Sie sind auf Seite 1von 13

Geological Society of America Bulletin

doi: 10.1130/0016-7606(1984)95<1413:MDOFTS>2.0.CO;2
1984;95, no. 12;1413-1424 Geological Society of America Bulletin

DAVID B. NASH

Yellowstone, Montana
Morphologic dating of fluvial terrace scarps and fault scarps near West

Email alerting services


cite this article
to receive free e-mail alerts when new articles www.gsapubs.org/cgi/alerts click
Subscribe
America Bulletin
to subscribe to Geological Society of www.gsapubs.org/subscriptions/ click
Permission request to contact GSA http://www.geosociety.org/pubs/copyrt.htm#gsa click
viewpoint. Opinions presented in this publication do not reflect official positions of the Society.
positions by scientists worldwide, regardless of their race, citizenship, gender, religion, or political
article's full citation. GSA provides this and other forums for the presentation of diverse opinions and
articles on their own or their organization's Web site providing the posting includes a reference to the
science. This file may not be posted to any Web site, but authors may post the abstracts only of their
unlimited copies of items in GSA's journals for noncommercial use in classrooms to further education and
to use a single figure, a single table, and/or a brief paragraph of text in subsequent works and to make
GSA, employment. Individual scientists are hereby granted permission, without fees or further requests to
Copyright not claimed on content prepared wholly by U.S. government employees within scope of their
Notes
Geological Society of America
on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from on September 30, 2013 gsabulletin.gsapubs.org Downloaded from
Morphologic dating of fluvial terrace scarps and fault scarps near
West Yellowstone, Montana
DAVID B. NASH Department of Geology, University of Cincinnati, Cincinnati, Ohio 45221
ABSTRACT
Analysis of fault and fluvial terrace scarps
in the West Yellowstone Basin south of Heb-
gen Lake, Montana, indicates that, during
degradation, the gradient of the scarp face
and the curvatures of the basal concavity and
crestal convexity of the scarp face decrease
(become more rounded) with time. The pat-
tern of degradation is identical with that
predicted by a simple analytical model for
degradation by creep, in which the change in
elevation at a point on a hillslope is propor-
tional to the curvature of the profile at that
point. The model therefore provides the basis
for morphologic dating, a technique for de-
termining the age of a scarp by the extent to
which its initial morphology has been de-
graded. In order to date a scarp, the coeffi-
cient of proportionality, c, must be known.
The technique is probably not suitable for use
on hillslopes degraded by processes other
than creep.
Both the potential and the limitations of
morphologic dating are demonstrated via
measurements of a sequence of terrace scarps
along the Madison River. The calculated rela-
tive ages are in agreement with the known
relative ages, despite the fact that the gradient
of the youngest scarp is significantly less than
that of the older scarps.
To calculate an absolute morphologic age,
c must be derived from analysis of a scarp of
known age having the same aspect, must be
underlain by the same material, and must be
within the same climatic zone as the scarp to
be dated. In the West Yellowstone Basin, a
value of c = 2.00 0.38 x 10"
3
m
2
/ yr was
calculated from a minimum date of 7100 50
yr B.P. determined by radiocarbon dating for
one of the terrace scarps.
Scarps formed by normal faulting in the
West Yellowstone Basin are much more diffi-
cult to date than are terrace scarps, because
of the possible complexity of their initial
morphology. Fault scarps formed during the
1959 Yellowstone earthquake commonly
consist of an assemblage of two or more
smaller scarps or are superimposed on pre-
existing scarps formed by one or more pre-
vious faulting events. By using a selective
sampling procedure, however, one can screen
out these fault scarps. In this way, it was pos-
sible to determine an age of 2,800 1,100 yr
for a pre-1959 fault scarp in the West
Yellowstone Basin.
Because the fault and terrace scarps in the
West Yellowstone area do not vary signifi-
cantly in height, the morphologic dating tech-
nique proposed earlier by Nash (1980a)
cannot be applied to them. The substantially
modified procedure developed for the present
study does not require a wide range of
heights for each scarp to be dated; it can be
performed using a relatively small number of
scarp profiles if c is known. It is easy to use,
and, when used with considerable care, it is
reasonably accurate (36%-59% in this
study). It is potentially applicable to scarps in
other areas.
INTRODUCTION
One of the problems in determining earth-
quake hazards for many parts of the western
United States is the dating of significant, prehis-
toric seismic events. Several techniques have
been developed for dating fault movements (for
example, see Sieh, 1978), but most depend on
datable materials, such as peats, associated with
the fault. Organic material, however, will rarely
be preserved in scarps formed by normal fault-
ing of alluvial fans within the Basin and Range
province, so that alternate methods of dating
fault scarps are required. One method, morpho-
logic dating, is based on a simple analytical
model of hillslope evolution (Nash, 1980a).
The purposes of this paper are to extend the
method and to apply it to a series of fault and
fluvial terrace scarps in the West Yellowstone
Basin, Montana.
On August 17, 1959, a large, destructive
earthquake occurred in Yellowstone National
Park in the vicinity of Hebgen Lake and the
town of West Yellowstone (Witkind, 1964).
Numerous fault scarps ranging from a few cen-
timetres to several metres in height formed
throughout the area. South of Hebgen Lake, the
fault scarps were generally small (<30 cm high)
and were superimposed on older scarps several
metres in height. Although the ages of the pre-
1959 fault scarps and the general seismic haz-
ards of the Hebgen Lake and West Yellow-
stone area are not particularly important in
themselves because the region is sparsely popu-
lated (although the 1959 earthquake resulted in
a significant loss of lives), the area was selected
for detailed study because the style of faulting
and the character of the material underlying the
fault scarps are similar to those of faulted alluv-
ial fan surfaces throughout the Basin and Range
province. Furthermore, in the West Yellowstone
area, there is a well-developed sequence of flu-
vial terraces separated by terrace scarps which
provide an excellent opportunity to assess the
accuracy of the morphologic dating technique
on hillslopes of known relative age. The study of
the scarps in the West Yellowstone area has also
provided the opportunity to re-examine and
modify the morphologic dating technique. A
fault scarp near Drum Mountain, Utah (Buck-
nam and Anderson, 1979), could be dated be-
cause the scarp profile had a wide range of
heights along the trace of the fault (Nash,
1980a). In the West Yellowstone area, the ter-
race scarps tend to be of uniform height along
their lengths, and the fault scarps tend to have a
complex initial morphology; it was thus neces-
sary to revise the morphologic dating technique
in order to accommodate these different condi-
tions.
The experience gained by morphologically
dating the fault scarps and the terrace scarps in
the West Yellowstone area should be relevant to
the morphologic dating of fault scarps in more
heavily populated regions of the Basin and
Range province and should provide a valuable
tool for assessing regional seismic recurrence
intervals.
STUDY AREA
The Hebgen Lake study area is in southwest-
ern Montana, in the south-central portion of the
Geological Society of America Bulletin, v. 95, p. 1413-1424, 15 figs., 3 tables, December 1984.
1413
1414 D. B. NASH
R.4E. R.5E.
T.12S.
T.13S~
BOZEMAN, MT. 83 Ml. f
SCALE R.4E. R.5E
WEST YELLOWSTONE, MT. 0.5 Ml.^
0. 5 1.0 1.5 2.0
MILES
2.5
0 1 2 3
KILOMETERS
Figure 1. Map of the West Yellowstone Basin south of Hebgen Lake. Study site 1 is located south of the Madison River, and study site 2 is
located east of the South Fork Madison River.
West Yellowstone Basin. This broad, remarka-
bly flat, outwash-inantled, structural depression
is bounded to the south by the Madison Plateau,
to the southwest by the Henry Mountains, to the
east and northeast by the Gallatin Range, and to
the west by the Madison Range. The elevated
central portion of the basin south of Hebgen
Lake is nearly featureless and is tilted a few
degrees to the north (Myers and Hamilton,
1964, suggested that the tilting is a result of on-
going regional tectonic warping). The only large
topographic features within this area of the basin
are north-facing, eiist-west-trending fault scarps
ranging in height from <0.5 m to almost 4 m
and fluvial terrace carps ranging in height from
1 to 11 m. The terrace scarps are along the
banks of the Madison and the South Fork Madi-
son Rivers.
During the Laramide orogeny, crustal com-
pression of the West Yellowstone Basin formed
large, overturned or recumbent folds and low-
angle thrust faults dipping to the southwest. In
the late Tertiary, arching and crustal extension
caused large-scale normal faulting and created
Basin and Range-type mountains (for example,
the Gallatin Range and the Madison Range)
(Hamilton, 1960; Eardley, 1960; Witkind,
1974; Smith and others, 1977).
Most of the West Yellowstone Basin is over-
lain by a remarkably homogeneous deposit of
obsidian sand and gravel (Waldrop, 1975). On
the basis of obsidian hydration rim dating,
Porter and others (1983) concluded that the sed-
iments were deposited 28,000-40,000 yr B.P. as
Pinedale outwash. Deposition may have been
by glacial outburst flooding, which Pierce and
others (1976) and Pierce (1979) stated would
account for the lack of surface distributaries or
channels and the generally featureless topog-
raphy of the basin.
Two sites south of Hebgen Lake (Fig. 1) were
selected for intensive study of fault scarps and
terrace scarps. Study site 1 (Fig. 2), immediately
south of the Madison River (where the river
enters Hebgen Lake north of West Yellow-
stone), contains five well-developed, north-
facing terrace scarps which formed during
still-stands of the Madison River. Study site 2
DATING OF SCARPS, WEST YELLOWSTONE, MONTANA 1415
Figure 2. Map of fluvial terraces south of the Madison River and north of West Yellowstone,
Montana. There are five well-developed sets of terraces, from the oldest and highest (Tl) to the
youngest and lowest (T5). Radiocarbon dating of wood collected from a conical depression on
terrace T4 provides a minimum age for that terrace of 7,100 50 yr. The short, numbered lines
indicate locations at which the profiles of the scarps above each terrace (shown in Fig. 9) were
measured.
(Fig. 3), several kilometres west of site 1 and
immediately east of the South Fork Madison
River, contains an intersecting fault scarp and
fluvial terrace scarp.
OBSERVATIONS
Initial Morphology of Terrace Scarps and
Fault Scarps
The initial morphology of degraded terrace
scarps and fault scarps may be inferred from
active cut-banks and recent fault scarps within
the same area and underlain by the same
material. The fault scarps formed during the
1959 earthquake south of Hebgen Lake and the
high, active cut-banks along the Madison and
South Fork Madison Rivers are assumed to have
formed in the same way as the older fault scarps
and terrace scarps and are underlain by the same
obsidian sand and gravel deposit.
The profiles of the modern cut-banks consist
of a straight midsection separating a straight,
nearly horizontal base and crest (defined in
Fig. 11a below). The profile midsection is kept
at the angle of repose of the underlying sand and
gravel, 33.5, by fluvial undercutting (Fig. 4).
The in situ angle of internal friction, <f>, mea-
sured with an Iowa Bore Hole Shear,
1
was
found to be identical to this angle of repose. The
angle of repose for most cohesionless material
will be equal to the angle of internal friction of
that material (Carson, 1977). The crest and the
base of the profile intersect the midsection to
form a sharp basal concavity and a crestal con-
vexity (Fig. 9a).
The initial morphology of the fault scarps is
more complex than that of the cut-banks. Nor-
mal faulting in the West Yellowstone Basin re-
flects ongoing, north-south crustal extension
(Smith and others, 1977). The near-surface fault
planes in the cohesionless, frictional sands and
gravels therefore form in a state of active Ran-
kine stress and dip at an angle of 45 + 0/ 2. For
most common cohesionless sands and gravels, <t>
will be between 30 and 40, resulting in an
initial scarp slope angle between 60 and 65
which exceeds the angle of repose of the under-
lying material. The initial scarp midsection,
termed the "free face" by Wallace (1977), re-
treats at a high angle (by raveling), progres-
sively burying its base with an apron of debris
sloping at the angle of repose of the underlying
material (Fig. 5). The same pattern of free-face
retreat and build-up of a basal debris apron is
observed along the base of the walls of borrow
'Handy Geotechnical Instruments, Inc., Ames,
Iowa.
pits in the West Yellowstone obsidian sand and
gravel deposit (Fig. 6). Variable time is required
for complete burial of the retreating free face by
the basal debris apron (Wallace, 1977, 1980).
After 20 yr, only a small remnant of the free face
remains on fault scarps formed during the 1959
earthquake (Fig. 7). After the free face has been
completely buried, a scarp profile identical to
that of a fluvial cut-bank forms: a straight, nearly
horizontal base and crest separated by a straight
midsection (the debris apron) inclined at the
angle of repose (33.5) of the underlying obsid-
ian sand and gravel. The basal concavity and
crestal convexity formed by the intersections of
the crest and base with the midsection of the
fault scarp profile will be quite sharp.
Immediately after faulting, the fault scarp pro-
file will seldom be as simple as shown in Figure
1416 D. B. NASH
alluvial * - v
s
/
fan
Terrace T6
2 / alluvial
/
fan
Fault Scarp F1
2
11110? B ? 6
5
Fault Scarp F1'
1
Terrace T6'
t -
N
A

H
0 5 0 1 0 0
Scale (Meters)
Study Site 2
By South Fork Madison River
Scarp continues
but was not mapped
Figure 3. Map of a faulted fluvial terrace of the South Fork Madison River west of West
Yellowstone, Montana. The short, numbered lines indicate locations at which profiles of the
terrace scarp arid the fault scarp (shown in Fig. 10) were measured. Faulting both predates
terrace formation (the segment of the terrace scarp T6', upstream from the fault, is higher than
the segment T6, downstream from the fault) and postdates terrace formation (the terrace
surface is faulted). Field evidence indicates that, since being removed during terracing, the F1
segment of the Ifault scarp has moved only once prior to 1959 and is therefore amenable to
morphologic dating.
5. Frequently, deep fissures or broad, shallow
grabens form at the scarp base (Wallace, 1980).
The progressive outward growth of the debris
apron will, however, bury these features, pro-
ducing a simpler profile morphology. Another,
more serious complication frequently results
from splaying of the fault surface near the
ground surface, yielding not one scarp but an
assemblage of two or more smaller scarps
(termed "multiple scarps" by Mayer, 1982). If
the individual scarps are closely spaced along
the profile, the re treat of the free faces will cause
them to run together, forming a single, continu-
ous debris apron. If the individual scarps are
widely spaced, they will not run together. Even
after the free face of each scarp has been elimi-
nated, the scarp profile will consist of two or
more separate debris aprons. South of the Duck
Creek Highway Maintenance Station, one may
observe such a fault scarp several metres high,
formed during the 1959 earthquake. Approxi-
mately 50% of the total length of this fault scarp
consists of a single debris apron (Fig. 8a), and
50% consists of two or more separate debris
aprons (Figs. 8b and 8c). All but one of the fa ult
scarps that formed in the West Yellowstone ob-
sidian sand plain during the 1959 earthquake
were superimposed on pre-existing fault scarps.
The present fault scarp thus is the result of at
least two separate faulting events. Mayer (1982)
terms scarps formed by more than one faulting
episode "composite scarps." Both multiple (see
above) and composite fault scarps pose liifiicul-
ties for morphologic dating. Offset due to fault
creep would pose a further complication, but it
is doubtful whether creep contributed signifi-
cantly to the total offset of the faults, as there is
no indication of any disturbance to roads or to
the runway of the West Yellowstone Airport
where they cross faults active during the 1959
earthquake.
Degraded Terrace Scarps in
Study Site 1
By studying the series of terrace scarps of in-
creasing age in study site 1, the pattern of hill-
slope degradation with time can be determined.
The five terraces are numbered from t h; oldest
and highest, Tl , to the youngest and lowest, T5
(Fig. 2). The terrace scarp separating a given
terrace from the next higher terrace is given the
label of the lower (younger) terrace surf ice (for
example, terrace scarp T4 separates temces T4
and T3). The scarps were formed by the same
process and are underlain by the same o bsidian
sand and gravel deposit as are the modern cut-
banks along the Madison and the South Fork
Madison rivers and therefore are assumed to
have had the same simple initial morphology: a
nearly horizontal base and crest separate by a
straight midsection sloping at the angle of repose
(33.5) of the underlying sand and gravel. Cer-
tain data were not collected. The oldest terrace
scarp, Tl , was not analyzed because its mor-
phology was substantially altered during clear-
cutting of the lodgepole pine cover, which
involved extensive bulldozing and removal of
soil. Scarps formed by the intersection ol" two or
more terrace scarps were not analyzed because
their initial morphologies may have bei:n quite
complex: undercutting during the formation of
the younger terrace scarp may not have com-
pletely removed the older terrace surface, result-
ing in an initial profile consisting of two or more
subsidiary terrace scarps. This poses a problem
for morphologic dating similar to that posed by
both multiple and composite fault scarps.
Terrace scarps T2, T3, and T4 were profiled
at 20-m intervals along their bases, using the
profiling technique described by Nash (1980b).
The profiles of these scarps (Figs. 9b-?d) are
markedly different than those of modern cut-
banks (Fig. 9a). They are no longer the sim-
ple straight-line intersection of base with midsec-
tion and of midsection with crest. Now the basal
DATING OF SCARPS, WEST YELLOWSTONE, MONTANA 1417
Figure 4. Cut-bank of the South Fork Madison River (SE1/4 sec. 14, T. 13 S., R. 4 E.).
Active fluvial undercutting keeps bank face at 33.5, the angle of repose and the angle of
internal friction of the underlying obsidian sand and gravel deposit.
concavity and crestal convexity have become
more rounded by growing outward at the
expense of the base, midsection, and crest. The
slope angle of the midsection has decreased
significantly from its initial angle of 33.5. The
straight, nearly horizontal crest and base of the
profile may still be seen beyond the limits of the
curvature of the crestal convexity and basal
concavity. It is interesting to note that the
youngest profiled terrace scarp, T4 (Fig. 9b), has
a significantly lower midsection slope angle than
the two older terrace scarps, T3 and T2 (Figs. 9c
and 9d), demonstrating the danger of assuming
that the midsection gradient alone may be used
to determine the relative ages of hillslopes.
The ages of the terraces are unknown, al-
though all must be younger than 28,000-40,000
yr, the age of the underlying obsidian sand and
gravel reported in Porter and others (1983). An
unsuccessful search was made for organic mate-
rial that had been buried in the river channel
when basal undercutting of the cut-bank ceased.
The only datable organic material was found in
a deep conical depression located on the surface
of terrace T4 (Fig. 2). Currently 20 m wide and
3 m deep, the depression is of unknown origin,
but it is probably related to seismic activity
(Myers and Hamilton, 1964; Nash, 1981b).
Digging at the bottom of the depression, we
found the top 75 cm to be made up of sand and
gravel containing specks of charcoal. At depths
from 75 to 175 cm, we found peat with occa-
sional logs, underlain, in turn, by logs mixed
with abundant volcanic ash (identified by R. E.
Wilcox, U.S. Geol. Survey, Denver, as Mazama
Ash). K. L. Pierce (U.S. Geol. Survey, Denver)
suggested an age of 6,600 to 6,700 yr for the ash.
At a depth of 190 cm, a large, well-preserved log
was encountered. A sample of the log, sent to
the U.S. Geological Survey Branch of Isotope
Geology (Lab Sample Number W-4795),
yielded a radiocarbon age of 7,100 50 yr.
More sand and gravel were encountered beneath
the log, presumably indicating the original bot-
tom of the depression. The dated log provides a
minimum age for the conical depression and
thus also for terrace and terrace scarp T4, but
the terrace may well be considerably older than
7,100 yr.
Intersecting Degraded Terrace Scarp
and Fault Scarp in Study Site 2
The second study site is the intersection of a
fault with a fluvial terrace and a terrace scarp
carved during a still-stand of the north-flowing
South Fork Madison River. The portions of the
terrace on the downthrown and on the up-
thrown sides of the fault are labeled "T6" and
"T6"\ respectively, in Figure 3. Where the fault
scarp crosses the terrace surface, it is labeled
"Fl , " and where it extends up and across the
area above the terrace scarp, it is labeled "Fl ' . "
The history of the terrace and fault scarp is
complex, involving several separate faulting
events. Fault scarp Fl - Fl ' consists of a fresh,
small scarp (or set of small scarps), ranging from
5 to 40 cm in offset, that was superimposed
during the 1959 earthquake on a pre-existing
scarp 1.5 to 3 m in height. The surface of the
terrace is offset by Fl , and so the most recent
pre-1959 movement of the fault must postdate
the formation of the terrace. The height of the
terrace scarp, however, averages 7.7 m on the
upthrown block of the fault (section T6' ) but
only 4.6 m on the downthrown block (section
Figure 5. Raveling of the initial loosening-limited fault scarp in cohe-
sionless, frictional material causes scarp face (or free face) to retreat,
forming a basal apron of debris. Because the material fails in an active
Rankine stress state, the initial scarp face should be at a slope angle of
45 = <f>/2 (4> is the angle of internal friction of the material). The slope
angle of the debris apron will be equal to, or very nearly equal to, <t>
(assumed to be 35 here). The diffusion model for hillslope degradation
is not appropriate for the raveling process. Rapid raveling buries the
steep scarp face with a continuous debris apron relatively quickly, usu-
ally less than 50 yr, resulting in a hillslope with a simple initial profile
amenable to morphologic dating.
1418 D. B. NASH
Figure 6. Borrow pit at the north end of runway of the West Yellow-
stone Airport (NE1/4 sec. IS, T. 13 S., R. 5 E.). Raveling of the steep,
loosening-limited'. free face is producing a transport-limited debris
apron sloping uniformly at 33.5, the angle of internal friction and the
angle of repose for the obsidian sand gravel deposit overlying the
West Yellowstone Basin south of Hebgen Lake.
Figure 7. Fault scarp, formed during the 1959 Yellowstone earth-
quake, south of ithe Duck Creek Highway Maintenance Station
(SE1/4 sec. 22, T. 12 S., R. 5 E.). Raveling of the loosening-limited
free face produces a basal debris apron at the angle of repose (33.5)
of the underlying obsidian sand and gravel deposit. After 20 yr, the
retreating fault scarp has been nearly buried by upward growth of the
basal debris apron.
T6 of the terrace scarp), indicating that an ear-
lier faulting event created a scarp prior to
terracing.
In 1959, an active cut-bank of the South Fork
Madison River provided a clear cross section of
fault scarp Fl - Fl ' , immediately west of profile
location Fl-11 (Fig. 3). Although the cut-bank
is now abandoned and the cross section com-
pletely buried, it was studied intensively by
Myers and Hamilton (1964, their Fig. 34). From
their observations of this exposure, they inferred
a sequence of events similar to that listed above.
Photos and diagrams of the exposure indicate
that the fault moved twice after the creation of
terrace T6-T6' : once in 1959 and once prior to
1959.
Profiles of terrace scarp T6-T6' and of fault
scarp Fl - Fl ' are shown in Figures 10a-lOd
(the location at vrhich each profile was mea-
sured is shown in Fig. 3). The same pattern of
degradation observed on the profiles of terrace
scarps in study site 1 was observed on profiles of
scarps T6-T6' and Fl - Fl ' : the gradient of the
midsection decreases and the initially sharp
basal concavity and crestal convexity become
more rounded. The original, nearly horizontal
crest and base may still be observed beyond the
limits of the crestal convexity and basal concav-
ity. The mean mi dsection slope angle of the
higher, T6' segment of the terrace scarp is 11.3,
significantly greater (see Table 2 below) than the
mean midsection slope angle of 14.5 deter-
mined for the lower T6 segment of the scarp.
The difference in angle for two hillslopes of the
same age again demonstrates the danger of using
midsection slope angle alone to determine rela-
tive ages of hillslopes.
After burial of the retreating free face by the
basal debris apron, the morphology of the F1
segment of fault scarp Fl - Fl ' presumably
showed the same proportion of single to multi-
a
b
M
c
I 1 1
0 10 20
Scal e (Meters)
(No Verti cal Exaggerati on)
ple scarp profiles as was observed along the fault
scarp that formed in 1959 south of the Duck
Creek Highway Maintenance Station (Figs.
8a-8c). Because the initial profile morphclogy is
more variable for fault scarps than for :errace
scarps, greater variability in the degraded profile
morphology would also be expected. This ex-
pectation is confirmed by comparing the stan-
dard deviation of the midsection slope angles of
profiles from T6 (0.9) and T6' (1.2) with that
of profiles from the F1 (2.8) segment cf fault
scarp Fl - Fl ' (Table 2).
Figure 8. Profiles of fault scarps south of
the Duck Creek Highway Maintenance Sta-
tion north of West Yellowstone, Montana.
All scarps were offset during the 1959 Yel-
lowstone earthquake but are superimposed
on a much older, much degraded scaip. (a)
Simple, single scarp. After 20 yr, the initial
scarp has been nearly buried by the basal de-
bris apron, (b) Multiple scarp consisting of
two smaller scarps, (c) Multiple scarp consist-
ing of three smaller scarps.
Approximately 50% of the lateral ext ent of
the recently formed fault scarp is a single,
simple scarp, and the other 50% is made up of
multiple scarps.
DATING OF SCARPS, WEST YELLOWSTONE, MONTANA 1419
30
SCALE (METERS)
(NO VERTI CAL EXAGGERATION)
F l - 3
0 SO 40
SCALE (METERS]
(NO VERTICAL EXAGGERATION)
Figure 9. Profiles of
terrace scarps cut by the
Madison River north of
West Yellowstone, Mon-
tana (location of each pro-
file is shown in Fig. 2).
Dots represent the profile
predicted by the hillslope
degradation model, equa-
tion 1, and from a simple
initial profile (Fig. 11)
using values for H, a, 3,
6, and tc given in Table 1.
(a) Modern cut-bank of
the Madison River, (b)
Terrace scarp T4. (c) Ter-
race scarp T3. (d) Terrace
scarp T2.
T 6 ' - 5
F l ' - B
F l ' - 5
Figure 10. Profiles of terrace scarp and fault scarp west of West Yellowstone, Montana
(location of each profile is shown in Fig. 3). Dots represent the profile predicted by the hillslope
degradation model, equation 1, and from a simple initial profile (Fig. 11) using values for H, a,
P, 0, and tc given in Table 2. (a) T6 segment of terrace scarp, (b) T6' segment of terrace scarp,
(c) F1 segment of fault scarp, (d) Fl ' segment of fault scarp.
MORPHOLOGIC DATING
OF HILLSLOPES
The general characteristics of the degraded
fault and terrace scarps in West Yellowstone,
including the decrease in curvature of the crestal
convexity and the basal concavity and the de-
crease in gradient of the midsection, are all con-
sistent with the predictions of a simple model of
hillslope degradation (Nash, 1980a, 1980b,
1981a, 1981b). Furthermore, as shown else-
where (Nash, 1980a), under some conditions the
morphology of a degraded scarp may be used to
date the scarp. By using the radiocarbon date
available for terrace scarp T4, the model can be
calibrated and then used to morphologically
date the other scarps.
MODEL OF
HILLSLOPE DEGRADATION
Morphologic dating is based on a simple
model of hillslope degradation that yields the
equation
dt dx
2
'
(1)
At any point on the hillslope, the rate at which
the elevation (j>) changes with time (/) is thus
linked by a constant of proportionality (c) with
the second derivative, in which x symbolizes the
horizontal coordinate of the point. Given a sim-
ple initial profile (Fig. 11a), the model predicts
that as time elapses, the curvatures of the crestal
convexity and the basal concavity must decrease
while their lateral extents increase, thereby low-
ering the gradient of the midsection (Fig. l i b).
Not only are these predictions consistent with
the shapes of the terrace scarps and the fault
scarps south of Hebgen Lake (the observed pro-
file and the profile predicted by the model are
displayed in Figs. 9 and 10), but they are also
consistent with the pattern of degradation ob-
served on abandoned wave-cut bluffs along the
shores of Lake Michigan (Nash, 1980b). This
model further explains the observation made by
Bucknam and Anderson (1979) that the midsec-
tion gradient of a degraded fault scarp and a
lacustrine wave-cut bluff near Drum Mountain,
Utah, increased with increasing height of the
scarp (Nash, 1980a).
The model is applicable only to transport-
limited hillslopes (slopes on which more loos-
ened debris is available for transport than the
transportational processes are capable of remov-
ing) (Gilbert, 1877; Nash, 1981a). Although the
model is inapplicable to loosening-limited hill-
slopes (slopes on which debris is removed as
rapidly as it is loosened from the surface), it is
nevertheless ideal for describing the degradation
of slopes underlain by unconsolidated, cohesion-
less sands and gravels on which the primary
1420 D. B. NASH
Midsection
Base
Figure 11. (a) Profile of
a scarp having a simple initial
morphology, consisting of a
straight base and crest at
a uniform slo|ie angle, 0, and
a straight midsection at a
slope angle if a + 6 (a is
termed the ''initial excess
midsection slo pe angle"). The
scarp offset, H, is the per-
pendicular distance between
the crest and base.
(b) With time, the curva-
ture of the crestal convexity
and the basal concavity of
a transport-limited slope is
reduced (becomes more
rounded). The slope angle of
the midsection, /?+ 0 (/3 is
termed the "degraded excess
midsection slope angle"), is
defined as the average slope
angle over the central 10% of the scarp offset. Note that although the sections of the crest and
the base immediately adjacent to the midsection are progressively consumed by outward
growth of the crestal convexity and the basal concavity, their initial slope angle, 0, and the
initial scarp offset, H, may still be measured from the remaining undisturbed, straight sections
(c = 1.5 x 10"
3
m
2
/yr, H = 1.2 m, t = 1,000 yr, a = 15, /3 = 11.8, 0 = 20).
(Meters)
(No Vertical Exaggeration)
debris-transporting mechanism is soil creep. The
model thus opens the way to morphologic dat-
ing of most fluv ial terrace scarps cut in alluvium
and of some sciirps produced by normal faulting
of alluvial fans.
PREDICTED RELATIONSHIP
BETWEEN HILLSLOPE AGE
AND MORPHOLOGY
Analysis of the hillslope degradation model
indicates that the rate at which the gradient of
the profile midsection decreases is related to c
(assumed to be independent of x and t) and to
the height and midsection gradient of the initial
profile. The following discussion of this relation-
ship will be limited to hillslopes underlain by
cohesionless materials and having a simple in-
itial profile, such as fluvial cut-banks, marine
and lacustrine wave-cut bluffs, and some scarps
produced by normal faulting.
The offset (H) of the initial hillslope profile is
defined as the perpendicular distance separating
the crest and base (Fig. 1 la). Even on a substan-
tially degraded hillslope, the initial offset may
still be determined by extending the profile
beyond the limits of the basal concavity and
crestal convexity to points at which the original
straight crest and base have not been altered
(Fig. l i b). The initial excess midsection slope
angle (a) is defined as the angle by which the
initial slope angle of the scarp midsection ex-
ceeds the slope angle (0) of the crest and base
(Fig. 11a). The original slope angle (a + 0) of
the midsection of a degraded scarp may be de-
termined either by measurement of the angle of
internal friction of the underlying material or by
observation of modern, actively forming, loosen-
ing-limited hillslopes underlain by the same
material. Assumed to be the same for crest and
base, the angle 0 may still be measured on a
degraded scarp by observing the profile upslope
from the crestal convexity or downslope from
the basal concavity. The degraded excess slope
angle of the hillslope midsection (/3) is deter-
mined by measuring the angle by which the ob-
served profile midsection exceeds 0. Simply
calculating /3 from the slope angle of the central
or steepest line segment of the profile midsection
is ill advised, because this angle will be a func-
tion of the segment length (the shorter the seg-
ment, the steeper j3 becomes). To avoid this
problem and other problems associated with
different-sized hillslopes, /} + 0 is here defined as
the average slope angle of the central 10% of the
profile midsection (Fig. l i b).
In an analysis of the hillslope degradation
model, Nash (1980a) found that the time (t) in
which the excess midsection slope angle de-
creases from a to j8 is proportional to the square
of the profile offset (ff). Further analysis of the
model shows that to any given value of tan
(/J)/tan(a), there corresponds a unique value of
the dimensionless quantity ( t c/ H
2
) tan
2
(a).
Thus any combination of values for c, H, t, and
a that yields the same value for {tc/H
2
) tan
2
(a)
necessarily results in identical values of tan
(/3)/tan(a) (Nash, 1981b). The morphologic
dating technique used in the present: study is
based on the plot of (tc/H
1
) tan
2
(a) versus tan
(/3)/tan(a), displayed in Figure 12, foi values of
0 < 10. From Figure 12, it is observed that the
decrease in the gradient of the midsec tion is in-
itially rapid until tan(/?)/tan(a) <0.2, after
which the rate of decrease is very slow. Very
small errors made in the measurement of a, /3,
or 0 will result in very large errors in the calcu-
lated value of tc, if tan(^)/tan(a) <0.2.
PROCEDURE FOR
MORPHOLOGIC DATING
To determine the age (/) of a degraded hill-
slope, it is necessary to have the appropriate
value for c. In all probability, c is a function of
climate, hillslope aspect, and the nature of the
underlying material. When c has been found for
hillslopes in a great many different areas, the
relation of c to climate, aspect, and material may
become so well established that c can be esti-
mated directly from these parameters. For the
present, however, c must be determined for each
new study area by examining a degraded hill-
slope of known age.
Only the final step of the following procedure
distinguishes the calculation of c when t is
known and from the calculation of t when c is
known:
Step 1: Determine H, /3, and 0 from a detailed
profile of the hillslope.
Step 2: Determine a either directly, by mea-
surement of actively forming, loosenirig-limited
hillslopes within the study area, or indirectly, by
measurement of the angle of internal friction (<f>)
of the material underlying the hillslope.
Step 3: Calculate tan(/J)/tan(a) and use it to
determine the corresponding value of ( t c/ H
2
)
tan
2
(a) from Figure 12.
Step 4: Calculate tc by multiplying the
{tc/H
2
) tan
2
(a) determined in Step 3 by the
fl
r2
/tan
2
(a) calculated from the vali.es for H
and a found in Steps 1 and 2, respectively.
Step 5: If t is known and c is to be calculated,
determine c by dividing the tc found in Step 4
by t
or
Step 5: If c is known and t is to be calculated,
determine t by dividing the tc found in Step 4
by c.
Though based on the same model of hillslope
degradation, the above procedure is superior in
many ways to the one proposed previously by
Nash (1980a). The above procedure does not
require a wide range of values for H and /}, it
does not require curve fitting, it is easy to carry
out, and it yields results amenable to descriptive
DATING OF SCARPS, WEST YELLOWSTONE, MONTANA 1421
o . s
V
8
'
0.5
<0 .
Figure 12. Relationship among initial excess mid-
section slope angle, a; degraded excess midsection
slope angle, /?; scarp offset, H; hillslope diffusivity,
c; and the age of the hillslope, t. This plot may be
used for dating hillslopes if c is known or may be
used for determining c if t is known. It is applicable
to hillslopes having a simple initial morphology if
the slope angle of the crest and base, 0, does not
exceed 10.
( t c / H ) t a n ( t t )
0 2 4
Scale (Meters)
(No Vertical Exaggeration)
Figure 13. (a) Degraded fault scarp having a simple initial profile
(c = 1.5 x lO"
3
m
2
/ yr, H= 4.2 m, t = 2,000 yr, a = 35, 0 = 28.6, 6 =
0).
(b) Degraded multiple fault scarp. Although the age, total scarp
offset, and the slope angle of each small scarp are the same as for the
simple scarp in Figure 4a, the degraded excess midsection slope angle,
P, is much lower. Multiple slopes are not amenable to morphologic
dating unless their initial morphologies are known in detail (/? = 22.1).
and testing statistics. The hillslopes profiled in
the present study were analyzed with a BASIC
program, SLOPEAGE. The program utilizes a
lookup table of values for tan(/3)/tan(a) and
(tc/H
2
) tan
2
(a) rather than the curve shown in
Figure 12.
2
Multiple fault scarps cannot be dated by the
above technique. As they degrade, the smaller
scarps are rapidly smoothed out and, after a rela-
tively short time, a degraded multiple scarp will
resemble a degraded single scarp. If both a single
and multiple scarp have identical values of a, c,
H, and t, 3 for the degraded multiple scarp
(Fig. 13b) will be considerably lower than it
would be for the degraded single scarp
(Fig. 13a). Due to this lower value for /3,
morphologic dating will yield too great an age
for a multiple scarp.
Selective sampling provides a possible ap-
proach to the multiple scarp problem. In the
vicinity of Hebgen Lake, -50% of the lateral
extent of scarps produced by the 1959 earth-
quake was observed to be multiple scarps. It is
not known whether a similar proportion of mul-
2
The program is written for use with the IBM PC
and is available from the author at no cost. Send a
stamped, self-addressed envelope for details.
Figure 14. (a) Degradation
of a simple fault scarp. Initial
profile is dashed line; profile
after 1,000 yr is dotted line;
and profile after 2,000 yr is
solid line (c = 1.5 x 10"
3
m
2
/yr, H = 4.2 m, a = 35,
final p = 28.6, 6 = 0).
(b) Degraded composite
fault scarp offset 1.1 m in-
itially (dashed line) and then
again by 3.1 m 1,000 yr later
(dotted line). After 2,000 yr
(solid line), the degraded
composite fault scarp appears
very similar to the scarp
formed by a single faulting
event (Fig. 5a). Although
both had the same initial ex-
cess midsection slope angle
and the same total offset, the
degraded excess slope angle,
/3, of the composite fault
scarp is steeper than the sin-
gle event fault scarp (final fi =
31.0).
0 4 8
Scale (Meters)
(No Vertical Exaggeration)
1422
tiple scarps characterized scarps that were pro-
duced by pre-1959 faulting, but that seems to be
a reasonable assumption. If, then, successive
profiles were made at fixed horizontal intervals
along the base of a fault scarp, the 50% of the
profiles with the lowest values of tc would pre-
sumably represent degraded multiple scarps, and
could be eliminated from the final analysis. This
selective sampling procedure is usable only if
one can make a reasonable estimate of the ratio
of single to multiple scarps. It cannot be as-
sumed, a priori, that this ratio will be 50% in
other areas, because the ratio is likely to be a
function of several parameters, including the na-
ture of the surface materials.
Morphologic dating of composite scarps also
yields erroneous ages. As in the case of multiple
scarps, a degraded! composite scarp will, after a
relatively short time, appear very similar to a
degraded simple scarp produced by a single
faulting event (Figs. 14a and 14b). No modifica-
tion of the presen t procedure is likely to yield
reliable dates for the individual faulting events
that produce a composite scarp. Hanks and
others (1984) proposed a method for calculat-
ing average offset rates, but different values of 3
may result from identical average offset rates
(Figs. 14a and 14b). In the present study,
evidence found in cross-sectional exposures of
the fault and from other independent observa-
tions permitted identification of composite
scarps and their elimination from the final
analysis.
Dating should not be attempted with a hill-
slope on which the angle of the crest and of the
base differ by more than a few degrees. The
technique is also inappropriate for use with hill-
slopes on which the rounding of the basal
concavity and of the crestal convexity differ sig-
nificantly. According to equation 1, hillslopes
having a simple initial morphology (Fig. 11a)
should show a symmetry in the degree of round-
ing of the crestal convexity and of the basal
concavity with time. Hillslopes with asymmetric
rounding cannot be modeled using equation 1
(profiles having a more rounded and extensive
basal concavity tfcian crestal convexity probably
developed by slojje wash and must be avoided).
ANALYSIS OF DATA
Calculation of c
For terrace sca.rp T4, Table 1 gives tc = 10.59
1.64 m
2
. Dividing this value by the best avail-
able age for that scarp, 1= 7,100 50 yr, yields c
= 2.00 0.24 x 10"
3
m
2
/yr. Because 7,100
50 yr is the minimum age for terrace T4, the
actual value of c may be smaller than calculated.
The age of a hillslope calculated with this value
D. B. NASH
TABLE 1. DATA DERIVED FROM PROFILES OF TERRACE SCARPS IN STUDY SITE 1
Profile Offset, H Average slope
(m) angle of crest
and base, 0
(degrees)
Present ( K/ H
2
) tan
2
(a> tc
midsection slope (m
2
)
angle, +B
(degrees)
T2- 1 3.41 2.1
T2- 2 3.76 0.1"
T2- 3 2.64 1.4
T2- 4 3.06 1.4
T2- 5 5.87 -0.9
T2- 6 6.10 -1.2
T2- 7 5.03 -0.4
T2- 8 4.52 0.8
T2- 9 5.71 -0.4
T2-10 3.97 1.8
Mean 4.41 0.5
S.D. 1.23(28%) 1.2 (240%)
T3- 1 4.39 1.0
T3- 2 5.40 0.0
T3- 3 5.20 -0.4
T3- 4 4.62 0.7
T3- 5 4.07 1.1
Mean 4.74 0.5
S.D. 0.55(12%) 0.6 (130%)
T4- 1 2.81 -0.7
T4- 2 3.10 -1.1
T4- 3 2.51 -0.7
Mean 2.81 0.8
S.D. 0.30(11%) 0.2 (25%)
13.3 0.714 22.5
14.0 0.523 16.9
13.1 0.688 12.3
11.7 0.906 21.5
16.1 0.356 26.1
15.6 0.376 29.1
11.6 0.753 42.2
14.5 0.509 25.1
17.7 0.293 21.0
16.8 0.380 15.7
14.4 23.2
2.1 (14%) 8.4 (36! d)
13.8 0.583 27.7
15.8 0.392 26.2
19.0 0.246 14.7
16.8 0.353 18.0
19.8 0.237 9.7
17.0 19.3
2.4 (14%) 7.6 (39%)
11.9 0.693 11.7
10.6 0.840 16.9
9.9 1.007 13.9
10.8 14.2
1.0 (9%) 2.6(181)
2 2
Note: calculation of (tc/H ) tan (a) and tc based on a+B = 33.5.
TABLE 2. DATA DERIVED FROM PROFILES OF TERRACE SCARPS AND FAULT SCARP IN STUDY SITE 2
Profile Offset, H Average slope
(m) angle of crest
and base, 6
(degrees)
Present (tc/H
2
) tan
2
(o) tc
midsection slope (m
2
)
angle, +B
(degrees)
T6-1 5.08 0.5
T6-2 5.18 -0.4
T6-3 3.70 0.9
Mean 4.65 0.3
Si ). 0.83(18%) 0.7 (233%)
T6'-l 8.02 -0.7
T6'-l 7.02 -0.2
T6'-3 7.65 -0.9
T6'-4 7.77 -0.6
T6'-5 7.79 -0.5
T6'-6 7.87 -0.7
Mean 7.69 -0.6
S.D. 0.35 (5%) 0.2 (33%)
Fl- 1 1.82 -0.4
Fl- 2 1.90 -0.1
Fl- 3 2.55 -0.8
Fl- 4 3.49 -1.7
Fl- 5 2.69 0.2
Fl- 6 3.16 -1.4
Fl- 7 3.45 -1.5
Fl- 8 2.11 0.1
Fl- 9 2.03 -0.3
Fl-10 1.82 0.0
Fl-11 1.40 1.4
Mean 2.40 -0.4
S.D. 0.71 (30%) 0.9 (225%)
10.8 0.974 59.4
10.8 0.874 52.1
12.3 0.758 25.4
11.3 45.6
0.9 (8%) 17.0 (33%)
13.2 0.557 77.3
13.3 0.572 63.0
14.6 0.442 55.2
15.1 0.419 54.7
14.3 0.476 63.1
16.3 0.350 47.2
14.5 60.1
1.2 (8%) 10.3 (17%)
6.0 2.814 20.6*
6.2 2.840 23.2*
7.5 1.698 23.9*
9.6 0.949 23.2*
17.4 0.316 5.2
17.7 0.280 5.8
9.5 0.990 24.0*
13.4 0.577 5.9
12.0 0.708 6.5
10.9 0.898 6.7
13.1 0.688 3.4
11.2 5.6
4.0 (36%) 1.2 C!l%)
Note: calculation of (tc/H
2
) tan
2
(a) and tc based on a*6* 33.5.
'Profile assumed to be that of a degraded multiple scarp and thus is deleted from the calculation of the mean tc for fault scarp Fl.
of c will thus be a minimum age. The above
value for c lies between c = 1.2 x 10~
2
m
2
/ yr for
abandoned wave-cut bluffs in the northern pr-
tion of the lower peninsula of Michigan (Nash,
1980b), and c = 4.4 x 10"
4
m
2
/ yr for a scarp
produced by normal faulting near Drum Moun-
tain, Utah (Nash, 1980a). This intermediate
value for c is not surprising, because the obsidian
sand and gravel deposit of the West Yellowstone
Basin is coarser grained than the sandy morainic
DATING OF SCARPS, WEST YELLOWSTONE, MONTANA 1423
1 * Terrace Scarp T6
_
Terrace Scarp T6'
2 5
- A1
3 . 4
A2
_
6
I I
* 3
I I I
TABLE 3. MORPHOLOGICALLY DETERMINED AGES OF
TERRACE AND FAULT SCARPS
Scarp Age of scarp
(yr)
T2 11,600 6,400(55%)
T3 9,600 5,600 (58%)
T6 22,800 13,000 (57%)
T6' 30,000 10,800 (36%)
F1 2,800 1,100(40%)
Note: ages calculated using c value determined from analysis of terrace scarp
T4 (c = tc/c = 14.2 2.6 m
2
/7,100 50 yr = 2.00 0.38 * I0"
3
m
2
/yr).
material underlying the hillslopes of Michigan,
but it is finer grained than the alluvium underly-
ing the Drum Mountain scarp. The climate of
the West Yellowstone area is similarly interme-
diate between the humid temperature climate of
Michigan and the semi-arid climate of north-
central Utah.
The coefficient c is likely to be a function of
climate, underlying material, and aspect. Be-
cause the climate and underlying material are
uniform throughout the study area, and (except
for the west-facing terrace scarp T6-T6' ) be-
cause the aspect of all of the studied scarps is to
the north, the value of c calculated for terrace T4
is probably appropriate for all of the scarps ex-
cept T6-T6' . It is difficult to determine how
significantly Holocene changes in climate af-
fected c; a difference of two orders of magnitude
between the c values derived from Michigan and
Utah suggests that c may be strongly affected by
climate. Pleistocene climate differed enough
from Holocene climate to make suspect mor-
phologically determined ages of > 10,000 yr for
hillslopes.
Morphologic Dating of Scarps in the
West Yellowstone Basin
For terrace scarps T2 and T3, the mean
values of tc are 23.2 m
2
and 19.3 m
2
, respec-
tively (Table 1). If c is assumed to be equal for
all of the scarps, then tc reflects their relative
ages. The values for tc calculated for the terrace
scarps accord well with their relative ages, as
inferred from their relative positions (Fig. 2). It
is encouraging that morphologic dating duly in-
dicates that T2 is older than T3, and that T3 is
older than T4, despite the fact that the mean
midsection slope angles of T2 and T3 (16.3 and
17.5, respectively) are both steeper than T4
(10.8). Using c = 2.00 x 10"
3
m
2
/yr, terrace
scarps T2 and T3 are dated at 11,600 and 9,600
yr B.P., respectively (Table 1). As the date for
T2 is pre-Holocene, its accuracy is questionable.
The mean values of tc for terrace scarps T6
and T6' are 45.6 m
2
and 60.1 m
2
, respectively
(Table 2). These two scarps are segments of the
80
Figure 15. Hillslope aspect
and tc calculated for the T6 and
7 0
T6' segments of terrace scarp
T6-T6'. There appears to be a
6 0
poorly defined tendency for hill-
slopes with southern aspects p
to have higher values of tc. 5 50
The age of the scarp, t, is "
the same for all profiles; thus,
changes in c with hillslope
4 0
aspect may account for some of
the observed trend of decreasing
3 Q
tc as scarp aspect changes from
southerly to westerly.
same terrace scarp (Fig. 3) and so they must be
of the same age and have similar values for tc.
The observed large difference in tc values results
primarily from the anomalously low value for
profile T6-3. Taking c = 2.00 x 10~
3
m
2
/ yr
yields an age of 22,800 yr for T6 and 30,000 yr
for T6' (Table 3). These dates are Pleistocene
and also are questionable.
K. L. Pierce (1983, personal commun.) found
a strong correlation between hillslope aspect and
c for terrace scarps on incised alluvial fans.
Some local confirmation arises from the mean-
der of terrace scarp T6-T6' , which varies the
aspect of the scarp along its lateral extent. For
the profiles collected from the T6 and T6' seg-
ments of the terrace scarp, a plot of scarp aspect
against tc suggests that tc decreases from a max-
imum for profiles with a southerly aspect to pro-
files with a westerly aspect (Fig. 15). With the
age of the scarp (t) necessarily the same for all
profiles, the observed difference in tc may be due
in part to differences in c. Presumably, c reaches
a minimum on scarp with a northerly aspect.
The value of c appropriate for the west-facing
terrace scarp T6-T6' is likely to be significantly
greater than that calculated for the north-facing
terrace scarp T4. Such a trend would be consist-
ent with several studies of valley asymmetry
summarized by Young (1972). These studies in-
dicate that for east-west-trending valleys in the
northern hemisphere, the valley wall with a
northerly aspect is generally steeper than the
wall with a southerly aspect.
The pre-1959 profile of a fault scarp may be
restored by removing from the profile the still-
fresh small scarps formed in 1959. If this re-
stored scarp profile is not composite (if it was
formed by a single pre-1959 faulting event), it is
suitable for morphologic dating. It is assumed
that approximately one-half of the lateral extent
180 210 240 270 300 330
Azimuth of hillslope aspect ()
of a scarp, when produced by normal faulting of
cohesionless material, will consist of multiple
scarps. In that case, if calculated from profiles
taken at evenly spaced intervals along the scarp
face, about one-half of the dates will overesti-
mate the actual age of the fault scarp. Of the
eleven profiles collected from the F1 segment of
fault scarp Fl - Fl ' , the six yielding the highest
values of tc are presumably degraded multiple
scarps. With elimination of these six profiles, a
mean tc = 5.6 m
2
is calculated from the remain-
ing five profiles. Use of c = 2.00 x 10~
3
m
2
/yr
then gives a date of 2800 1100 yr B.P. for the
pre-1959 faulting that produced the F1 segment
of fault scarp Fl - Fl ' .
CONCLUSIONS
Morphologic dating of fault scarps presents
some special difficulties that were not fully ap-
preciated prior to this study. Morphologic dating
will result in an erroneously old age if the initial
scarp is multiple: consisting of two or more
smaller scarps. Moreover, the selective sampling
procedure suggested here for dealing with mul-
tiple scarps must not be used indiscriminately in
other areas. Composite scarps, formed by recur-
rent movement along the same fault, also are not
amenable to morphologic dating and should be
avoided. Given the problems presented by mul-
tiple and composite fault scarps, an indirect ap-
proach may, in the long run, offer simpler and
more accurate morphologic dating. When dat-
ing scarps produced by normal faulting of allu-
vial fans, it may be preferable to date the
upstream terrace scarps formed by fan-head en-
trenchment rather than to attempt to date the
fault scarp itself.
With only one independently dated hillslope,
terrace scarp T4, a quantitative assessment of the
1424 D. B. NASH
accuracy of morphologic dating cannot be
made. The present study does, however, provide
qualitative evidence for its accuracy. Despite the
fact that the slope angle of the youngest scarp is
significantly less than that of the older scarps, the
morphologic dating procedure yields, for the ter-
race scarps in study site 1, ages consistent with
the known relative ages of the terraces. Since its
formation, terrace scarp T6-T6' has been offset
by the faulting that produced the F1 segment of
fault scarp Fl - Fl ' . The terrace scarp thus pre-
dates the F1 segment of the fault and, as before,
the morphologic dates for the terrace scarp and
fault scarp are in accord with their known
relative ages.
The morphologic dating technique used here
goes beyond Niish (1980a) by developing a
procedure that is simpler to use, is more widely
applicable, is more accurate, and is operable
with a smaller number of scarp profiles. The
modified technique provides a rapid and inex-
pensive method far dating hillslopes that are not
amenable to dating by other means. Although it
must be used with much caution and is not ap-
propriate for use on all hillslopes, morphologic
dating offers a valuable tool for determining pa-
leoseismic chronology and for assessing regional
seismic risks.
ACKNOWLEDGMENTS
This work was sponsored by the U.S. Geolog-
ical Survey, Earthquake Hazards Reduction
Program, under Contract 14-03-0001-19109. I
thank Ralph O. Meyer, District Ranger, and his
staff at the Hebgen Lake Ranger District of the
Gallatin National Forest for their co-operation
and help. Robert E. Wallace (U.S. Geol. Survey,
Menlo Park) made many helpful suggestions
during two productive visits to the field area.
Kenneth L. Pierce (U.S. Geol. Survey, Denver)
is thanked for his help with the field work; for
finding material for radiocarbon dating; for hav-
ing that material analyzed by the Radiocarbon
Laboratory of the U.S. Geological Survey; and
for his interest, encouragement, help, and enthu-
siasm. I am particularly grateful to my two field
assistants: Michael J. Bolton, who provided am-
iable and capable assistance for three grueling
months of field work during the summer of
1980; and my father, Leonard K. Nash, who did
yeoman's service as an unpaid field assistant dur-
ing the financially strained but productive
summer of 1981. M. A. Carson, Arvid M. John-
son, and Leonard K. Nash are thanked for their
helpful criticisms of this manuscript.
REFERENCES CITED
Bucknam, R. G, and Anderson, R. E., 1979, Estimation of fault-scarp ages
from a scarp-height-slope-angle relationship: Geology, v. 7, p. 11-14.
Carson, M. A., 1977, Angles of repose, angles of shearing resistance and a igles
of talus slopes: Earth Surface Processes, v. 2, p. 363-380.
Eardley, A. J., 1960, Phases of orogeny in the deformed belt of southwestern
Montana and adjacent areas of Idaho and Wyoming, in Campau, I). K .
and Anisgard, H. W., cds., Billings Geological Society 11th annual field
conference, West Yellowstone-earthquake area: p. 92-105.
Gilbert, G. K., 1877, Geology of the Henry Mountains: U.S. Geographical and
Geological Survey, 160 p.
Hamilton, W., 1960, Late Cenozoic tectonism and volcanism of the Yellow-
stone region, Wyoming, Montana, and Idaho, in Campau, D. E., and
Anisgard, H. W., eds.. Billings Geological Society 11th annual field
conference, West Yellowstone-earthquake area: p. 86-91.
Hanks, T. C Bucknam, R. C., Lajoie, K. R and Wallace, R. E 1984,
Modification of wave-cut and faulting-controlled landforms: Journal of
Geophysical Research, v. 89, no. B7, p. 5771-5790.
Mayer, L., 1982, Quantitative tectonic geomorphology with application to
northwest Arizona [Ph.D. dissert.]: Tucson, Arizona, Arizona Univer-
sity, Department of Geosciences, 512 p.
Myers, W. B., and Hamilton, W., 1964, Deformation accompanying the Heb-
gen Lake earthquake of August 17, 1959: U.S. Geologica Survey Pro-
fessional Paper 435-1, p. 353-360.
Nash, D. B., 1980a, Morphologic dating of degraded normal fault scarps:
Journal of Geology, v. 88, p. 353-360.
1980b, Forms of bluffs degraded for different lengths of time in Emmet
County, Michigan, USA: Earth Surface Processes, v. 5, p. 331-345.
1981a, FAULT: A FORTRAN program for modeling tht degradation
of active normal fault scarps: Computers and Geostience, v. 7,
p. 249-266.
1981b, Fault scarp morphology: Indicator of paleoseismit chronology:
U.S. Geological Survey Final Technical Report, Contiact Number
14-08-0001-19109,132 p.
Pierce, K.L.,1979,Historyand dynamics of glaciation in the nort lern Yellow-
stone National Park area: U.S. Geological Survey Professional Paper
729-F, 90 p.
Pierce, K. L., Obradovich, J. D., and Friedman, I., 1976, Obsiditn hydration
dating and correlation of Bull Lake and Pinedale glaciatio is near West
Yellowstone, Montana: Geological Society of America Bulletin, v. 87,
p. 703-710.
Porter, S. C., Pierce, K. L., and Hamilton, T. D., 1983, Late Wiscansin moun-
tain glaciation in the western United States, in Porter, S. C., ed., The late
Pleistocene: Minneapolis, Minnesota, University of Minnesota Press,
p. 71-111.
Sieh, K. E., 1978, Pre-historic large earthquakes produced by slip on the San
Andreas fault at Pallett Creek, California: Journal of Geophysical Re-
search, v. 83, p. 3907-3939.
Smith, R. B., Shuey, R. T Pelton, J. R and Bailey, J. P., 1977, Yellowstone
hot spot: Contemporary tectonics and crustal properties from earth-
quake and aeromagnetic data: Journal of Geophysical Research, v. 82,
p. 3665-3676.
Waldrop, H. A., 1975, Surftcial geologic map of the West Yellowstone quad-
rangle, Yellowstone National Park and adjoining areas Montana,
Wyoming, and Idaho: U.S. Geological Survey Miscellaneous Geologic
Investigations Map 1-648, scale 1:62,500.
Wallace, R. E., 1977, Profiles and ages of young fault scarps, north-central
Nevada: Geological Society of America Bulletin, v. 88, p. 1267-1281.
1980, Degradation of the Hebgen Lake fault scarps of 19i 9: Geology,
v. 8, p. 225-229.
Witkind, I. J., 1964, Events on the night of August 17, 1959The human
story: U.S. Geological Survey Professional Paper 435-A, p. 1-4.
1974, Some details of the Hebgen Lake, Montana, earthq lake of Au-
gust 17, 1959, in Voight, B., ed., Rock mechanics: Th: American
northwest, 3rd Congress Expedition Guide: University Pa*k, Pennsyl-
vania, Experiment College of Earth and Mineral Sciences, 1 lie Pennsyl-
vania State University Special Publication, p. 126-133.
Young, A., 1972, Slopes: Edinburgh, Scotland, Oliver and Boyd, p. 288.
MANUSCRI PT RECEI VED BV THE SOCIETY J UNE 17, 1 9 8 3
REVI SED MANUSCRI PT RECEIVED APRI L 5 , 1 9 8 4
MANUSCRI PT ACCEPTED APRI L 17, 1 9 8 4
Printed n U.S.A.

Das könnte Ihnen auch gefallen