa profile are taken to be the
'oceanic' layer (layer 3) of the crust.
Although its velocity varies
significantly along the Kilauea profile, its thickness is surprisingly
uniform.
Layer 3 is present in nearly all oceanic seismic refraction
profiles and is probably formed at the mid ocean ridge (Raitt, 1963;
Shor et al., 1980).
Studies of ophiolites suggest that this layer is
composed of gabbroic rocks (Salisbury and Christensen, 1978).
Beneath
the island, the bottom few kilometers of the 7.1 km/sec (density =2.9
g/cc) material probably makes up layer 3of the ocean crust.
The
upper boundary of the layer has been obscured by the presence of the
volcanic pile and probably altered somewhat by the passage of magma on
its way to the surface.
The fine structure of the upper mantle determined from the
amplitude modeling (see chapter III) is remarkably similar to the
seismic structure determined for the 'upper mantle' from the ophiolite
study of Salisbury and Christensen (1978).
measurements of Vp and Vs versus depth.
Figure 5.1 shows their
A comparison of the velocity
structure between 6 and 7 km depth with the structure of the Moho and
108
109
upper mantle shown in figure 3.27 shows that in both cases there is a
large step in velocity followed by a gradual decrease in velocity
gradient to a final slight velocity gradient.
This correspondence of
the data helps to support the theory that ophiolites are remnant
pieces of ocean crust.
Figure 3-27 also suggests another explanation of the low Pn
velocity (7.9 km/sec) measured in the Kilauea profile (see chapter
111, section on velocity structure:
southeast flank).
The Pn
velocity was measured over a short range close to the critical
distance for Pn.
In contrast, Pn in the Mauna Loa profile is measured
over a larger range.
It is possible that Pn in the Kilauea profile
sampled only the very upper part of the mantle while in the Mauna Loa
profile, the mantle was sampled at a greater depth. If the proposed
structure in figure 3.27 is correct, one would expect a higher Pn
velocity in the Mauna Loa profile without having to appeal to upper
mantle velocity anisotropy.
Hill (1969) analyzed a number of seismic refraction profiles
oriented parallel to the major coast lines of Hawaii.
Two of these
profiles, the Kau-Puna profile and the Kona profile, intersect the
Kilauea and Mauna Loa profiles respectively, at roughly right angles.
At the point of intersection, the structure from this study should be
consistent with Hill's structure.
Along the Kau-Puna coast the two
profiles agree well with each other (figure 3-2).
However, this is
not the case along the Kona coast. Hill obtains a crustal thickness
about 3 km greater than what was determined in this study.
This
110
discrepancy could be due to the high velocity region beneath the Kona
coast.
The existence of this region was not well defined by previous
researchers.
Consequently, the thickness of the crust would be
overestimated to compensate for the abnormally high velocities in the
crust.
HIGH VELOCITY AND DENSITY REGIONS IN THE CRUST
Modeling of the rift zone and summit structure indicates that the
velocity and density of the rock beneath these regions is higher than
the surrounding erupted rocks. Geologic evidence (Macdonald and
Abbott, 1970; Wentworth and Jones, 1940) suggests the rifts are
composed of a tightly packed sequence of near-vertical dikes which are
feeders for flank eruptions.
Presumably, material left behind in the
dikes is under sufficient pressure to prevent the lava from degassing
to form vesicles.
This would leave the dike rock denser and with a
higher seismic velocity than the surrounding erupted rocks.
Modeling of the rift zones also suggests that they widen with
depth and that they coalesce such that a major part of the volcanic
pile is composed of high velocity and density rocks.
This phenomenon
is more difficult to explain. Kinoshita et al. (1963) realized that
the large positive Bouguer gravity anomalies over Hawaii require that
a major part of the crust be composed of high density rocks.
Their
preferred structure hypothesized an inter fingering of intrusive sills
with flows from the volcanic centers. However, this hypothesis does
not take into account the observed widening of the of the rift zones
111
with depth.
An alternate explanation genetically related to the
formation of the rift zones seems more plausable.
One explanation,
simply related to the growth of the rift, is possible.
As was
discussed earlier, magma is transported through the rift zones in
subvertical dikes.
The magma left behind in the dikes probably cools
and hardens within a few months, so it is likely that the dikes are
not reoccupied by magma during following eruptions. It is likely that
new dikes form for each eruption, which would tend to push the older
dikes to the side.
Repeated eruptions would tend to push the older
dikes further and further to the side, widening the rift zone with
time.
Futhermore, if the rift zone grows vertically with time, there
would be fewer dikes near the top of the rift zone, which would
explain the widening of the rift zone with depth.
Comparison of figures
3.4 and 3-5 shows that the
east rift of Kilauea is about
8%
velocity in the
greater than the velocity in
Kilauea' s southwest rift. The lower velocity in the southwest rift
suggests that fewer dikes have formed there.
the geologic data.
This is consistent with
Lipmann (personal communication, 1979) points out
that the southwest rift is less developed geologically than the east
rift. Fewer pit craters, cinder cones, and volcanic fissures have
formed in the southwest rift compared to the east rift.
Modeling of the Mauna Loa profile indicates the existence of high
velocity and density rocks within the north flank of Mauna Loa.
suggests that the north flank is a rift zone
generally recognized as one.
This
- although it is not
This conclusion is supported by the
112
geologic data which shows a handful of cinder cones and volcanic
fissures
occuring along the crest of the flank.
Kona Coast Anomaly
The region of high density and high velocity rock that underlies
the Kona coastline is anomalous because there is no topographic
expression of this feature; high velocity and density rocks are
usually found in rift zones.
The absence of corresponding topography
suggests that this is an older feature of the crust which has been
buried by the more recent lava flows from Mauna Loa. From the gravity
data alone, Kinoshita et al. (1963) proposed that this elongate
anomalous region is an old rift zone, possibly genetically related to
Haulalai volcano, which lies at its north end.
This is a reasonable
hypothesis and is consistent with the data presented here.
Forceful Injection of Magma into Rift
Zones and the
Kalapana Earthquake
The depth and orientation of the volcanic-pile/ocean-floor
contact relates to the fault parameters of the November 29, 1975,
magnitude 7.2 earthquake, which shook the island of Hawaii.
This
contact should be located somewhere within the midcrustal layer.
During the initial formation of the island, lava flows spread out over
the then roughly 75-m.y.-old seafloor, presumably on the top of the
sediment layer. As the volcano continued to grow, the sediment layer
became sandwiched between the old lava flows formed at the spreading
113
ridge and fresh flows forming the volcanic pile. This sediment layer
is difficult to detect by the technique used here because a
low- velocity layer is presumably formed by the sediments.
If the
average thicknesses of the 'volcanic' and 'sediment' layers in the
average Pacific crust of 2 to 3 km (Shor et al., 1970) are appropriate
for the southeast coast of Hawaii, then the volcanic-pile/ocean-floor
contact is near the base of the midcrustal layer at a depth of 5 to 6
km below sea level in the Kilauea profile.
The focus of the M 7.2 earthquake was located at a depth of 5 km
underneath the southeast coast of the island near the town of Kalapana
(Tilling et al., 1976; my figure 2.1). Furumoto and Kovach (1979) and
Ando (1979) have completed detailed studies of the source model of the
event.
Both papers present fault-plane solutions based on several
types of data and conclude that the south flank of Kilauea was pushed
seaward along a near-horizontal plane by forceful injection of magma
into Kilauea 's rift zones.
A connection between the mobility of the south flank of Kilauea
with the dilation of its rift zone was first proposed by Fiske and
Kinoshita (1969).
Since then, several authors have used diverse data
sets to help confirm the hypothesized connection.
Koyanagi et al.
(1972) have analyzed earthquake occurrences along the rifts.
They
found that the main clustering of events during volcanic activity is
along the rift zones and that the greatest compressive stress axes of
the events are located perpendicular to the trend of the rift.
Swanson et al. (1976) have used geodetic data to show that the south
114
flank of the island is moving seaward while just to the north of the
rift zones the crust is stable, suggesting dilation of the rift zones.
Furumoto and Kovach (1979) suggest that the dilation of the rift
zones pushed the south flank of Hawaii seaward along the
volcanic- pile/ocean- floor contact.
Their fault plane solution
indicates a thrust mechanism along a plane dipping gently toward the
island. The inferred 20 dip of the volcanic-pile/ ocean-floor
contact from this study supports this mechanism.
The depth of the
focus (5 km) reported by Tilling et al. places the seismic event near
where the top of the ancient seafloor is believed to be.
CRUSTAL THICKNESS
A similar pattern of variation in crustal thickness is observed
in both profiles.
The ocean crust begins to thicken landward at the
break in slope between the volcanic pile and the ancient sea floor.
At this point the Moho dips gently toward land, and the water layer
thins.
At the coast, the crust is about 13 km thick in both
profiles. Beneath the subarial part of the volcano, the Moho dips
more steeply until it reaches a maximum depth of 17 km along the
Kilauea profile and 18 km along the Mauna Loa profile.
The two profiles cross in the interior of the island just north
of the summit of Mauna Loa (figure 4.1).
The seismic coverage in both
profiles is limited at the intersection, but the gravity coverage is
good, which allows the extrapolation of the seismic structure with the
gravity data.
The depth to the Moho at the intersection is about 16
115
km, which corresponds to a total crustal thickness of about
19.5 km.
This point is close to the summit of Mauna Loa, where the maximum
crustal thickness probably occurs for the south part of the island.
These parameters are an integral part of the lithospheric flexure
models that have been constructed for the Hawaiian Ridge (Walcott,
1970; Watts and Cochran, 1974).
The total vertical deflection of the
crust, from the load of the volcanic pile, is an essential parameter
in these models.
From the present study, the depth to the Moho, in
the ocean crust around Hawaii, is found to be about 10 km.
a total deflection of about 6 km.
Walcott in his model.
This gives
This is the same value used by
Using gravity data, Watts and Cochran estimate
the deflection along the older parts of the Ridge at 8 to 11 km.
This
is consistent with the deeper Moho depths observed at Oahu and other
islands (Furumoto, et al., 1971; Furumoto, et al., 1973). This
observed increase in Moho deflection towards the older islands
probably is related to the subsidence of the islands due to the load
on the lithosphere from the presence of the volcanoes.
In other
words, the older islands have had longer to subside and have
correspondingly larger Moho deflections.
LOSS OF HIGH FREQUENCY ENERGY AT MAUNA LOA
The six recordings for the smallest source-receiver distances at
station DAN (on Mauna Loa's southwest rift zone) and SWR (on its
summit) from the Kilauea profile are shown in figure 5.2.
Both
stations are located on high-velocity rift zone rocks and are about
equally distant from the six shots.
The seismograms recorded at SWR
116
-
Figure 5.2
The six recordings at stations DAN and SWR for the
smallest source-receiver distances from the Kilauea profile
117
are clearly deficient in the high-frequency components present in the
seismograms recorded at DAN.
Stations MOK and SLA, also located on
the summit of Mauna Loa, exhibit similar deficiencies in highfrequency energy. Seismograms recorded at these stations from the
Mauna Loa profile show the same phenomenon. In the Kilauea profile,
since the ray paths to DAN and SWR are similar, except for the upgoing
part, the loss of high-frequency energy is most likely to occur along
the upgoing part of the seismic rays.
A natural explanation of this
phenomenon is that the ray incident at SWR passes through a magma
chamber.
However, as Klein (personal communication, 1979) points out,
this is believed to be unlikely. From the Kilauea profile shots, at
least, the ray paths pass south of Mauna Loa's inflation center.
Whatever the case, this loss of high frequencies at Mauna Loa's summit
are not readily explained by the data presented here and awaits
further study.
CONCLUSIONS
Eight main conclusions can be drawn from this study:
1)
Hawaii Island is a pile of volcanic rocks lying on top of a
relatively rigid lithosphere that has been depressed by some
6 km
in response to the load of the pile. Stratification is found in
the volcanic pile associated with the varying ages and properties
of the rocks that comprise it. The pile is also intruded by
sub-vertical dikes in the volcanic rift zones which are feeders for
flank eruptions.
The growth of the island is produced from the
118
eruption of new lavas onto the surface of the volcano, adding to
the pile. However, growth of the island also occurs through the
expansion of the volcanic rift zones.
Profiles perpendicular to
the rift zones suggest that these zones widen with depth and
coalesce with other rift zones to comprise a significant portion of
the crust.
A simple model of repeated injection of dikes into the
rift zones over time can explain the observed widening of the rift
zones with depth.
2)
It has been recognized for many years that magma moving through
the rift zones is capable of generating enough pressure to shove
the flanks of the island seaward.
Recently, it was suggested that
this pressure caused the 1975 Kalapana earthquake by slippage of
the south flank of Hawaii seaward along the
volcanic-pile/ocean- floor contact.
Support for this hypothesis is
provided by identification of the volcanic-pile/ocean-floor contact
beneath the south flank at a depth and orientation consistent with
what was suggested from the fault plane solutions.
3)
The existence of a high velocity and density region has been
identified beneath the Kona coast.
Before, this region was only
speculatively identified from the gravity data.
This region is
believed to be an old rift zone which has been buried by later lava
4)
The southwest rift zone of Kilauea is less distinct
geophysically and geologically than the east rift of Kilauea.
This
suggests that fewer feeder dikes have formed within the southwest
rift.
119
5)
The depth to the Moho is about 13 km along the Kona and
Kau-Puna coasts of Hawaii, deepening to a value of about 19 km
close to a point beneath the summit of Mauna Loa.
6)
The oceanic layer (layer 3) is uniform in thickness around
Hawaii.
This fact, coupled with the roughly 2 to 3° landward
observed dip of the layer, is believed to reflect the flexure of
the ocean crust due to the load of the island.
7)
The crust-mantle transition beneath Hawaii is a sharp velocity
step followed by a decrease in velocity gradient into the upper
mantle.
This is consistent with velocity structure studies of
ophiolites.
8)
The north flank of Mauna Loa is cored with high velocity and
density rocks.
This fact together with the geologic data suggests
that the north flank is a rift zone.
RECOMENDATIONS FOR FUTURE WORK
There are still some important problems yet to be solved in
Hawaiian crustal structure.
anisotropy is still unclear.
The existence of upper mantle velocity
The data presented here suggest the
possibility of anisotropy but the evidence is incomplete. Upper
mantle velocity anisotropy is observed for most of the Pacific (Shor
et al., 1970).
If it does, or does not, exist under Hawaii could have
important implications for the behavior of mantle rocks under abnormal
loads. More complete knowledge of Pn velocities for a range of
azimuths around Hawaii could help resolve this question.
120
A large body of P-wave data are now available for Hawaii island
and most of the older islands.
To get a more complete picture of the
elastic properties of the crust and upper mantle it will be neccessary
to obtain S-wave data also. Long period surface wave analysis and
S-wave delays could provide the general S-wave structure for the
island and the underlying lithosphere. For local problems, earthquake
sources could be recorded with horizontal geophones to construct
refraction profiles.
With such S-wave velocity information, the
attenuative properties of the volcanic pile could be determined and
the paths of transport of magma through the volcano could be more
accurately mapped.
121
REFERENCES CITED
Ando, M. (1979). The cycle of eruptions and large earthquakes on the
island of Hawaii, Hawaii Symposium on Intraplate Volcanism and
Submarine Volcanism, Hilo, Hawaii, July 16-22, p. 55.
Braile, L.W., and R.B. Smith (1975). Guide to the interpretation of
crustal refraction profiles, Geophys. J. Roy. Astron. Soc.
, 40,
145-176.
Broyles, M.L., W. Suyenaga, and A.S. Furumoto (1979). Structure of the
lower east rift zone of Kilauea Volcano, Hawaii, from seismic and
gravity data, J. Volcanol. Geotherm. Res., 5,
317-336.
Cady, J.W. (1977). Calculation of gravity and magnetic anomalies along
profiles with end corrections and inverse solutions for density
and magnetization, U.S. Geol. Surv. Open-File Rep. 77-463, 110 p.
Cerveny, V. (1966). On dynamic properties of reflected and head waves
in the n-layered Earth's crust, Geophys. J. Roy. Astron. Soc,
11, 139-147.
Cerveny, V., I.A. Molotkov, and I. Psencik (1977). Ray method in
seismology, Prague, Karlova University, 214 p.
Chase, T.E., H.W. Menard, and J. Mammaerickx (1971). Topography of
the north Pacific, Chart TR-17, Inst, of Mar. Res., Univ. of
Crosson, R.S., and R.Y. Koyanagi (1979). Seismic velocity structure
below the island of Hawaii from earthquake data, J. Geophys.
122
Dalrymple,
G.8., E.A. Silver, and E.D. Jackson (1973). Origin of the
Hawaiian Islands, American Scientist, 61,
294-303.
Detrick, R.S., and S.T. Crough (1978). Island subsidence, hot spots,
and lithospheric thinning, J. Geophys. Res., 83,
1236-1244.
Dietz, R.S., and H.W. Menard (1953). Hawaiian swell, deep and arch,
and subsidence of the Hawaiian Islands, J. Geology, 61, 99-113
Dobrin, M.B. (1976). Introduction to geophysical prospecting, 3rd cd.,
Mcgraw-Hill, New York,
630
p.
Eaton, J.P. (1962). Crustal structure and volcanism in Hawaii, in 'The
crust of the Pacific basin', Am. Geophys. Un. Mon. 6, 13-29.
Eaton, J.P., M.E. O'Neill, and J.N. Murdock (1970). Aftershocks of the
1966
Parkfield-Chalome, California, earthquake: a detailed study,
Bull. Seis. Soc. Amer., 60, 1151-1197.
Ellsworth, W.L. (1977). Three-dimensional sructure of the crust and
mantle beneath the island of Hawaii, Doctoral Thesis, Mass. Inst,
of Tech., Cambridge, Mass., 327 p.
Ellsworth, W.L., and R.Y. Koyanagi (1977). Three-dimensional crust and
mantle structure of Kilauea volcano, Hawaii, J. Geophys. Res.,
82,
5379-5394.
Fiske, R.S., and W.T. Kinoshita (1969). Rift dilation and seaward
displacement of the south flank of Kilauea Volcano, Hawaii, in
'Symposium on volcanoes and their roots', Oxford, England,
Ilnternat.
Interior,
Assoc. Volcanology and Chemistry of the Earth's
53-54.
123
Fiske, R.S., and E.D. Jackson (1972). Orientation and growth of
Hawaiian volcanic rifts: the effect of regional structure and
gravitational stresses, Proc. R. Soc. Lond. A., 329,
299-326.
Forsyth, D.W. (1977). The evolution of the upper mantle beneath
mid-ocean ridges, Tectonophysics, 38,
89-118.
Furumoto, A.S., G.P. Woollard, and J.F. Campbell, and D.M. Hussong
(1968). Variation in the thickness of the crust in the Hawaii
Archipelago, in 'The crust and upper mantle of the Pacific area,
Amer. Geophys. Un. Mon 12,
94-111.
Furumoto, A.S., J.F. Campbell, and D.M. Hussong (1971). Seismic
refraction surveys along the Hawaiian ridge, Kauai to Midway
Island, Bull. Seis. Soc. Amer.,
6J_, 147-166.
Furumoto, A.S., W.A. Wiebeng, J.P. Webb, and G.H. Sutton (1973).
Crustal structure of the Hawaiian archipelago, northern
Melanesia, and the central Pacific basin by seismic refraction
methods, Tectonophysics, 20,
153-164.
Furumoto, A.S., and R.L. Kovach (1979). The Kalapana earthquake of
November 29, 1975: an intraplate earthquake and its relation to
geothermal processes, Phys. E. Plan. Int., 18,
197-208.
Herrin, E. (1968). Seismological table for P, Bull. Seis. Soc. Amer.,
58, 1196-1219.
Hill, D.P. (1969). Crustal structure of the island of Hawaii from
seismic refraction measurements, Bull. Seis. Soc. Amer., 59,
101-130.
124
Kinoshita, W.T., H.L. Krivoy, D.R. Mabey, and R.R. MacDonald (1963)
Gravity survey of the island of Hawaii, U.S. Geol. Surv. Prof.
Pap. 475-C,
114-116.
Koyanagi, R.Y., D.A. Swanson , and E.T. Endo (1972). Distribution of
earthquakes related to mobility of the south flank of Kilauea
volcano, Hawaii, U.S. Geol. Surv. Prof. Pap. 800-D,
89-97.
Koyanagi, R.Y., X Meagher, F.W. Klein, and G.S. Puniwai (1978).
Hawaiian volcano obsevatory summary 76, January to December
1976:
U.S. Geological Survey., Menlo Park, CA, 68 p.
Lee, W.H.K., and J.C. Lahr (1975). HYPO7I (revised): a computer
program for determining hypocenter, magnitude, and first motion
pattern of local earthquakes, U.S. Geol. Surv. Open-File Rep.
75-311.
Macdonald, G.A., and A.T. Abbott (1970). Volcanoes in the sea; the
geology of Hawaii, University of Hawaii Press, Honolulu, Hawaii,
441
p.
Malahoff, A., and Woollard, G.P. (1970). Geophysical studies of the
Hawaiian ridge and Murray fracture zone, in 'The Sea', part 2,
vol. 4, edited by A.E. Maxwell, pp 73-131, Interscience, New York
Manghnani, M.H., and G.P. Woollard (1968). Elastic wave velocities in
Hawaii rocks at pressures to 10 kilobars, Am. Geophys. Un. Mon.
125
McMechan, G.A., and W.D. Mooney (1980). Asymptotic ray theory and
synthetic seismograms for laterally varying structures: theory
and application to the Imperial Valley, California, Bull. Seis.
Soc. Amer., 70, 2021-2035.
McDougall, I(1971). Volcanic island chains and sea floor spreading,
Nat. Phys. Sci., 231, 141-144.
Moore, J.G., W.R. Normark, and P.W. Lipman (1979). Loihi seamount- a
young submarine Hawaiian volcano (abstract), Hawaii Symposium on
Intraplate and Submarine Volcanism, Hilo, Hawaii, July
16-22.
Morgan, W.J. (1971). Convection Plumes in the lower mantle, Nature,
230,
42-43.
Nichols, J.N., N. Warren, B.P. Luyendyk, and P. Spudich (1980).
Seismic velocity structure of the ophiolite at point Sal,
southern California, determined from laboratory measurements,
Geophys. J. Roy. Astron. Soc, 63, 165-186.
Raitt, R.W. (1963). The crustal rocks, in 'The Sea', M.N. Hill editor,
vol. 3, 85-101.
Richter, C.B. (1959). Elementary Seismology, Freeman,
768
p.
Ryall, A., and D.L. Bennett (1968). Crustal structure of southern
Hawaii related to volcanic processes in the upper mantle, J.
Geophys. Res., 73,
r
4561-4582.
Salisbury, M.H., and N.I. Christensen (1978). The seismic velocity
structure of a traverse through the Bay of Islands ophiolite
complex, Newfoundland, an exposure of the oceanic crust and upper
mantle, J. Geophys. Res., 83,
805-817.
126
Shaw, H.R., and E.D. Jackson (1973). Linear island chains in the
Pacific: Result of thermal plumes or gravitational anchors?, J.
Geophys. Res.,
78, 8634-8652.
Shaw, H.R. (1973). Mantle convection and volcanic periodicty in the
Pacific: evidence from Hawaii, Geol. Soc. Amer. Bull., 84,
1505-1526.
Shor, G.G., H.W. Menard, and R.W. Raitt (1970). Structure of the
Pacific ocean basin, in 'The Sea', 4, pat. 11, Interscience, 3-27
Strange, W.E., G.P. Woollard, and J.C. Rose (1965). An analysis of the
gravity field over the Hawaiian islands in terms of crustal
structure, Pacific Science, 19,
381-389.
Sutton, G.H., J. Kasahara, W.N. Ichinose, and D.A. Byrne (1977). Ocean
bottom seismograph development at the Hawaii Institute of
Geophysics, Marine Geophysical Res., 3, 153-177.
Swanson, D.A., W.A. Duffield, and R.S. Fiske (1976). Displacement of
the south flank of Kilauea Volcano: The result of forceful
intrusion of magma into the rift zones, U.S. Geol. Surv. Prof.
Talwani, M. , J.L. Worzel, and M. Landisman (1959). Rapid gravity
computations for two-dimensional bodies with application to the
Mendocino submarine fracture zone, J. Geophys. Res., 64, 49-59.
Tilling, R.E., R.Y. Koyanagi, P.W. Lipman, J.P. Lockwood, J.G. Moore,
and D.A. Swanson (1976). Earthquake and related catastrophic
events, island of Hawaii, November 29, 1975: a preliminary
report, U.S. Geol. Surv. Cir. 740,
33
p.
127
Walcott, R.I. (1970). Flexure of the lithosphere at Hawaii,
Tectonophysics, 9,
435-446.
Watts, A.8., and J.R. Cochran (1974). Gravity anomalies and flexure of
the lithosphere along the Hawaiian-Emperor seamount chain,
Geophys. J. Roy. Astron. Soc, 38,
119-141.
Watts, A.B. (1975). Gravity field of the northwest Pacific ocean basin
and its margin, Hawaii and Vicinity, Geol. Soc. Amer., MC-9.
Watts, A.B. (1976). Gravity and bathymetry in the central Pacific
ocean, J. Geophys. Res., J3_, 1533-1553.
Watts, A.B. (1976). An analysis of the isostasy in the world's oceans;
1, Hawaiian-Emperor seamount chain: J. Geophys. Res., 83,
5989-6004.
Wentworth, C.K., and A.E. Jones (1940). Intrusive rocks of the leeward
slope of the Koolau Range, Oahu, J. Geol., 48,
975-1006.
Wilson, J.T. (1963). A possible origin of the Hawaiian islands, Canad.
J. Phys., 4_,
863-870.
Yoshii, T., Y. Kono, and K. Ito (1976). Thickness of the oceanic
lithosphere, in 'The geophysics of the Pacific ocean basin and
its margin', Am. Geophys. Un. Mon., 19,
423-430.
Zucca, J.J., D.P. Hill, and F.K. Duennebier (1979). A compilation of
the data from the
1976 Hawaii seismic refraction
experiment, U.S.
Geol. Surv. Open-File Rep. 79-771.
Zucca J.J., and D.P. Hill (1980). Compilation of the data from the
1978 Hawaii seismic refraction
experiment on the west flank of
Mauna Loa, U.S. Geol. Surv. Open-File Rep. 80-451.