Gediz Earthquake: 1970

Geophys. J. R.astr. SOC.(1972) 30, 229-252.
Seismotectonic Aspects of the Gediz, Turkey, Earthquake
of March 1970
N. N. Arnbraseys and J. S. Tchalenko
(Received 1972 October4)
Summary
In 1970 March 28, Asia Minor was shaken by a severe earthquake of
magnitude 7. The shock affected the region of Gediz where 1100people
were killed. The earthquake was associatedwith 45 km of surfacefaulting,
most of which occurred on pre-existing faults and zones of weakness. The
sense of movement was predominantly normal, with a relatively small
horizontalcomponent. Maximumrelativedisplacementsof 220 cmvertical
and 30 cm horizontal were measured. The overall regional deformation
seems to be an extension in the north-east-south-west direction. The
earthquake was preceded and followed by a long series of shocks.
Introduction
In 1970 March 28, at 21 h 02min GMT, Asia Minor was shaken by a severe
earthquake of magnitude 7. The shock affected the Province of Kiitahya, Turkey,
particularly the region of Gediz, and it was felt over an area of about one million
square kilometres. The macroseismic effects in the epicentral region were severe;
within an area of about 3000km' about 15000 housing units were destroyed or
damaged beyond repair. The main shock and its numerous and comparatively large
aftershocks killed 1088 and injured 1290 people. Much of the damage was due to
landslides, aftershock activity and to fires started soon after the earthquake. The total
damage is estimated at about f10 million.
The area affected by the earthquake is shown on Fig. 1. The authors were in the
Gediz area during April 1970investigatingthe seismotectonicand engineeringaspects
of the earthquake, as well as late in 1970and in the spring of 1972.* They made no
attempt to assess intensities in the epicentral area. As stated elsewhere, the authors
doubt the existanceof isoseismalsand the valueof Intensity Scaleswithin the epicentral
region of a strong earthquake, (Zhtopek & Ambraseys 1969; Ambraseys 1969). In
the case of the Gediz earthquake, as in other cases, (Javaheri 1970),no two intensity
maps prepared by different field parties looked remotely alike, as can be seen by
comparing the intensity maps of the Gediz earthquake prepared by Aytun &
Tasdemiroglu 1970; Uz & Gii~lii1970; Yarar et al. 1970; Uzsoy & Celebi 1970;
Tezcan & Ipek'l971. The maximum intensity assigned to the main shock by these
writers is VIZI-IX (MM).
As a result of the earthquake, major faults on an east-west and to a lesser extent,
north-south direction, were reactivated, Fig. 2. The dominant motion on both fault
systems was dipslip, with the block north and east of the reactivated system down-
thrown by a few tens of centimetres to 2.2m.
* The effects of the Gediz earthquake on man-made structures and the purely geological and
engineeringaspects of the event are omitted from this paper; they will appear elsewhere.
229
1
byguestonJune29,2014http://gji.oxfordjournals.org/Downloadedfrom

230 N. N. Ambraseys and J. S. Tchalenko
--
FIG.1. Location map.
Seismicity of the Gediz area
Gediz is situated in an area of complex block faulting, typical of Asia Minor, with
no well-defined tectonic pattern. For example, the Izmir sheet of the 1:500000-scale
Geological Map of Turkey (MTA 1964),reflects very little, if any tectonic background
and other geological maps (CTIA 1968; CTIE 1964; Ketin 1964, 1967, 1970) are too
general and on a scale too small for the purpose intended. Published geological
reports on specificregions, such as on the region between Kutahya and Gediz (Akkus
1962), on the Simav depression (Zeschke 1954), on the western Taurus (Graciansky
1966)and the references given by Dubertret (1970) for Asia Minor, are valuable and
of considerable interest but the synthesis of all these reports into a comprehensive
tectonic model of Asia Minor is still lacking.
Certain aspects of the tectonics of the Gediz region have been studied recently
(Abdiisselamoglu 1970;Ketin 1970;TaSdemiroglu 1971)and they will not be discussed
here. The reader is referred, however, to the geological map of Asia Minor (MTA
1964;Dubertret 1970)and to the sketch given in Fig. 3. Part of the region considered
in this paper has been studied in detail by L. Dvbertret, to whom we are grateful for
comments on the geology of the greater Gediz area.
The dominant faults in the area are the Simav fault in the west and the Muratdag
fault-system in the east. The former fault, which shows a large vertical offset,separates
the subsided Upper Simavand Hamamsuvu valleysfrom the metamorphic Demircidag
massif. This fault continues, in an east-south-east direction, under the Neogene cover
of the Hamamsuvu and possibly further east, under the Murat valley; its trace is in
fact marked by the Ilica and Gazino thermal springs.
The Muratdag region is crossed by two main faults which converge east of Gediz.
They separate the subsided TavSandag wedge, which is covered largely by ophiolitic
rocks, from the metamorphic block of Muratdai: in the south and from the TavSandag
in the north. It does not appear that these two faults continue beyond the vicinity of
Gediz.

Seismotectooicaspects of the Gediz, Turkey 231
ri
I
f
..'

232 N. N.Ambraseys and J. S. Tchalenko
29OOO' 29030' 30000'
'4u,(Eorthquake)
Fault 0 T h e r m a l spring
Km
---
0 10 20 30 (Geologic1
FIG.3. Sketch of general tectonic features of the Gediz region.
Many of the rivers in the Gediz region run along north-south trending basins,
filled with upper Tertiary lacustrine deposits and Quaternary alluvium. Between these
riversone findsthe granitodioritic Eiigozdag and the metamorphic massifsof Kocadai:
and Saphanedag. To our knowledge, the latter two have not been mapped in any
detail, but the earthquake brought to light at least one important fault zone running
south-south-east,along the eastern slope of the Saphanedag, i.e. the Agikpaga-
Muhipler fault break, Figs 2, 4 and 5.
The town of Gediz is located at the intersection of the Simav-Muratdag fault
system with the north-south trending Quaternary alignment of the Emet and Murat
river basins. A line of andesitic and basaltic volcanic plugs extends from Gediz to the
west along the southern periphery of the Saphanedag.
The Gediz earthquake occurred in Central Asia Minor were destructive earth-
quakes are known to have caused damage intermittently for 24 centuries (Ambraseys
1971). Like most parts of Asia Minor, the region affected by the earthquake is
predominantly a complex of young block faulting, vestiges of which can be found in
the surface faulting associated with the earthquakes of April 1896 (Agamennone
1900), September 1899 (Schaffer 1900), June 1944 (Foucht & Pinar 1944; Press
Reports 1944), and with earlier events (Hamilton 1837).
Of these earthquakes, the shock of 1944is of particular interest as it occurred in
the immediatevicinityof Gediz, Fig. 6. The magnitudeof thisearthquakewas about 6,
and the shock was felt within a radius of 230 km; it was followed by many aftershocks
of which at least one showed a magnitude of 53. The instrumental location of the
epicentreof these events is in error and the dataavailable today are totallyinadequate
for analysis. The lack of information about earthquakes prior to 1960is probably in
no way unusual for seismic events in Asia Minor. Much of the information about
the 1944 earthquake was gathered from press reports and from interviews with the
inhabitants of the region. From this information it appears that the 1944earthquake

Seismotectonicaspectsof the Gediz, Turkey 233
I
2970' 2Y30'2970'
I
FIG.6. Damage and faulting associated with the earthquake of 1944 June 25.
caused heavy damage to the extreme western part of the epicentral area of the 1970
earthquake and also faulting which extended intermittently from Ayvak to Muhipler,
Pinarbasi, Yelki and Sazak, Fig. 6. Part of these ruptures were reactivated during the
Gediz earthquake and they were still visible in the spring of 1972. The villages of
Celtikci, Erdogmus and Pinarbasi were almost totally ruined and they were rebuilt
some time after the 1944earthquake.
Foreshock and aftershock sequences of the Gediz earthquake
An interesting feature of the Gediz earthquake is that this shock was the largest
event in a series of earthquakes in Central Asia Minor which began almost a year
earlier and continued for more than a year after the earthquake. Fig. 7 shows the
distribution of epicentres for the period March 22 1969-1971 May 11. During this
period, the seismological networks of the Kandilli Observatory and of the Institute
of Earth Sciences at the University of Istanbul recorded more than 6000 aftershocks
of all magnitudes. A considerable number of shocks of magnitude greater than 3$
were also recorded by the network of the National Observatory in Athens. The
location of about 450 shocks of magnitude over 4 has been determined by the National
Oceanic and Atmospheric Administration, the Bureau Central Internationale de
Seismologie, as well as by local and combined networks of Turkey and Greece. The
best located epicentres in Fig. 7 may be 5-1 5 km in error. Epicentral determinations
based on less than four stations or with residuals greater than 30 km were not used
(Uz & Gii~lii1970; Ergin, Uz & Guclii 1972). All shocks seem to be superficial and
most of them have been assigned to depths between 10and 25 km; a large part of them,
especiallyin the early part of the series, are even shallower. Magnitudes determined
by various stations vary as usual, and for the smaller events, ML values given by
Athens have been adopted.
The Demirci sequence
The first shock of the series occurred near Demirci on 1969 March 23. It was a
shallow shock of magnitude 5+ which, onlMarch 22 was preceded by a foreshock of
magnitude 43, and followed by a series of relatively small events, two of which showed
M = 5 and 53 respectively. The Demirci earthquake caused considerable damage,

234 N. N. Ambraseys and J. S. Tchalenlco
b
% E
x ‘
3

Seismotectonic aspects of the Gediz, Turkey 235
destroying 1700 houses in the Demirci region without casualties. There is some
unauthenticated evidence that the main shock was associated with ground ruptures
near Sevciler, which extended for a few kilometres striking N-30" W up the Akdag
mountains (Press reports, Can 1969; Ketin & Abdiisselamoglu 1969).
The Ala~ehirsequence
A few days later, on the 1969 March 28,a large shock ofmagnitude 6occurred near
Alagehir, about 60 km south of Demirci, Fig. 7, starting a new and very long sequence
of low magnitude shocks which, with intermissions, lasted for almost a whole year.
Two comparatively large aftershocks near Demirci in April and October 1969 were
followed by many small ones from the same region, continuing for months. According
to local reports, at least three swarms of small shocks occurred during that period near
Demirci, each swarm lasting a few days.
The Alagehir earthquake and its aftershocks caused considerable damage in the
Gediz River Valley and particularly in the smaller valleys to the north of the Gediz
River where 40people were killed and a few thousand houses destroyed. Slumping of
the ground in the Valley and landslides on higher ground were responsible for some of
the damage. The maximum intensity reported for the main shock did not exceed
VII (MM), (Aytun 1971). Some of the later aftershocks of the Alagehir sequence
caused much damage with casualties in the Demirci region as well as in the Simav
Valley (Can 1969; Ketin & Abdiisselamoglu 1969).
The Demirkoprii dam, a 77-m high earthfill construction completed in 1960, and
located 15km from the instrumental epicentre of the main shock, suffered no damage.
Minor slumping and accelerated settlements of the fill were noticed. It is not known
whether in 1961the impounding of the 165-million-cubic-metresreservoir (1600mill.m3
maximum volume) triggered any shocks.
As a result of the Alagehir earthquake, a series of ground ruptures developed on
either side of the Valley near Alagehir. All of them are purely tensional, still visible,
suggestinglarge scale slumping of the Valley slopes. The longest and most continuous
rupture, extends for 12km from a point 2 km from Alagehir on the road to Kula, to
near Kurudere, Fig. 7. This rupture, probably of tectonic origin, crossestheAlasehir-
Salihli railway tracks which were damaged and the service was disrupted after the
earthquake. The origin of these and other ruptures further west, is not clear. Arpat
& Bingol (1969) imply that the motion was in fact purely tensional and that the
generic cause was the tectonic subsidence of the Valley floor. A cursory inspection of
the area early in 1972, however, suggests that some of these newly-formed ruptures
originated through settling and slumping of the alluvium on the flanks of the low hills
bordering the Valley, while other ruptures, beyond those already reported by Arpat
& Bingol (1969), were of tectonic origin.
The Gediz sequence
The largest shock of the series, M = 7, ociurred almost exactly one year after the
Alasehir earthquake, on 1970March 28. The epicentre is reasonably well located and
lies to the east of the area of the preceding aftershock activity, about 20 km north of
Gediz, Fig. 7.
More than 350 aftershocks of the Gediz sequence, of magnitude equal to or greater
than 3$, were recorded through February 1971and located. About three quarters of
the epicentres, including those of the larger aftershocks, can be associated with the
east-west trending fault system and with the central part of the downthrown block.
The larger and more numerous aftershocks occurred a few kilometres to the north of
the east-west trending surface ruptures and of their extension towards Muratdag in
the east, and towards Simavin the west. A smaller number of shocks occurred on the
northern extension of the north-south trending line of surface ruptures.

236 N.N.Ambraseysand J. S.Tchalenko
During the year following the Gediz earthquake, aftershocks continued inter-
mittently; large ones of magnitude between 5 and 6, being followed by many smaller
ones. All aftershocks for the period March 1970-March 1971 fall within a zone
40-50 km wide and 300 km long, trending west-north-west from the area of the main
shock near Gediz on the east, through the Demirci aftershock region, to Edremit on
the west, Fig. 7. Initially, aftershocks in the western part of the zone, near Edremit
and BigadiG, were few and rather weak. Near the close of the Gediz sequence, they
became numerous and strong, a shock of magnitude 5+ on 1971 February 23 and its
aftershocks near Edremit, marking the end of the sequence.
Since 1944major crustal strains have not been relieved by earthquakes in Central
Asia Minor as would normally be expected in a highly active region. Moderate
shocks have originated at relatively unexpected points, but this may be due to errors
in the location of epicentres prior to 1960. The 1969-1971 activity, being one of the
very few instances where the data is at all adequate for analysis,provided the opportu-
nity to study, in a very cursory manner, the developmentand pattern of seismicactivity
in Central Asia Minor. A series of strain energy release maps were prepared, starting
with the Demirci shock, in a manner somewhat similar to that described originally by
BAth (1953) and later by St Amand (1956), i.e. by contouring the normalized sum of
the square roots of the energies of all shocks occurring in the region for different time
periods.
A comparison of the strain energy release maps shows that the earthquake of
1969 March 3 effectively marks the beginning of a systematic strain release process in
Central Asia Minor. With time, the readjustment of strain, progressing gradually
outward from Demirci, delineates an west-north-west trending zone of strain relief,
not apparent from the tectonics of the region, extending from Gediz to Edremit.
Fig. 8 shows one of the strain release maps for a later stage of the Gediz sequence
from which the main shock of 1970 March 28 has been excluded. The effect of
excluding the main shock is apparent in the conspicuously low value of strain in the
epicentral area and in the area contained by the surface ruptures where, apparently,
strains were relieved through faulting. The strain released by the aftershocks alone,
which is comparable to that released by the main shock, follows a pattern very similar
to that of the stress concentration at the tips of a crack. The relatively low strain
release in the region between Saphane and Simav, may be due to a much earlier
relaxation brought about by the 1944activity.
Ground deformation accompanying the Gediz earthquake
The account given in the Appendix is a brief description of the ground effects
associated with the Gediz earthquake. It supplements the information already
published by Arsovski (1970),Ambraseys &Tchalenko (1970), Aytun &Tasdemiroglu
(1970) and Tasdemiroglu (1971) and it is given here in greater detail in order to put
on record facts that may facilitate further field work in this part of Turkey, Figs 9
and 10.
The ground fractures associated with the Gediz earthquake may be grouped into
three distinct types according to their relation with pre-existing faults and with the
terrain in general.
TypeA fractures were formed by re-activation of fault zonesin hard rocks, Fig. 11.
The rocks involved were serpentines (Pinarbasi, Giimiilekoy, Sazkoy, Hamamlar),
basalts and tuffs (AkGaalan) and metamorphic limestones (Asikpasa). The surface
effects varied from severe shattering and disordered fissuring of the rocks, Fig. 5, to
the creation of clean scarps.
Type B fractures were as type A, but they formed in regions where the rock was
overlain by later Tertiary and Quaternary sediments (Giimiilekoy,Sazkoy, Pinarbasi)
or by weathered slope material (Asikpasa), Figs 4 and 5. On sloping ground the fault

FIG.8.StrainreleasemapoftheGedizaftershocksequence:contoursin109(erg)*.

FIG. I I . Striations on andesite outcrop caused by relative motion of tuffs:
looking north-east.

scarp was generally accompanied by secondary fractures which formed subsidence
zones and gravity grabens of the type described by Slemmons (1957). On flat ground
the surface fractures formed elongated grabens of comparatively small throws or
en echelon series of open ruptures. At first glance this en echelon pattern seemed to be
evidencefor strike-slip fault displacement. However, the detailedmapping of a number
of B type fractures on flat ground, such as those shown in Fig. 12,demonstrated that
invariably fissures were arranged in a zig-zag pattern rather than en echelon and that
all of them were open showing no preferential lateral displacement on a large scale.
A likely explanation is that zig-zag fractures opened directly; they cannot be explained
in terms of strike-slip displacement, for such motion would necessarilyhave produced
compression features or at least significant changes in the opening of these fractures
along their trace.
Type Cfractures formed primarily in sub-horizontally bedded Neogene marls with
intercalated limestone beds, running uninterrupted as single fractures from a few
hundred metres to three kilometres (Odacamligi, Muhipler, Tuzluburun, Nennikiri),
Fig. 13.
Fig. 14 shows a map and a diagrammatic profile of some of these ruptures. In
most cases the ruptures are double, running in pairs parallel to each other, a few
metres to 300 m apart, with the block in between them downthrown by a few tens of
centimetres to over 2 m. Invariably, they run along the sides of a hill, the flanks of
which were raised in a sense opposite to the direction of slope, with the hill crest
downthrown. The floor of the graben thus formed is broken up into long, narrow
strips, traversing the whole length of the crest and both end sides of the hill. At first
sight, these factures seemed to be the upper boundaries of large-scale double slides, of
the type shown in the lower part of Fig. 14. A search on the flanks of the hills,
however, showed no evidence of emerging landslide toes; nor was there any evidence
of the upper boundaries turning in the same direction, with opposite sense of lateral
motion. Therefore, double landsliding or spreading of the hills seems improbable.
At a time it was thought probably that C-type fractures were due to an oblique
rather than vertical fault movement, of the type described by Tarr & Martin (1912)
in connection with the Yakutat earthquake of 1899, brought to our attention by
A. Grantz. However, further examination of these ruptures showed no signs of
scissoring or apparent reversal of throw along their trace. Consequently, oblique
movement also seems improbable.
It is perhaps significant that a closer examination of exposures of the Nennikiri
C-type fractures, brought to view in road excavations, showed that two of them werein
fact pre-existing normal faults in the mark. This coincidence is so striking that it
seems probable that all C-type fractures are in fact the result of normal faulting of
the underlying bedrock rather than of large-scale landsliding or strike-slip motion.
It is not difficult to show that normal faulting along a concealed bedrock fault may
produce a graben structure in the overlying overburden which may exhibit C-type
features with throws greater than those in the bedrock. Consider an overburden of
thickness h overlying bedrock which is subjected to normal faulting on a fault plate
dipping at an angle 4. If the angle of active failure in the overburden is b and the
throw due to normal faulting in bedrock d, then the throw in the overburden would
be of the order of
D = d/2tan($). tan(b).
Assuming that the value of (d/h)is negligeably small, the width ofthe graben structure
will be of the order of
L = 212. tan(b).
Thus for b = 10" and 4 = 45", the throw in the overburden will be almost three
times greater than that in bedrock. For a steeper bedrock fault, say 4 = 70°,both

FIG.13. Fault-scarp in marls, north-east branch of Tuzluburun fractures; looking
south-east.
[Focirig p. 240

29.22' 2Q623' 29024'E
a,P
Y
Flo.14. Map and diagrammatic profile of type C ruptures.

throws, in bedrock and overburden, are equal and the width of the graben at the
surface is about one third of the thickness of the overburden.
Although the tectonic origin of type C fractures cannot be established beyond
doubt, their attitude and location fit a pattern of regional extension. This pattern is
the same for the motions observed on the main north-west-south-east Asikpasa-
Muhipler and the east-west trending Erdogmus-Hamamlar ruptures. The dominant
motion on both major tectonic ruptures is dip-slip, with the Gediz block downthrown
along planes dipping between 60"and 70" to the north-east in the first case and to the
north in the second, (see Appendix). The dips that can be deduced for these two main
ruptures from surface evidence can hardly extend to a great depth; they do suggest,
however, a block surface displacement to the north-north-east.
Another indication of regional but considerable extension, most probably of
tectonic origin, is provided by the reversals of the lateral motion observed along
certain parts of the fault breaks. At Hamamlar, Ecekoy and elsewhere at Nennikiri,
Figs 2,9 and 10,where one fault segment exhibits for 1 or 2 km strike-slip offsets of a
few tens of centimetres, the next segment has the same amount of opposite displace-
ment.
The Gediz earthquake seems to be the result of a re-adjustment of a relatively new
structural system in which east-west trending blocks are being segmented by a north-
eastward regional extension, a crustal movement not related in any obvious way with
that of the Aegean Sea or the Anatolian Fault Zone. It is very probable that Tertiary
features in the Gediz area are closely affected by a pre-existing complex of basement
faults, perhaps created initially by a different stress pattern, which predispose the
region to react today in an irregular and complex fashion to what appears to be a
north-easterly extension. Fault plane solutions of the Demirci and Alasehir shocks
and of two aftershocks of these events do show normal faulting. They suggest that
present day faulting in Central Asia Minor is the consequence of the interference of a
new direction of motion with old structural trends which has reactivated pre-existing
faults in a complicated manner to accommodate the new motion (McKenzie 1973
to be published).
It is significant perhaps that the Simav and Muratdag faults, Fig. 3, are arranged
in an en echelon pattern, and the conjecture for the existance originally of an east-west
trending right-lateral fault system in Asia Minor, is tempting. The new fractures
associated with the 1970Gediz earthquake between Erdogmus and Hamamlar could
then be interpreted as the surface expression of a fault linking the en echelon
Simav-Muratdai: system.
Structural interpretations such as those made above will, however, remain purely
speculative until further field mapping clarifies the tectonics of the region.
The question of whether the fault displacements measured a few days after the
main shock are in fact the displacements that did actually occur during the event, is
difficult to answer. One thing is, however, certain; in the immediate vicinity of fault-
breaks the ground motion should have been equal to, and more often less than at some
distance away. At sites straddling the fault-break, at Civarigiirlek, Pinarbagi, Sazak
and Erdogmus, Fig. 15, the ground motion was unquestionably of very low particle
velocity if we judge from the damage inflicted upon these sites. Elsewhere, fault-
breaks showing throws up to two metres in incompetent deposits were found standing
upright, Figs 4 and 13. It seems, therefore, that the actual relative displacements that
took place on faults during the Gediz earthquake were much smaller than those
measured a few hours or days after the event and that the latter must contain a
considerable component of rapid creep. However, until more details are available
on the transient part of surface faulting, any estimates of the motion will be subject
to some ambiguity of interpretation.

FIG.15. Fault break approaching Erdogmiis from the west.
,Fucinx p. 242

Acknowledgments
Following the earthquake, and with the agreement of the Government of Turkey,
UNESCO sent a reconnaissance mission to Gediz. The chief objectives of the mission
were to make a preliminary study of the seismotectonic and engineering aspects of
the earthquake. In addition to the writers, members of the mission were, Dr M.
Arsovski of the Earthquake Research Institute of the University of Skopje, Jugoslavia;
Professor J. Grases of the University of Caracas, Venezuela; Mr A. Moinfar of the
Plan Organization, Tehran, Iran, and Mr G. Valenzuela of the University of Santiago,
Chile. The Mission was the guest of the Earthquake Research Institute of the Ministry
of Housing in Ankara, headed by Mr A. Aytun and the late M. Tasdemiroglu. The
writers acknowledge the use of considerable help from many Turkish engineers and
geologists and the assistance of the United Nations Office in Ankara. Subsequent
field trips of the senior author to the epicentral area late in 1970 and again in the
spring of 1972were partly supported by the Natural Environment Research Council.
Dedication
This paper is dedicated to Mehmet Tasdemiroglu, whose protracted illness and
premature death deprived Turkey of one of its leading field geologists and the authors
of an old friend and companion of a great many reconnaissance missions.
Engineering Seismology Section,
Imperial CoIlege of Science,
London, S.W.7
References
Abdusselamoglu, S., 1970. Observations of the seismic activity of the Gediz region,
Gediz Depremi Simpozyumu, Ankara 8-1 6 (in Turkish).
Agamennone, M., 1900. Tremblements de terre dans 1’Empire Ottoman, Beit.
Geophys., Leipzig, 4, 136.
Akkus, M., 1962. The geology of the area between Kiitahya and Gediz, Bull.
Min. Explor. Inst. Turkey, 21-30.
Ambraseys, N., 1969. On the seismicity of earthquake motion, Izo. Fiziki Zemli,
7, 86-90, Moscow.
Ambraseys, N., 1971. Value of historical records of earthquakes, Nature, 232,
375-379.
Ambraseys, N. & Tchalenko, J., 1970. The Gediz earthquake of March 28, 1970,
Nature, 227, 592-3, London. Also Proceedings 3rd European Symposium
Earthq. Eng., 1, 193-199, Sofia.
Arpat, E. & Bingol, E., 1969. The rift system of the Western Turkey, Bull. Min.
Explor. Inst. Turkey, No. 73, 1-9.
Arsovski, M., 1970. Tectonics of the Gediz earthquake, UNESCO Reconnaissance
Mission Report, Paris.
Aytun, A., 1971. Experience gained from recent earthquakes in Turkey, Proc.
NATO Earthquake Conference, Paper 7, San Francisco.
Aytun, A. & Tasdemiroglu, M., 1970. First report on the Gediz earthquake of 28
March 1970, Imar ve Iskan Bakanligi, Afet Isleri Gene1 Miidiirlugu, Ankara
(in Turkish).
Bith, M., 1953. Seismicity of Fennoscandia and related problems, Gerlands Beitr.
Geophys., 63, 186.
Can, R., 1969. The earthquakes of Demirci and Alasehir of March 1969; a field
report. Unpublished Report, Engineering Seism. Section, Imperial College.

244 N. N. Ambraseys and J. S.Tchalenko
Celebi, M. & Uzsoy, S., 1970. Structural damage at Gediz earthquake on March 28
1970,Publ. Middle East Tech. Univ.,Ankara.
CTIA, 1968. Carte Tectonique Internationale de I’Afrique, 1 : 5,000,000 scale,
Feuille 2, UNESCO, Paris.
CTIE, 1964. Carte Tectonique Internationale de l’Europe, 1 : 2,500,000 scale,
Feuille 15, Moscow.
Dubertret, L., 1970. Explanatory text of the Geological Map of Turkey,
1 :500,000 scale, Izmir Sheet, Maden Tetkik ve Arama Enstitiisii, Ankara.
Ergin, K., Uz, Z. & Gii~lii,U., 1972. 28 Mart 1970 Gediz depremi art sarsintilarin-
in incelenmesi, Madan Fakiiltesi Publ. no. 29, Istanbul Teknik Univ., Istanbul.
FouchC, M. & Pinar, N., 1944. Etude gtologique et mCtCorologique du tremble-
ment de terre d’Adapazar du 20 juin 1943. Istanbul Univ. Fen Fakiilt.
Mecmuasi, Seri A, Cilt 7 (dated 1943).
Gilbert, G. K., 1890. Lake Bonneville, Monogr. U.S. Geol. Survey no. 1, p. 354.
Graciansky, P. de, 1966. Le massif cristallin du Menderes, un exemple possible de
vieux socle granitique remobilid, Rev. Geogr. Phys. Geol. Dynam., 8, fasc. 4,
Hamilton, W., 1837. Extracts from notes made on a journey in Asia Minor in 1836,
Javaheri, J. H., 1970. Dasht-e Bayaz earthquake-Numbluk Valley. DIC Thesis,
Ketin, I., 1964. Carte de Turquie, Tectonique de I’Europe, 258, Moscow.
Ketin, I., 1967. Relations between general tectonic features and the main earth-
quake regions of Turkey. Publ. Dept. Geol. Tech. Univ., Istanbul, also Bull.
Min. Res. Explor. Inst. Turkey, no. 71 (1968), 63-67.
Ketin, I., 1970. Seismic activity of the neogene regions of Western Anatolia, Gediz
Depremi Simpozyumu,Ankara, 8-16 (in Turkish).
Ketin, I. & Abdiisselamoglu, S., 1969. 23 Mart 1969 Demirci ve 28 Mart 1969
Alasehir-Sarigol depremleri hakkinda makrosismik gozlemler, Maden Mecu-
muasi, 4, no. 5 Geology Dept., Univ. Istanbul.
MTA, 1964. Tiirkiye Jeoloji Haritasi, Izmir 1 : 500,000 Maden Tetkik va Arama
Enstitusii, Ankara.
St Amand, P., 1956. Two proposed measures of seismicity, Bull. seism. SOC.Am., 46.
Schaffer, F., 1900. Das Meanderthalbeben von 20 September 1899, Mittheil. Kais.
Konigl. Geograph. Gesellsch. Wien,43, 221-230.
Slemmons, D. B., 1957. Geological effects of the Dixie Valley-Fairview Peak,
Nevada, earthquake of December 16, 1954,Bull. seism. SOC.Am., 47, 368.
Tarr, R. S. & Martin, L., 1912. The earthquakes at Yakutat Bay, Alaska, in
September 1899, U.S. Geological Survey, Profess. Paper 69, 36-40.
Tasdemiroglu, M., 1971. The 1970 Gediz earthquake in Western Anatolia, Turkey,
Bull. seism. SOC.Am., 61, 1507-1528.
Tezcan, S. & Ipek, M., 1971. March 28 1970, Gediz Turkey earthquake and its
long distance effects, Publ. Research Centre, Robert College, Istanbul.
Uz, Z. & Gii~lii,U., 1970. Distribution of aftershocks and intensity of the Gediz
earthquake, Gediz Depremi Simpozyumu, Ankara, 3348 (in Turkish).
Uzsoy, S. & Celebi, M., 1970. Structural damage caused by the Gediz-Kiitahya
earthquake of 28 March 1970, Publ. Middle East Tech. Univ., 75, Ankara (in
Turkish).
Yarar, R., Demir, H., Kumbasar, N. & Trusia, A., 1970. Preliminary report of the
Gediz earthquake, Publ. Istanbul Tech. Univ., Istanbul (in Turkish).
ZAtopek, A. & Ambraseys, N., 1969. On the determination of macroseismic intensi-
ties, Izv. Fiziki Zemli, 7, 86-90, Moscow.
Zeschke, G., 1954. Der Simav-Graben und seine Gesteine, Bull. geol. SOC.
Turkey, 179-189.
289-306.
Journ. Royal Geogr. SOC.,7, 34.
Engineering SeismologySection, Imperial College, London.

Seismotectonic aspects of the Gediz, Turkey
Appendix
Description of the ground deformationsassociated with the Gediz earthquake
245
The reader is referred to Figs 9 and 10,and also to Fig. 2. Vertical and horizontal
displacements have been rounded off to the nearest 5 cm, and their actual value is
shown on Figs 9 and 10. Values of vertical motion refer to the vertical component of
the fault displacement and are shown on the side of relative uplift of the fracture.
Unless otherwise stated, relative horizontal displacements refer to individualfractures
at a point and they are not representative of the general motion of the two sides of
the rupture. Dotted tracesindicate that theyare not definitelyconnectablethroughout
their whole length.
The description of the ground deformations begins at the extreme north of the
ASikpaSa-Muhipler section, Fig. 2. It is followed by the description of the shorter
features to the east of Ballibaba-Ecekoy, AkGaalan,Tuzluburun, Nennikiri, and ends
with the description of the main east-west section of Erdogmus-Hamamlar.
Mapping was carried out on a 1 :25 000scalewith the exceptionof the Erdobus-
Hamamlar section east of 29" 34', which was mapped on a scale 1 :100000, and
which was not completed. It is very probable that ground ruptures extended for
2-3 km towards Ugurluca; also details of these fractures were not studied. Orienta-
tions were checked magnetically, and altitudes by aneroid barometers. Along the
immediate vicinity of the fault-break the geology was also plotted, but with less
accuracy.
In what follows we give a description of the fault zone together with the effects
of the earthquake on man-made structures and on the ground itself. In Figs 9 and 10,
numbers refer to specific localities on the fault break; the reader may find these
numbers helpful as references.
AyikpaSa-Muhipler section
The northernmost ground ruptures observed in the area lie between the old site
of Soguksu and the new village to the north, point 0, Fig. 9. These ruptures are
masked by slides and show a very small throw to the east; their origin may not be
tectonic. One of these ruptures, showing tensional features, crosses the road from
Old Soguksu to the north, the whole region including the site of the old village being
situated on a vast landslide of comparatively recent development. It appears that as
a result of a major flood, the old site was abandoned shortly before the earthquake,
and the village moved across the Daryeri stream and rebuilt there. The new village,
built on travertine, was totally ruined by the earthquake while stillunder construction.
Another series of rupture lies about 1 km south-west of Agikpaaa. They occur
along the contact between crystallineand upper Tertiary limestonesand marls which,
without any sign of lateral displacement,have been downthrown to the east by about
20cm. Slumping of the marls has produced second line of ruptures which runs
parallel with and 50 m to the north-east of the contact. These two parallel fractures
run down the slope of the hill for a few hundred metres and die out before reaching
the Asilik Dere, point 1.
Another set of ruptures was found further up the thickly wooded slopes of the
Efeklik Tepe, near point 2. From this point the fault-trace can be followed to the
south, almost continuously for 5km. Between point 2 and its crossing with the
Asilik Dere, the trace is again double. One branch follows the contact between
crystalline limestones and marls and the other runs parallel with the contact, cutting
through weathered limestones and topsoil. The motion on these two ruptures is
mainly tensional, with the block in between downthrown by about 20 cm and with a
perceptible but not persisting left-lateral offset of a few tens of centimetres.
After crossing the Asilik Dere, the trace becomes single for almost 1km and
2

246
follows a pre-existing fault in Mesozoic limestones. Between the Asilik and the
Karasu Dere the trace develops a neat vertical throw at the foot of an existing terrace
in weathered red limestones and topsoil, of 100to 180centimetreson the east, Fig. 4.
It exposes at irregular intervals parts of an old fault surface of endured limestone.
Here, en echelon cracks in the downthrown topsoil indicate a possible, but small,
left-lateral offset of about 20cm. However, this en echelon pattern seems to be
connected with the separation of the two sides of the trace, which opened directly
rather than with a genuine strike-slip displacement.
At its crossing with the Karasu Dere, the trace frays out in a number of irregular
tensionalfractureswhich produced a series of smallwaterfallsin the bed of the stream.
At the time of our first visit, about 2 m3s-l was flowing into these open fractures;
for about six days after the earthquake, the Karasu Dere dried up below this point,
and the whole discharge of the stream, about 3-4 m3s-', disappeared into these
fractures.
South of the Karasu Dere, the trace continues as a single rupture in crystalline
limestones, running up the thickly wooded slopes of the Ayikayasi Tepe to the south.
Here the trace is a reactivated old fault on which there has been repeated movements
during late Quaternary time. Within 200 m from its crossing with the Karasu Dere,
the throw to the east increasesfrom a few tens to 125cm with a clear left-lateraloffset
of about 20cm, (see Fig. 2 in Ambraseys et al. 1970). Dips of the fault plane in
crystalline limestones are between 65" and 70" east.
The exposed face of the fault surface shows slikensides and striations caused by
earlier movements on this fault in this sense.
Betweenpoints 3and 4 the tracefollowsvery closelythe contact between crystalline
limestones and Tertiary conglomerates,mainly weathered limestones. From point 3,
the throw to the east increases from 125 to 270cm, and near the summit of the
Ayikayasi Tepe the trace becomes a gaping fissure, 4 m deep at the foot of the scarp,
large enough to admit a man. The vertical offset is variable; the east side, that is the
downhill side, is downthrown, in places up to 200cm and the total height of the
hangingwall to the bottom of the gap is more than 6 m. The morphology of the scarp
along this section of the trace is very similar to that described by Gilbert (1890) as a
subsidence zone and gravity graben; one or more fractures cutting through the
conglomerates of the hanging wall, produce the spreading and settling of the Tertiary
superficialmaterial.
Past the summit of the AyikayasiTepe the trace runs up and down over a number
of low ridges, following the contact between limestones and conglomerates, main-
taining a throw to the east of 150-200 cm and an apparent left-lateral offsetof about
30 cm.
To the south of point 4, the trace becomes difficult to follow; it passes through
upper Tertiary tuffs and volcanic agglomeratesin the form of a rather wide fracture
zone, in places 100m wide. Here, it was not possible to measure the total amount of
vertical movement across the zone, but it should be more than 150cm with a few tens
of centimetres of cummulative left-lateral offset. The vertical offset on few fractures
is about 30 cm, down on the east.
Between points 4 and 5 the trace lies in upper Tertiary volcanics and runs at the
base of a series of low hills. It shows a throw to the east of 20-100 cm and a small
left-lateraloffsetthe magnitudeof whichis difficultto measurebecauseofthe shattering
in the fracture zone. In places this zone of fractures shows a tendency to bifurcateto
the south-east, particularly near point 4 as well as west of the SelimayaylasiTepe, but
individual fractures never leave the main zone for more than a few hundred metres
before they die out. Near its crossing with the Kalenlik Dere, the trace showsstrong
tensional features; it becomes double and develops a comparatively large gravity
graben which caused a small lake to form.
Near point 5, just before reaching the Madendagi Dere, the trace makes a sharp
N.N.Ambraseys and J. S.Tchalenko

turn to the south-eastandfor afewhundred metresfollowsasa singlecrack the contact
between upper Tertiary volcanics and an andesitic outcrop which was downthrown
by a few tens of centimetreswith no evidence of lateral motion.
After crossing the Madendaki Dere, the trace frays out into a zone of small
fractures and, uninterrupted for about a kilometre, resumes its southward course
through shattered upper Tertiary volcanics and weathered limestones. This fracture
zone consists of numerous irregular cracks on a very gentle topography, the east side
being generally downthrown by 5-25 cm. On some cracks there is evidence of left-
lateral movement but the overallmovementis tensional. Near the Ahir Dere, the zone
of ruptures dies out and it cannot be found on the opposite banks of the stream.
South of the Ahir Dere, at point 6, another trace in andesites winds along the
base of a seriesof low scarps on the west slopeof the Kayabasi Tepe. Here, the up-hill
side has dropped by 10-25cm with respect to the downhill side with no conclusive
evidence of preferential lateral movement. Further south the trace becomes very
weak to follow.
At point 7, west of Yumurtaskoy, a long rupture in young volcanics was found
running for a few hundred metres. It is probably of landslide origin and its east side
is downthrown by 10-50 cm, showing a small left-lateral movement.
West of Yelkikoy the trace reappears in marls mainly as a wide zone of relatively
weakfractures. It continuesforabout 1km, until east ofPinarbasi the trace encounters
a series of ophiolitic outcrops which it shatters. At this point the trace turns in a
southeasterly direction and for about one kilometre cuts through a thin cover of
marls which were intensely sheared and offset in a left-lateral sense by about 40 cm,
downthrown on the north-east by a few tens of centimetres, Fig. 5.
The trace continues towards Muhipler, and in a number of places becomes double
and discontinuous. On all traces the direction of uplift is fairly consistent and it is
on the west or south-west. However, occasionallythe trace, where it becomes double,
shows strong extensionalfeatures and the formation of small gravity grabens. North-
west of Muhipler, the road to Pinarbasi runs on the summit of a low hill of limestones
and marls, the axis of which nears N-30" W. The trace here is double and follows the
flanks of the hill which were raised by 10-30 cm opposite to the direction of slope so
that the downhill sides of the hill were uplifted. Approaching Muhipler the trace is
obscured by numerous shallow landslides.
Near point 9, a series of discontinuous cracks was found running along the
contact between serpentines and marls, leading towards Pinarbasi. In places a
combination of open cracks and pressure ridges were found to radiate away from
ophiolitic outcrops suggesting a significantdisplacement of the outcrops with respect
to the overlyingmarls which, however, did not produce any definite surface pattern
of fractures.
Approaching Pinarbasi, the trace consists of a series of long but weak fractures,
two of which pass through the back of the'village,showinga throw to the south-east of
5-15 cm. Five timber frame houses straddling one of the fractures were distorted but
they did not collapse. Of the 102houses of the village, two collapsed but no one was
killed. Villagersstatethat the earthquakeof 1944had causedconsiderabledamageand
that the ground ruptures in the region, particularly those in the immediate vicinity of
the village, appeared in 1970at the same place as in 1944.
Between Pinarbasi and Civarigiirlek the trace disappears and re-appears above
Civarigurlek,running along a reactivated contact between serpentines and tuffs, the
latter downthrown to the south-east by 10cm. Small en echelon fissures show no
preferentiallateral displacementand may be dueto openingup of the contact. Between
these two villages a spring of water increased its yield after the earthquake flooding
the downstream part of the village. At Civarigiirlek, only four out of the 110houses
collapsed and no one was killed.

248
Ballibaba-Ecekoy section
Some of the most complicated cracks and fissures produced by the earthquake
were found east of the fault-sectionjust described. A discontinuous seriesof fractures
was found near point 10, running along a wide saddle between Efrem and Ballibaba
Tepe.
Much of the Efrem Tepe consists of Mesozoic limestones, in places coveredwith
andesitic lavas. The whole system is heavily fractured and weathered, making it
extremely difficult to identify structural elements activated by the earthquake. The
fractures observed follow the contact between crystalline schists and upper Tertiary
tuffs, down on the east by a few tens of centimetres with no sign of lateral motion.
The trace here is discontinuous and dies out in tuffs before reaching the Madendagi
Dere.
After branching off at point 11, the trace trends N40" E and follows faithfully
the 1300-m contour all the way to its crossing with the Madendaiji Dere. Up to
point 12 the trace follows the contact between crystalline schists and upper Tertiary
tuffs, after which point it follows a reactivated contact zone of upper Tertiary lime-
stones and tuffs. The trace is weak and discontinuous, with no clear evidence of
lateral movement but with a vertical throw on the east of 10-40cm.
Near its crossing with the Madendaii Dere the trace is obscured by numerous
landslides which make it impossible to determine the nature and sense of movement
After crossing the river the trace runs up and down a series of low saddles towards
Sazak, passing through upper Tertiary tuffs which it shatters, showing a small throw,
down on the north-east side by 10-30cm with no evidence of consistent lateral
motion. At point 13, the trace turns south-east and follows for 500 m a reactivated
contact between andesites and upper Tertiary tuffs.
From point 13, showing small and irregular vertical displacements in andesite
and topsoil, the trace continues to the.south and crosses the Ahir Dere. It then
passes behind and to the west of an old massive landslide, covered in places with
screeand slopedebris,utilizingthe upper boundary of the slideand inplacesmodifying
it.
Here, a number of springs of water were found issuing from behind the slide. On
entering the slope debris area, apparent vertical displacements increase from 20 to
about 150cm, and the trace, following the upper boundary of the slide, turns near
point 14 due east and passes between Karapinar Tepe and Sazak. This segment of
the trace is marked by left-lateral en echelon cracks and the lateral movement of the
conjugate fracture, north of the Karapinar Tepe, is right-lateral.
From point 14the fractures fray out and skirt Karapinar Tepe with small vertical
displacements. Fractures leading east pass between the villages of Ecekoy and Sazak
and show tensionalfeatures. In Sazak 38 out of the 75 houses collapsed and 14people
were killed. Out of 70 houses in Eckoy, 20 collapsed killing 13people. Near Sazak
the trace becomes double; one rupture zone passes through the village and forms a
graben 150m wide and 20 cm deep, with h o evidence of strike-slip movement of
opposite walls.
N. N. Ambraseys and J. S. Tchalenko
Akcaalan section
Structures west of AkGaalan that were activated during the earthquake all follow
pre-existing contact zones along at least some of which there has been repeated
movement during late Quaternary time. At point 15 a series of irregular cracks in
tuffs, which followthe 1050-mcontour for about 500 myshow that the north-east side
was downthrown by about 30cm with no evidence of lateral displacement. Near
point 16, these cracks die out and they re-appear higher up the hillside in the form of
a double trace between andesites and tuffs. Here the zone of shattering exhibits
primarily extensional characteristics and demonstrates a 20-cm right-lateral offset,

Seismotectonic aspectsof the Gediz, Turkey 249
with the tuffs between the two walls downthrown by the same amount in the form of
a gravity graben, Fig. 11.
Old striations on the south-west facing wall suggest that since the overflow of
andesites-dacitesin early Pleistocene there has been a total vertical movement of at
least 500cm associated with a right-lateral movement of the same amount.
To the south of this point, another trace can be followed for 500 m before it dies
out within a zone of landslides and rockfalls. The valley north-east of this trace is
littered with large blocks of volcanics that have rolled down from above before and
during the earthquake.
The ground deformations found at the Tuzluburun Tepe are of particular interest.
This hill of white upper Tertiary mark and limestones rises about 60m above the
surrounding topography; to the north-east it is bounded by the Ak Dere and to the
south-west by the Madendagi Dere. Near the north-west edge of the hill Point 17, a
trace heading south-eastward, was noted. The ground deformations here are greatly
complicatedby landslidinginto the Cakmak Dere, the slidingutilizingthe trace as an
upper bound and modifying it. Near point 17the trace crosses, at a very small angle,
the axis of the hill and as a single fracture follows the midheight of the south-west
slope for about 1 km; the south-west side of this fracture went up. This means that
the upthrow was on the downhillsideby 60-120 cm. Another trace wasfound running
parallel with and along the midheight of the north-east slope of the hill, also with the
downhill side uplifted by as much as 220 cm, Fig. 13. Both traces follow the 1050-m
contour and suggest that as a result of the earthquake the flanks of the hill spread
out so that the central part of the massif slumped. Elsewhere the central part of the
Tuzluburun Tepe abounds with tensional features, gravity grabens, slumping and
gaping fissures 200 cm wide. The origin of these ruptures, although in appearance
tectonic, is not clear. Both of them die out before they reach the valley below.
About 300m south-west of AkGaalan, at Point 18, another trace can be followed
for a few hundred metres to the south-east along the contact between tuffs and
Tertiary marls. The northeast side of this series of ruptures was downthrown by
5-40 cm without any perceptible lateral motion of the two sides. Approaching the
Ak Dere the trace skirts a landslide area showingirregular vertical displacements,in
places 100cm down on the north-east. After crossingthe Ahir Dere, the trace follows
the contact zone between andesitesand tuffs and showstensilefeatures, but again with
no evidence of lateral motion, continuing up on the south-west by 30 cm. For another
500 m the trace follows faithfully the contact zone between andesites and upper
Tertiary volcanics on the south-east slopes of Alibaba Tepe before it becomes too
faint to follow. A number of ground fractures passing through Akqialan were found
to be due to landsliding. In this village 126people out of 1200were killed and 371
houses out of 490 were destroyed either by collapseor by fire which was started by an
aftershock.
Nennikiri section
Another interesting series of fractures was found north of Gediz. The Kepez and
Nennikiri Tepe form a continuous narrow ridge of white lacustrine mark and upper.
Tertiary limestones, about 3-5 km long. These hills stand 40-60 m above the general
topography. To the south-westthey are bounded by the Ak Dere and to the north-east
by the Cukur Dere. The road from Gediz to Kutahia follows the crest of these hills.
A major fracture flanking the hills on both sides was found running a few tens of
metres below the crest. The uphill side was downthrown by 40-1 10cm.
The scarps of these fractures show clear breaks in indured calcareous marls,
slickensided by the downward movement of the central part of the hills on a plane
dipping about 80"SW or 80" S. On east-west trending fractures the motion was right-
lateral becoming left-lateral on north-west-south-east ruptures.

250 N.N.Ambraseys and J. S.Tchalenko
Again here, the tectonic origin of these large-scalefeatures is not clear. However,
road excavations near the point where the road from Gediz to Kiitahia crosses these
ruptures for the third time, exposed cross-sections of the ruptures which show that
movements at this point occurred on pre-existing failure surfacesin calcareousmarls
on which repeated movements in the past were much larger than those observed on
the ground surface after the Gediz earthquake.
Approaching the Zelifiyaren Tepe, the trace turns to the north and follows the
steep slopes of the Yayla Dere. Here the trace has developed characteristics of a
landslideor slumpfeature in marls and limestoneswith throws of the order of 100cm.
Similarfeatures of hill spreading, but on a very much smallerscalewere found on
the Odacamligi Tepe, north-east of Nennikiri.
Erdogmus-Hamamlarsection
Near point 20, north of the Bakir Tepe, in the plains of the Gediz river, a series
of ground ruptures passing through cultivated fields can be followed for almost a
kilometre. In places the trace is marked by open cracks with 10-cm gaps but with
no evidence of lateral movement. The vertical movement varies from 10 to 20 cm,
down on the east side. In at least two places the trace is double with the central
block downthrown by 10-30 cm.
Another trace, near point 21, followsthe west bank of the Gediz river. It shows a
double scarp one of which appears to be due to slumping of the river banks. Here it
was not possible tojudge the amount of verticalmovement, both traces being compli-
cated by secondary fractures in the fields.
At point 22, a longrupturein cultivatedfieldswasfound toleadtowards Erdogmus,
displaying a throw to the north-east of 10-20 cm. At point 23 this rupture becomes
double asit cuts across a spur of outcropping marls and forms a small gravity graben.
Approaching Erdobus, at point 24, the trace becomes double again showing strong
tensile features and gaping fissures,with the central block downthrown by 20-40 cm.
Of the two traces, the one that appears near point 24 exhibits, on flatground,consider-
able tensional characteristics with gaps 10-4Ocm wide, Fig. 15. Both traces pass
through Erdobus destroying two out of 276 houses and distorting another 15timber
frame dwellingsincluding a mosque and a large barn which were straddling the traces
No one was killed in the village. The ground deformations caused steps to form in
narrow streets and a spring of water to dry up permanently. Eye-witnessesstate that
a group of people standing together outside a coffee-shopin the village at the time of
the earthquake felt the shock but were not thrown down; one of the main fractures
dividedthegroup, twopersonsfound themselveson the downthrownside,stillstanding,
and a third person was thrown into the gap injuring his leg. It is significant that
Erdobus had been rebuilt after almost total destruction during the earthquake of
1944June 25.
Between Erdogmus and its crossingwith the Canbulat Cay, the trace is discontinu-
ous, with the north side downthrown by 10-40cm. It crosses flat ground for almost
3.5 km showing fissures, open in places, gaping by as much as 30 cm but without
any evidenceof lateral movement. Some of these traces are double with the block in
between downthrown and the depressionflooded.
Near the Canbulat Cay the trace becomes double. One branch in gravel crosses
the river at point 26, forming a series of rapids and shows a total throw to the north
of about 20 cm. The other branch appearsatpoint 25. Both branchescrosstheroad to
Gumulekoy exhibitinga left-lateral offset of 5-15 cm with the north side downthrown
by a few tens of centimetres.
From point 26 the trace is continuous and it can be followed through cultivated
fields running up the slopes of Hill-878, cutting through limestonesandmarls,withthe
north side downthrown by 20-50cm. In places the trace is an open fissure and

Seismotectonicaspects of the Gediz,Turkey 251
exhibits a small discontinuous left-lateral movement which does not exceed 25 cm.
Near the summit of Hill-878 the trace breaks up into a series of tensional cracks and
joins the northern branch in a zone of intense fracturing. Here it was not possible to
judge the amount of total vertical movement, but it should have been more than
100cm, down on the north.
The northern branch, also in marls, shows larger vertical displacements, down on
the north, in places by 100cm. The trace is discontinuous and consists of long
en echelon cracks demonstrating a left-lateral displacementnot exceeding20 cm.
Near the summit of Hill-878 the trace becomes single again and runs down the
slope into the Genez valley in marls and limestones.
The vertical displacement here is lW120cm, down on the north. The trace
continues due east, mainly as a tensional feature, with no well-defined evidence of
lateral movement. Near its crossing with the Genez Dere a series of small gravity
grabens, flooded after heavy rains, produced a shallow pond.
The trace runs up the Naliburnu Hill and for about 2km follows the 900-m
contour on relatively flat ground. In places the trace is an open fissure, the two sides
having been pulled apart by as much as 100cm with no sign of lateral offset, down
on the north by 20-70 cm. Elsewheretensional features have produced small gravity
grabens downthrown by a fewtens of centimetres. A number of footpaths which cross
the trace were found offset not more than 25 cm in a left-lateral sense.
Approaching the Cevizli Dere the trace becomes more linear and continuous and
it showsa throw to the north of about 100cm. Here, as wellasin a number of isolated
exposures, the fault plane in limestone dips 60" to 65" to the north. Near the Cevizli
Dere the trace follows the base of an old reactivated escarpment of pink brecciated
limestones which it follows with a comparatively small vertical displacement. Near
point 27 the trace crosses the Cevizli Dere and runs parallel with and about 10m to
the north (downstream) of the contact between serpentines and marls exposed in the
river bed and further east on Hill-927. Here the trace shows a throw to the north of
about 20 cm with some evidence of left-lateral movement of 10-20 cm. Open cracks
and smallfissuresin the serpentinessuggestsomereactivation of this contact. Between
Genez and Cevizli the manner in which the fault break crosses ridges and smallvalleys
suggeststhat the fault plane dips about 60" to the north.
Further east the trace runs parallel with the serpentines and dies out before
reaching the Muratdaii Cay in a zone of heavy fracturing.
Here the trace is double for a short distance with the block in between down-
thrown.
Between Giimuslu and Sazkoy, at point 28, another trace was found following
the 800-m contour, running parallel with and to the north-west of Akdere. The trace
is in marls and limestones and shows a discontinuous vertical offset of 10-40cm on
the south-east. An en echelon system of fissures suggest left-lateral motion of small
amplitude. This trace dies out near the source of Akdere; no attempt was made to
search for ground deformations north-east of this point.
Two days after the main shock, near point 28, a new thermal source burst out in
the banks of the Muratdaii Cay, not far from anotherpermanent sourcenear the Mill
at point 29. For a few days after the earthquake, this source behaved like a geyser,
ejecting at intervals, sand mixed with warm water, occasionally spouting to heights
of a few metres. At the time of our visit the source continued to flow, discharginga
few tens of litres per second of relatively cool water at 2OC.
From point 29another trace can be followedeast for almost 5 km. From this point
to its crossing with the Goklet Dere, the trace is in marls and limestones; it runs up
and down several thickly wooded hills and gradually climbs the south side of Sazkoy
Hill, showing a throw of 30-60 cm to the north with a 25-cm left-lateral offset.
After crossing the Gokler Dere, the trace becomes double for a short distance.
A short rupture in limestonesshows small vertical movements with some evidence of

252 N. N. Ambraseys and J. S.Tchalenko
left-lateral movement on offset forest foot-paths. Another trace runs parallel with and
a few tens of metres to the north of a contact zone between upper Tertiary limestones
and serpentines, showing a throw of 20-30 cm to the north with little or no evidence
of lateral motion.
Past point 30, the trace consists of a series of long en echelon fractures, wholly in
limestones, showing a small throw to the north of 10-20 cm but with clear evidence
of left-lateral movement of the same amount.
As it approaches the Kogur Dere, the trace branches off and skirts a wide zone of
shattered limestones which is bounded to the north and south by prominent fractures,
the whole system giving the impression of an incipient gravity graben. Here, as well as
in other parts of this trace, a number of springs of water appeared shortly after
the earthquake, some of them behaving like geysers.
South of Somakli a series of discontinuous ground ruptures can be followed with
difficulty, the terrain in places being almost inaccessible. Of the two traces south of
Somakli, the one to the south-east can be seen in places but it was not followed up.
Between Yumurak Kayazi and Murat Daii-Hamanlar, the ground deformations
follow pre-existing structures, along at least some of which, notably east of Fikiriz,
there has been repeated movements in a right-lateral sense during late Quaternary
time and again during the earthquake. Displacements of 40 cm of right-lateral
movement on some of these faults with a throw to the north of about 40 cm were
measured. Where the trace crosses the Baybiyen Dere the sense of lateral movement
is reversed. Here the trace follows the contact zone between crystalline schists and
serpentines and shows a left-lateral motion of 70 cm with the north side downthrown
by 60 cm.

Gediz Earthquake: 1970

Empfohlen

Empfohlen

Weitere ähnliche Inhalte

Was ist angesagt?

Was ist angesagt? (11)

Andere mochten auch

Andere mochten auch (6)

Ähnlich wie Gediz Earthquake: 1970

Ähnlich wie Gediz Earthquake: 1970 (20)

Mehr von Ali Osman Öncel

Mehr von Ali Osman Öncel (20)

Kürzlich hochgeladen

Kürzlich hochgeladen (20)

Gediz Earthquake: 1970