Content uploaded by Amos Salamon
Author content
All content in this area was uploaded by Amos Salamon on Jul 16, 2014
Content may be subject to copyright.
JSeismol
DOI 10.1007/s10950-009-9168-9
ORIGINAL ARTICLE
Patterns of seismic sequences in the Levant—interpretation
of historical seismicity
Amos Salamon
Received: 4 September 2008 / Accepted: 27 April 2009
© Springer Science + Business Media B.V. 2009
Abstract There are historical accounts of about
a hundred damaging earthquakes that occurred
during the last two millennia in the Levant, in
and around the Dead Sea fault system, and about
half of which were associated with additional felt
shocks. Several modes of earthquake sequences
can be distinguished in them: (a) In 46 accounts,
only one single event is noted. These are not
known from tectonic settings similar to that of
the Levant, and may just be a result of incom-
plete reporting. (b) In four cases, quakes pre-
ceded the mainshock by minutes, hours, and up
to several weeks—possibly foreshocks. (c) Thirty-
five mainshock–aftershock sequences were noted,
lasting hours, days, weeks, months, and even more
than a year; four of these also have foreshocks. No
typical delay time was recognized for the largest
or most significant aftershocks: they appeared up
to several months later. (d) Six of the reported
mainshock–aftershock sequences appeared in a
“storm.” Another 13 sequences are insufficient to
specify further.
A. Salamon (
B
)
Geological Survey of Israel, 30 Malkhe Israel St.,
Jerusalem 95501, Israel
e-mail: salamon@gsi.gov.il
Keywords Aftershocks ·Dead Sea fault system ·
Earthquake sequences ·Foreshocks ·
Historical seismicity ·Mainshocks
1 Introduction
How strong aftershocks are and how long they
last are big concerns after the mainshock. Strong
events are infrequent in the Levant, but reports
accumulated for more than two millennia do
describe them, although not necessarily in mod-
ern terms. Earthquakes appear in various tem-
poral and spatial distributions, mostly in dense
clusters consisting of one large event, the main-
shock, which is usually followed by many smaller
ones, the aftershocks. Occasionally, the main-
shock is preceded by a single or several fore-
shocks (preshocks). A sequence in which there
is no predominant single earthquake is a swarm,
and a successive occurrence of several mainshock–
aftershock sequences is a second kind of a swarm
or a seismic storm (e.g., Kisslinger 1996; Utsu
1961, 2002, and references therein). There is no
universal definition or procedure for classifying
seismic sequences or determining the exact na-
ture of any single earthquake in a cluster. Most
researchers refer to aftershocks as events imme-
diately following a large earthquake within a dis-
tance of one to two rupture lengths from it and
at a rate of occurrence that is higher than the
JSeismol
background seismicity of that area prior to the
mainshock. Practically, however, aftershocks are
defined ad-hoc, according to the scope of the given
study (Utsu 1961, 2002).
In general, the stronger the mainshock, the
larger and more numerous the aftershocks are and
the longer their period of occurrence is, even up to
several years. These short-term relations have tra-
ditionally been expressed by three empirical scal-
ing laws, namely: (1) the frequency–magnitude
power law relation (Gutenberg and Richter 1954);
(2) the average size of the largest aftershock is
∼1.2 magnitude units smaller than the mainshock
(Båth 1965), but this may vary widely from 0.3 to
three units or more and cannot be predicted; and
(3) the frequency of aftershock occurrence decays
at a rate proportional to t
−p
,wheret is the time
after the mainshock and p ∼ 1 (Omori 1894). The
generalized Omori law attempts to incorporate all
three relations into one (Shcherbakov et al. 2004).
The timing of the largest aftershocks, delaying
days, weeks, and even months, cannot be pre-
dicted (e.g., Hough and Jones 1997). Moreover,
often, an earthquake sequence triggers a new se-
quence nearby in which the mainshock is some-
times even stronger than the previous one. This
might be explained by a stress-triggering effect,
but so far, its timing is not predictable either. The
area affected by an aftershock occurrence is rela-
tive to the size of the mainshock (e.g., Utsu 2002),
in accordance with the notion that the length of
the seismogenic rupture scales with the magnitude
of the mainshock.
Much reliable historical data of seismic se-
quences in the Levant has been gathered in re-
cent times. All this information, which extends
over a period paralleled only in a few other areas
worldwide, has not yet been analyzed. Although
limited and partial, evaluation of these data could
be more productive than waiting for future events,
and may give us greater insight into the seis-
mogenic nature of the Dead Sea fault system
(DSFS), the major seismotectonic element in the
Levant. Therefore, this study focuses on typifying
the sequence of events that preceded and followed
strong historical earthquakes in and around the
DSFS. How the list of damaging earthquakes was
compiled is explained and the patterns of seismic-
ity that were possible to identify are discussed.
2 Historical seismicity
There are several sources that report or indicate
past earthquakes; each has advantages and limi-
tations, and none lack uncertainties. Most com-
mon are historical documentations, paleoseismic
evidence, and archaeoseismic findings; the for-
mer constitute the main body of information that
details patterns of the past seismicity. Although
written in various languages from many places and
cultures, described in many different ways and
points of view, and expressing political and reli-
gious views, these are the most credible sources
available. Bringing all the events together to a
common denominator in one language and in
seismological terms is essential in order to build
as unified and complete a catalogue as possible,
but this is not a trivial matter. This aim has al-
ready been targeted by several modern investiga-
tors (e.g., Ambraseys, Guidoboni, Karcz), and this
work is based on their studies.
2.1 Limitations of the historical data
The notion that many catalogues of historical
earthquakes contain erroneous entries has al-
ready been raised by several researchers (e.g.,
Ambraseys et al. 2002; Guidoboni and Comastri
2005;Karcz2004) and that of the Levant is of no
exception. They demonstrated that several of the
events noted in the published literature originated
from misprinting, misinterpretation, duplication,
etc., and, in fact, are false. Their critical studies
were able to filter out many of the questionable
events and verify the record of many significant
earthquakes. Overall, it seems that sufficient re-
liable information has already been accumulated,
enabling the reconstruction of a dependable and
comprehensive list of earthquakes that originated
in the Levant area. The evaluation made here is
based upon these works.
The main limitation of the historical reports
is the subjective description, which does not
detail the source parameters. Moreover, parame-
terization of the past accounts is also an unavoid-
able subjective process involving considerable
uncertainties. In contrast to modern instrumen-
tal recording, where the threshold of detection is
more or less known and completeness of the list
JSeismol
can be defined, it is not possible to know which
category or type of event is missing from the his-
torical record. Hence, the reported earthquakes
are probably a partial accounting of the actual
occurrences. It is therefore important to mark out
a clear line between the original story and the
evaluated parameters. It is not even possible to
quantify the uncertainty associated with the eval-
uation because it is a result of personal judgment
rather than of a measuring procedure. Therefore,
one should be aware of the unknowns typical to
the historical material, and realize how far an in-
terpretation can go. The uncertainties associated
with each of the source parameters are discussed
below.
2.2 Sources of data
In general, the studies that relied directly on
contemporary and primary sources and extracted
an accurate description of an event with mini-
mal necessary interpretation (e.g., translation, his-
torical perspective, etc.) are the more reliable.
Studies that provide detailed and accurate refer-
encing to the primary sources are also dependable.
On the other hand, the authenticity of events that
appear in lists with no reference to the original ac-
counts should be doubted. Several in-depth inves-
tigations (e.g., Ambraseys et al. 2002; Guidoboni
et al. 1994;Karcz1987, 2004; Karcz and Lom
1987) examined most of the existing catalogues,
pointed out their advantages and shortcomings,
and stressed the caution needed in studying their
information. Following these works, it was possi-
ble to prioritize the available catalogues, compile
lists of events whose authenticity could be well
verified, and point to the doubtful events. There
was no attempt here to reexamine the original
sources, and so, the present work relies on com-
paring and evaluating different studies that rely on
primary accounts, while not escaping their uncer-
tainties. Nevertheless, as new data are discovered
and better interpretation offered, there will be
a need to reexamine the original sources and to
reevaluate the reliability of the present listings.
Most of the information regarding events
around the Mediterranean up to the fifteenth
century A.D. has already been collected, com-
piled, analyzed, and presented in the catalogues of
Ambraseys et al. (1994), Guidoboni et al. (1994),
and Guidoboni and Comastri (2005). Later events,
however, have not yet been analyzed in a sim-
ilarly systematic approach. Other invaluable re-
views were published by Ambraseys (1989, 2004),
Ambraseys and Finkel (1995), Poirier and Taher
(1980), and others. Reappraisals by Ambraseys
(2005a, b), Ambraseys and White (1997), and
Karcz (2004), as well as focused investigation on
specific events (e.g., Ambraseys and Barazangi
1989; Ambraseys and Karcz 1992; Ambraseys and
Melville 1988; Darawcheh et al. 2000; Guidoboni
et al. 2004a, b), are also available. Many other
lists draw from both primary and secondary
sources (e.g., Amiran et al. 1994; Ben-Menahem
1991; Khair et al. 2000; Plassard and Kogoj 1968;
Sbeinati et al. 2004), and they are referred to
after cross correlating their data with the primary
sources.
There have been far fewer direct field investiga-
tions of past earthquakes, and these may provide
a better estimate of the rupture zone, magnitude,
and mechanism (e.g., Akyuz et al. 2006;Amit
et al. 1999; Daëron et al. 2005, 2007; Elias et al.
2007; Ellenblum et al. 1998; Gomez et al. 2001,
2003
; Klinger et al. 2000; Marco et al. 1997, 2003,
2005; Meghraoui et al. 2003; Neimi et al. 2001;
Nemer and Meghraoui 2006; Reches and Hoexter
1981; Zilberman et al. 2004, 2005). Attributing
source parameters to historical events, however, is
not a straightforward process. In most cases, field
evidence is associated with the candidate from a
known list of historical earthquakes that best fits
it. Taking the paleoseismic data back to prove
and support the existence of the selected histor-
ical event should be done with care because, for
example, this event may not have been reported
at all.
Other seismogenic effects, such as lacustrine
seismites and deformed layers from the Dead
Sea basin (Marco et al. 1996; Enzel et al. 2000;
Ken-Tor et al. 2001; Migowski et al. 2004), may
attest to the strength of shaking or to the dis-
tance from the source of many of the historical
events. These natural features may also record
strong earthquakes that were possibly missed,
ignored, lost, or not reported in the course of
history. Similarly, archaeoseismological evidence
also contains useful information. These sources
JSeismol
are limited in pointing to the exact date, source
area, and strength of the event, yet they provide
invaluable information that certainly should be
addressed in future studies (Ambraseys 2006a;
Karcz et al. 1977).
3 List of events
On the whole, all the events reported by historical
sources to have occurred in the Levant and to
have damaged at least one site were considered
(Appendix A). Events inferred from seismites or
archaeoseismic findings were not considered in
this study. The collected data were correlated
with the existing literature in order to validate or
question each of the events. In most cases, stud-
ies that relied on primary sources and extracted
the original description more accurately enabled
distinguishing real events from questionable ones.
In cases still unresolved, personal judgment was
counted on. This resulted in a compact, yet more
reliable list of earthquakes than if every possible
event were listed. Clearly, this is not an error-free
evaluation, and the list will need to be updated if
new sources are discovered and better interpreta-
tions are offered.
As the nature of the DSFS was focused on, and
given the limited resolution of the historical data,
all the events that affected the Levant and that
seemed to originate from this source were listed.
For example, Elias et al. (2007) suggested that the
damage in the coastal cities of Lebanon caused
by the earthquake of 551 A.D. was due to the
rupture of the offshore Mount Lebanon thrust,
and that this thrust belongs to the DSFS system.
This implies that, in addition to shear events that
originated from the Dead Sea Transform and its
parallel and branching faults, thrusts and normal
faults were related to as well. On the other hand,
events reported to affect nearby regions that be-
long to another seismotectonic regime were ex-
cluded, such as, for example, the event of 1222
from Cyprus (Guidoboni and Comastri 2005), that
of 1568 from the Mediterranean west of the Syrian
coast (Ambraseys and Finkel 1995), and the 1269
earthquake in Cilicia (Guidoboni and Comastri
2005), which may have all originated from the
Cypriot Arc. Similarly, the 1114 sequence may
have come from the East Anatolian Fault zone
(Ambraseys 2004); the 749 event in Mesopotamia
(Karcz 2004) seems to be closer to the Arabia–
Anatolia collision zone; and earthquakes in Egypt,
which are clearly west of the DSFS, resulted from
structures such as the Suez Rift and others.
There is no consistent or systematic way to
determine the location and strength of histori-
cal earthquakes, and they are mainly based on
felt shocks and damage reports. Several quanti-
tative methods were developed in cases where a
good set of data is available (e.g., Shebalin 1973;
Bakun and Wentworth 1997; Gasperini et al. 1999;
Sirovich and Pettenati 2001), as well as empiri-
cal intensity–magnitude relations (e.g., felt area–
magnitude, by Frankel 1994). In addition, eval-
uation of paleo- and archaeoseismological data,
in the context of the areal neotectonics, can help
in constraining the source parameters. Running
these procedures, however, requires special at-
tention and is beyond the scope of this work.
Nonetheless, the preliminary estimates already
available in the existing literature suit the purpose
of this study. The present list (Appendix A) shows
the date of occurrence of each of the events, local-
ity of the most severely affected areas, estimated
size class, sources of information, and the pattern
of seismicity, as explained herein.
3.1 Origin time
As trivial as this parameter may be, determin-
ing it is not at all simple. For example, the ex-
ceptionally intense seismic sequences along the
northern Dead Sea transform during the twelfth
century were extensively studied by Ambraseys
(2004) and Guidoboni et al. (2004a, b). Basically,
the two researchers drew from more or less the
same original contemporaneous sources, but each
arrived at different origin times (by days) of what
seems to be the same sequence of events.
In general, origin times were taken from stud-
ies that carefully examined the given event and
addressed errors in the published literature, be-
cause misinterpretation may simply have resulted
in duplicating that event and inflating the actual
list. In cases that were hard to resolve, the best
JSeismol
estimate, in our opinion, was adopted. Still, some
uncertainties remained, thus making the present
interpretation only preliminary to future resolu-
tion of uncertainties. The selected events are cited
by their time of occurrence and, to the extent
known, by the year, month, day, part of the day,
hour, and minute.
3.2 Location of the historical earthquakes
Ideally, the center of the most severely damaged
area should coincide with the rupture zone of the
given earthquake. However, this was proved to
be too simplistic an assumption, and in fact, the
relationship is much more complicated. The popu-
lated area is not homogeneously scattered around
the earthquake epicenter or in the area of interest,
the spread of the seismic waves is not symmetric
(e.g., velocity structure, focal mechanism, direc-
tivity), site effects dramatically vary from place
to place, and historical reports in most cases are
biased and incomplete. Locating a historical event
is, therefore, a dependent procedure and, if not
carefully done, may result in duplicating an event,
moving it to a different area, or borrowing it from
elsewhere (e.g., Karcz 2004).
The original reports vary from referring gener-
ally to the affected area, pointing to one site only,
and listing (all?) the affected localities and struc-
tures. Single sites might have been mentioned
for being the most severely affected; those only
known to be affected; most important or of special
interest for the reporter; and, in some cases, the
site where the earthquake was felt by the reporter.
In some cases, it is possible to delineate the most
severely affected area (I
0
), while in others, where
only a partial coverage of the affected zone was
available, only the site of maximal damage (Imax)
can be identified. The last is possibly the closest
to the epicenter, but the seismogenic source could
well be far away. Some localities, such as Antioch
or Jerusalem, were mentioned to have been hit
again and again, possibly due to being important
cultural, political, and religious centers. For the
very same reason, we may assume that other less
important sites might have just been neglected.
Given the above constraints, the simplest and
least biased preliminary interpretation would be
to assume that the center of the most severely
damaged zone represents the seismic epicenter.
The estimates given here regarding the center of
the affected area are adopted mainly from previ-
ous studies and presented in Appendix A. Future
work is needed to determine more accurately the
intensities of all damaged localities and calculate
the macrocenter (intensity center or barycenter)
or the epicenter.
3.3 Size of the historical earthquakes
Obviously, the original reports do not contain full
coverage of the damage and effects and there
is no simple calibrated damage–intensity rela-
tionship available. With the present understand-
ing, where a magnitude of historical seismicity is
mainly based on macroseismic data rather than on
measurable parameters, it is impossible to assign a
clear magnitude.
The descriptions, both full and partial, vary
from place to place, time to time, and event
to event. Therefore, some large events may be
under-reported and listed as moderate. Similarly,
inflated reports of moderate earthquakes, some
closely timed moderate events, a mainshock fol-
lowed by an intensive aftershock sequence, an
earthquake swarm with several strong events, and
a sequence of strong events may all be described
as one large earthquake. Moreover, factors such
as site effects and directivity that increase damage
are not considered either.
Given such large uncertainties, it was possible
only to estimate the size of the historical event
rather than assign definite or discrete magnitude
grades of modern scales. The present estima-
tions were adopted from historical, geological, and
paleoseismological studies and, if not available,
made by personal judgment. They are more or less
correlated with the broad categories suggested by
Ambraseys and Jackson (1998) as follows: V, very
large event (Ms ≥ 7.8); L, large (7.8 > Ms ≥
7.0); M, moderate (7.0 > Ms ≥ 6.0); and S, small
(6.0 ≥ Ms).
Minimal magnitudes of historical earthquakes
are also important, mainly for the study of small
events such as fore- and aftershocks. Experi-
ence shows that the lowest magnitude of felt
JSeismol
earthquakes is on the order of M3.5–4, given that
the observer is near to the source. As distance
from the epicenter increases, the threshold of felt
earthquakes increases as well.
3.4 Is the historical list reliable?
The shortcomings presented here are inherent
properties of the historical data and may strongly
affect the quantification of aftershock activity.
Future studies may reveal new historical sources
and narrow uncertainties, but nevertheless are not
expected to parameterize descriptive information.
Thus, in our opinion, it is important to extract the
most out of the historical data cautiously and limit
the conclusions accordingly, rather than to ignore
their potential.
Regarding the type of sequence activity, it
all depends on whether the reporter intuitively
grouped several events together and conveyed the
impression that this was a seismic sequence. The
information is given in terms of relative spatial
and temporal relations between each of the events
in the given sequence, rather than the exact source
parameters. In this sense, the accuracy of the ori-
gin time is not essential as long as the event is not
duplicated; the exact location is not critical either
because the simultaneous occurrence of strong
seismic sequences in nearby regions is rare or it
would be reasonable to assume that all the felt
events should be related to the same area. The
magnitude, however, is a significant parameter for
it determines whether a selected event is a fore-,
main-, or aftershock and, thus, typifies the nature
of the seismic sequence. Of course, even though
detailed parameters are much preferred, that it is
generally possible to interpret the historical data
in terms of modern seismology means that there
is still a realistic core. Perhaps the most important
deficiency is that the historical data available are
only a part of what really happened.
4 Synthesis—the historical sequences
Altogether, there were close to a hundred dam-
aging earthquakes in and around the DSFS area
that were identified since about the second cen-
tury B.C. up to the end of the nineteenth century
A.D. (Appendix A). Modern, twentieth century
felt fore- and aftershock activities were added
for comparison. References to all the events and
quotations mentioned hereafter are listed in the
following tables and appendices.
Looking at the spatial and temporal appear-
ance of the earthquakes listed, in most cases,
a single earthquake is mentioned, but interest-
ingly enough, there are reports of sequences of
several events, as well as of lighter shocks that
followed the strongest one. Yet, so far, the cumu-
lative behavior of these sequences has been over-
looked. Here, these sequences are arranged firstly
in a table that presents the duration they lasted
(Table 1). A noted event, possibly the largest fore-
or aftershock, is also listed by the time it preceded
or followed the mainshock. Secondly, the vari-
ous sequences are classified in order to identify
characteristic patterns (Appendix A)andcom-
pare them with modern sequences (Appendix B).
The total number of sequences per duration and
the number of noted aftershocks per time delay
after the mainshock are presented in the summary
of Table 1 and in histograms in Figs. 1 and 3.
Historical patterns with no modern parallels are
also discussed. Interestingly, people related to the
strength and length of the shaking of the main
event, and these patterns are classified as well.
4.1 Single events
Many reports describe single earthquakes with no
mention of any other shocks (Table 2). Present-
day single events in the DSFS area are mostly
of small magnitude, and as seismic networks im-
prove, aftershocks are detected for even smaller
events. On the other hand, almost all mod-
ern large earthquakes that occurred in tectonic
settings similar to that of the DSFS were asso-
ciated with aftershocks. It is therefore reason-
able to assume that large historical earthquakes
were also associated with aftershocks, but the re-
ports skipped these. Therefore, in our opinion,
the recounting of large (damaging) historical
earthquakes as single events is due to incom-
plete reporting. All together, 46 single events were
counted, five of which occurred in the years B.C.
and 41 in those A.D.
JSeismol
Table 1 Duration of inferred foreshock and aftershock sequences
Foreshocks Main event Aftershocks
Wks Days Hours Min Size Date Hours Days Weeks Months Unk
12 34 561 2 3 1 40d2 34 5 678 9 101112>12
M 341 + ⊗
M–L 363
•
,6,⊗
S–M 458 ?
M 526 ? ⊗ 18?
M 528 1,⊗
S 587–8 ?
S–M 634 ⊗
M 713 ⊗
M–L 746 ⊗
M 991 ⊗
M–L 1033
•
, ∼6
••
⊗
L 1068 03
•
,2.5,⊗
M 1091 ?
M–L 1138–9
•• • •
,⊗
S 1156 09 + + ⊗
S–M 1156 10 + + + + + + + ⊗
S–M 1156 12 ⊗
S–M 1157 04 + ⊗
M 1157 07 ⊗
L 1157 08 + + + + + + + + + + 21, ⊗
L 1170 ⊗
L 1202 + ⊗
•
M 1212 ⊗
S 1259 ?
M–L 1344 ⊗
9,
•
M–L 1404 + ⊗
M 1546
••
,⊗
•
M 1705 + ⊗
M 1759 10 ⊗, ?
L 1759 11
••••
⊗
JSeismol
Table 1 (continued)
Foreshocks Main event Aftershocks
Wks Days Hours Min Size Date Hours Days Weeks Months Unk
123456 1 2 3 140d2 345 6 789 10 11 12>12
L 1796 ⊗
+ + 30,
•
L 1822 + 30, ⊗
S–M 1834 ⊗
M–L 1837
•• •
,⊗
L 1872 + + ⊗
Modern events
M 1927 + + + + + + + +
•
+ + 38? ⊗
S 1956 ⊗
+ + M 1969
•
21, ⊗
+ + + S 1983
•
⊗
S 1984 ⊗
+ 12,
•
S 1993 ⊗
L 1995
•
24, ⊗
S 2004
•
AB
•
⊗
Summary of the duration of historical sequences (⊗)
2– 2
A
– 36sequences 5 –1–1–– 2 1 – 52 2 –22 – –11 1 – 3 2 4
Summary of the time delay of the noted aftershocks in historical sequences (
•
)
1 – 2 1 4foreshockand 5 111––– 1 2 2 1– 1 –11 – –1– – – – – –
7 aftershock
Information on the historical and modern events relate to felt events (M > 3.5–4), except 1956, where the minimal threshold is not known, and 1984, for which the
recorded events were possibly of 2 < M < 3; see details in Appendices and .
+Reported activity;
•
noted event, possibly the largest fore- or aftershock; ⊗ reported or inferred end of sequence; Wks weeks; Min minutes; Size estimated magnitude,
see Appendix for explanation; Unk unknown; 40d 40 days
JSeismol
Fig. 1 Duration of
historical aftershock
sequences in the Levant,
in and around the DSFS
(simplified from Table 1)
0
2
4
6
8
Unknown> 12 6 - 12 2 - 5
Months MonthsMonths
40
Days
1
Month
1 - 3
Weeks
1 - 6
Days
Hours
Number of sequences
4.2 Mainshocks
Many descriptions relate to the strength and
length of the shaking of the main event, whether
reported as a single event or as part of a sequence.
4.2.1 Duration of the mainshock
Nine descriptions tell how long the shaking lasted
and, in a few cases, even distinguish phases
within the given event, and it was possible to
arrange these by seconds, minutes, and even hours
(Table 3). Modern earthquakes that “lasted for
two and a half hours,” such as those reported
for March 18, 1068, are not known. Strong mo-
tion of earthquakes on the order of M7 may last
several tens of seconds, and even the 2004, M9.2
Sumatra earthquake lasted no more than about
10 min. Presumably, the immediate strong after-
shocks were thought of as part of the mainshock
and gave the people at the time the impression
that the earthquake lasted for hours.
Some reports describe a very close sequence of
shocks, such as on January 3, 1344: “Two shocks in
close sequence.” It is not possible to further clas-
sify these as a large, shortly delayed aftershock, or
even speculate these as subevents. Nevertheless,
this is invaluable information since it tells us that
some destructive earthquakes may have started
“low” and intensified in later stages (e.g., 1872),
thus giving the people short but precious time to
evacuate into the open.
4.2.2 Strength of shaking
Ten reports described earthquakes “such as has
not occurred before” (e.g., 31 B.C.) and the great
impact of the shaking (Table 4). No tools are
available to quantify terms such as “tremendous
shaking” or “mighty earthquake”; these just tell us
that this was the impression people got at the time.
Even today, in places where a great earthquake
is inevitable, the timing and shaking will most
probably be a complete surprise. Yet, when the
reports tell that “everything had been tossed,” it
is reasonable to assume that the motion was con-
siderably strong. This, of course, may also result
from being at or near the seismogenic source or
even from a significant site effect.
The strongest shake may arrive right at the start
of an event (e.g., 1796 A.D.) or in later phases
(e.g., 1872). Whenever the strong shaking appears,
Table 2 Historical events reported as a single earthquake
Origin time of single historical events
B.C.: 760–750, mid-second century, February 21, 148 (or 130), c. (69–) 65, 31.
A.D.: March 23, 37; c. 47; December 13, 115; c. 127–130; 303 or 304; 348 or 349; 450–457; August 22,502; July 9, 551;
c. 570; 580 or 581; 601–602; 634; June 659; September 659–August 660; June 12, 853–June 1, 854; 972;
November 10, 1002–October 29, 1003; August 21, 1042–August 9, 1043; May 29, 1068; June 26, 1117;
August 17 1140–August 6, 1141; c. 1150; August 1163; October 13, 1284; March 8, 1287; March 22, 1287;
January 11–February 8, 1293; January 13–February 11, 1339; April 9–May 8, 1407; December 29, 1408;
November 8 or 16, 1458; March 9, 1537; February 1557; September 13, 1563; January 4, 1588;
January 21, 1626; March 1719; 1722–1723; April 15, 1726; September 25, 1738
Details and references in Appendix A
JSeismol
Table 3 Durations of the mainshock
Duration Descriptions Comments
Seconds to January 4, 1588 (AMA): A strong shock of earthquake Duration can be correlative
a minute was felt in Cairo, where it was of long duration to magnitude
April 26, 1796 (AF2): Lasted with intermissions for about 1 min
August 13, 1822 (Am3): The main shock happened in three
phases lasting altogether 40 s
January 1, 1837 (Am4): The main shock lasted between 10 and 30 s
Minutes November 29, 528 (GCT): The earthquake...lastedfor1h Canbeasuccessionofstrong
to hours November 24, 847 (GCT, one of the events): Dreadful earthquake, aftershocks immediately
lasted for 3 h after the mainshock
March 18, 1068 (GC): An earthquake...lastedfortwoandahalfhours
August 13, 1822 (Am3): ...themainshockwasfollowedforabout 8 min by
successive shocks, about 30 in all, each of short duration but of
damaging intensity
Number January 3, 1344 (GC): Two shocks in close sequence Can be subevents, triggered
of shocks January 1, 1837 (Am4): The earthquake consisted of two distinct shocks events or immediate strong
about 5 min apart aftershocks
April 3, 1872 (Am3): Between the first shock and the latter part of destructive
shaking, many people managed to run out of their houses into the open
References abbreviations in Appendix D
and as trivial as it can be in our eyes, the his-
torical reports clearly tell that the destruction was
associated with the strong shakings, and this, in
fact, is the essence of antiseismic engineering.
4.3 Foreshocks
Only four accounts report on earthquakes that
preceded the main event by several minutes to
a few months (Table 5), all of which also con-
tained aftershocks. There could also have been
foreshocks that occurred a few minutes or seconds
before the mainshock, but the historical data do
not distinguish them from what happened during
the main event.
The reports do not tell if people did perceive
in real time that a strong earthquake was to
come,butclearlysaythat theyfeltworried:“...a
strong shock was felt in the region: this caused
considerable concern and warned the people of
Table 4 Strength of main shaking
The shaking Descriptions
Strength of shaking 31 B.C. (GCT): Such as has not occurred before
December 13, 115 (GCT): Unusually powerful, tremendous quaking
341 (GCT): Most violent earthquake
April 3, 1872 (Am3): Between the first shock and the latter part of destructive shaking,
many people managed to run out of their houses into the open
Effects May 18–19, 363 (GCT): A mighty earthquake tore up the stones of the old foundation
of the temple
July 9, 551 (GCT): Mountains were uprooted and violently split open
December 5, 1033 (GCT): “We have seen the mountains shake, leap like stags,
their stones broken into pieces, the hillocks swaying to and fro, and the trees bending down. ..
Insomeplacesthewatersinthecisternsreachedthebrim...”
Strong motion? September 13–14, 458 (GCT): Everything had been tossed and terribly shaken
May 20/29, 526 (GCT): Foundations of buildings were struck by thunderbolts, thrown up,
lifted, and collapsed
April 26, 1796 (Am3): In Latakia so violent that almost everything collapsed with the first shock
References abbreviations in Appendix D
JSeismol
Table 5 Possible historical foreshocks
Time before the main event Descriptions
Minutes August 13, 1822, 20:40 (Am3): ...At8h10mpmonAugust13,astrongshockwasfeltin
the region: this caused considerable concern and warned the people of what was to follow.
The main shock happened 30 min later
Hours 749 (or 750, or March 9, 757) (GCT): There was a tremor at night...andeveryone had gone
outofthecitytoprayatthetemple...therewasasudden tremor and the temple collapsed
on top of them. . .
a
May 1, 1212 (GC): Foreshock at sunset 30.4, mainshock at dawn 1.5
November 24, 1705 (AF1, SDM): Three main different sized shocks happened on Tuesday
night: The first caused general panic, the second was the strongest, causing the damage
Modern example: August 3, 1993, 12:43 (a swarm or a sequence) (Ho, GII): The largest
event (M
L
= 5.8) appeared third, 2.5 h after the first one (M
L
= 3.5), and 12 min after
the second one (M
L
= 4.8). The second largest event (M
L
= 5.6) was recorded 4.5 h after
the sequence started
Days, a few weeks August 13, 1822, 20:40 (Am3): Slight shocks, began on August 5 and continued
intermittently until August 12. . .
Modern example: March 31, 1969, 7:16 (Suez Rift) (BMA, Sa1, Ke): Preceded 2 weeks
before by 35 large foreshocks, including 3 M
L
≥ 4
February 3, 1983, 23:30 (Gulf of Aqaba): Preceded 2 weeks before the strongest event in
the swarm
Months February 20, 1404 (GC): Before that, there had been an earthquake (December 18, 1403),
at midday
References abbreviations in Appendix D
a
This event occurred in Mesopotamia, outside the present study area, but is mentioned here as an example
what was to follow. The main shock happened
30 min later...” (in 1822) or “The first caused
general panic, the second was the strongest, caus-
ing the damage” (1705). In the case of 198–199
B.C., although there is no explicit mention of a
preceding event, people realized the worst was
to come and reacted: “. . .but the number of vic-
tims was limited, because it did not happen in a
single shock.” Different instincts, unfortunately,
were not so seismic-proof: “There was a tremor at
night...and everyone had gone out of the city to
pray at the temple...there was a sudden tremor
and the temple collapsed on top of them...”
(749, in Mesopotamia, Guidoboni and Comastri
2005).
Identifying foreshocks in real time is extremely
important because this is a true alarm, and it
seems that, somehow (fear, intuition, cumulative
experience?), people realized that, although they
did not always react right. Omitting the 46 reports
of single events that are believed to be incomplete
and the 13 unresolved clusters, about a ninth (four
out of 35) of the sequences were preceded by
foreshocks! Obviously, this is of great importance
in early warning evaluation and needs further
examination.
4.4 Aftershocks
The appearance of secondary shocks that closely
followed the main shock is common in histori-
cal reports, mostly from the second millennium
A.D., and refers to about a third (35 of 94, six
of which appeared in a storm, see Section 4.5)of
the listed events. Former reports are indirect, and
the occurrence of aftershocks is inferred from the
reaction of the people at the time: “Danger for
three days” (341 A.D.) or “the inhabitants forced
to take refuge in the desert, where they stayed for
forty days” (746 A.D.).
4.4.1 Duration of aftershock sequences
The reported aftershocks are classified by the
time they lasted after the main shock, in
hours, days, months, etc. (Table 6). Large earth-
quakes (M > ∼6) are known to produce in-
tensive aftershock sequences, which may last
JSeismol
Table 6 Duration of sequences that followed historical earthquakes
Duration Descriptions
Several hours 363 (GCT): Event took place on the third hour, and partly on the ninth hour of the night
March 18, 1068, 8:30 (GC): The earthquake was followed by two more shocks within the same
period (two and a half hours?)
Several days December 5, 1033 (GC): For 8 days, the mind has not been satisfied and the soul is not at rest.
On that night (the earth) shook again (December 5–6, 1033); on Friday (December 6, 1033),
as well as on the following night (December 6–7, 1033), the shocks recurred
May 20, 1202 (AM1): The major earthquake was followed by brief shocks towards noon on the
same day, which were slightly felt in Cairo. . . earthquake at Hamat on May 21 was followed
by another shock in the afternoon. Altogether, the shocks lasted for 4 days
May 26, 1834 (AAT): Strong, many aftershocks during 10 days
A month September 634 (GCT): An earthquake with a series of tremors lasted for a month
April5,991(GCT):Theshockswentonrepeatedlytill...(May5)inthesameyear
January 3, 1344 (GC): Two shocks in close sequence. In relation to the worst affected area, it is
recorded that the places concerned were abandoned by their inhabitants for more than a month
November 24, 1705 (AF1, SDM): Light shocks continued to be felt till Ramadan (1 month)
October 30, 1759 (AB): Series of strong aftershocks (time not specified), some of which were felt
as far as Aleppo, that added to the damage. Note that, after a month, this event was followed by
the November 25, 1759, strong earthquake
40 days February 28/March 10, 713 (GCT): Earthquakes began throughout the world and lasted
for 40 days
January 18, 746 (GCT): A strong earthquake in Syria...theinhabitantsforcedtotakerefugeinthe
desert, where they stayed for 40 days. ..
Several months 341 (GCT): Danger for 3 days, shocks for a whole year
to a year May 20 or 29, 526 (GCT): Lasted for 6 days (possibly the fire)...theearthshook for a year. ..
The earthquake continued every day and night for a year and a half without ceasing
October 11, 1138, until June 1139 (Am5, GBC): Destructive seismic sequence until June, 1139.
The main event was followed by three large events on the first day. A total of 80 shocks were felt
during the whole seismic sequence
June 29, 1170, 3:45 (GBCB): The earthquake lasted for 3 or 4 months, or perhaps longer. There
were times when three or four or even more shocks were felt by day or night
May 1, 1212 (GC): Aftershocks for a year
February 20, 1404 (GC): The most violent earthquake at Aleppo was followed by a sequence of
less powerful shocks, which lasted until early July 1404
January 14, 1546 (AK): Then, on March 13, 1546, there was another alarm, the noise of which was
greater before it died out. Then, on May 13, there occurred another shock felt by some people
more than others, apart from the continuous shocks of previous days, some of which occurred at
night and some during the day
November 25, 1759 (AB): Aftershock sequence until August 1760. Damaging shocks on
November 26, December 5, December 12, December 30
April 26, 1796 (Am3): Aftershocks continued to be felt for 2 months
January 1, 1837 (Am4): Aftershocks continued to be felt for almost four (five?) months. Important
are: January 16 widely felt and caused considerable damage in the south, January 22 and 25
reported from the north and caused panic in Damascus. May 20 was reported from the north and
caused considerable damage at Hashbeya
April 3, 1872 (Am3): Aftershocks continued to be felt with decreasing severity throughout
April and May, but did not cease altogether until February 1873
Modern seismicity: July 11, 1927, 13:04 (Av): Twelve 3.5 < M < 5 aftershocks until February 1928,
and two such more until September 1930. Most powerful were on July 17 and February 22, 1928
March 16, 1956, 19:32, and March 16, 1956, 19:43 (ISC, PK): Thirty weak aftershocks until
November 1956
March 31, 1969, 07:16 (in the Suez Rift) (BMA, Sa1, Ke): Mainshock with more than 2,000 events.
Preceded 2 weeks before by 35 large foreshocks including 3 M
L
≥ 4, followed by 19 M
L
≥ 4 in
half a year and four more in the next 15 months until December 1970
JSeismol
Table 6 (continued)
Duration Descriptions
February 3, 1983, 23:30 (Ho, Sa1): A swarm, lasted 8 months, largest event M
L
= 5.3 after 2 weeks,
28 events M
L
≥ 4; 94 events M
L
> 3
August 3, 1993, 12:43 (Ho, GII): A swarm or a sequence? The two largest are M
L
5.8 and 5.6,
in the first hours. Strongest event preceded by a foreshock. Overall, 420 events of M
L
> 3
November 22, 1995, 4:15 (Ho, AT): Mainshock followed by >5,000 aftershocks, largest aftershock
after 3 months on February 26, 1996, Mw = 5.6, most M
L
> 4 occurred in the first 100 days,
afewM
L
> 4 continued for 2 years
February 11, 2004, 8:15 (Sa2, GII): Mainshock with a few tens of aftershocks over half a year, the
largest, M
L
3.7, occurred 2 days later
Several years August 13, 1822 (Am3): It was followed by an aftershock sequence that lasted almost 2.5 years
Unknown period September 13–14, 458 (GCT): When the earthquake ceased, everyone of those who fled regained
of time his confidence
November 29, 528 (GCT): The earthquake that now occurred lasted for 1 h and was
accompanied by a terrible sound. Then He appeared to a pious man, who told the survivors
to write at the top of their doors ‘Christ is with us. Stop’. When this was done, the wrath
of God abated
587 or 588 (GCT): Later shocks. ..
September 26, 1091 (GC): There was an earthquake and 86 towers in the walls of Antioch
collapsed. . . there were numerous earthquakes in the Syrian territory
March 22, 1259 (GC): There were numerous shocks in Syria at the time when the Tartars arrived
References abbreviations in Appendix D
several months and longer, including M3.5–4
events, which are considered as the lowest limit
for felt quakes (Appendix B). Obviously, the
threshold of historical records is also of felt events,
and it is therefore reasonable to accept historical
sequences of similar length.
About a half (17 of 35) of the sequences lasted
a month or less, a sixth lasted 2–5 months, and
a quarter lasted almost a year or longer. In four
reports, it was possible to infer the occurrence of
intensive activity after the main destructive event,
but the time each lasted was not specified. Short
durations are not known in modern times from
in and around the DSFS, and it is reasonable to
assume that the historical record is incomplete
for such sequences. Possibly, these reports may
have related to the immediate strong aftershocks,
neglecting the later and weaker stages of the se-
quence. Only when the observer is far away from
the epicenter does the aftershock sequence seem-
ingly “become” shorter, simply because shaking
decreases with distance. The notion of 1 month or
40 days could be a metaphoric sign of a significant
period of time derived from religious or cultural
views, rather than what actually happened. For
example, many Biblical events lasted “40” days,
nights, years, etc. It therefore should be taken as
a symbolic time frame rather than as a nominal
figure.
Correlating the duration of the aftershock se-
quences with the estimated magnitude of the his-
torical mainshocks (diamonds in Fig. 2), many
short sequences (<40 days) appear in all the
range of magnitudes. Long durations, however,
appear with moderate (M > ∼6)andlarge
(M > ∼7) earthquakes where the historical se-
quences lasted up to about 500 and 1,000 days,
0 200 400 600 800 1000 1200
Aftershock duration (in days)
Estimated mainshoc
k
magnitude
L
M-L
M
S-M
S
Fig. 2 Aftershock duration–estimated mainshock mag-
nitude relations. Diamonds denote historic events and
ellipses signify modern events. The line delineates the en-
velope of maximal duration of the historical aftershock
sequences
JSeismol
Fig. 3 Time delay of the
noted (largest?) historical
aftershocks (simplified
from Table 1)
0
2
4
6
6 - 8
Months
1 - 5
Months
1 - 3
Weeks
1 - 6
Days
Hours
Number of aftershocks
respectively. These maximal durations seem to
correlate, more or less, with the lower bound of
duration of modern aftershock sequences (ellipses
in Fig. 2) and, therefore, can be considered to be
more realistic.
4.4.2 Time delay of the noted (strongest?)
aftershocks
Of special interest after the mainshock is over
is the timing of the strongest aftershock because
it may add to the damage. Indeed, the historical
accounts mention several aftershocks increasing
the damage or panic or being exceptionally strong,
and these can sometimes be suspected to have
been the strongest in the sequence and possibly
even to have been the mainshock. All together,
there were 18 notable aftershocks mentioned in
seven sequences. Some sequences thus include
several aftershocks, but it was not possible to de-
termine which one was the strongest. Therefore,
all the “noted aftershocks” were included in the
evaluation. This, of course, biases the examination
but eventually seems not to have affected the main
conclusion much. Table 1 and Appendix A,and
Figs. 3 and 4,presentthetimedelayofthemost
significant aftershocks of the historical events.
Most of the noted aftershocks appeared just
after the mainshock and in the following days
and weeks, but several others were delayed a few
months, or even as much as 8 months. There
is no simple correlation between the time delay
and the mainshock magnitude (Fig. 4), even for
aftershocks in modern times, and no simple expla-
nation either. The lack of correlation may result
from the incompleteness and bias of the database;
nevertheless, that strong aftershocks were delayed
days, weeks, and months after the mainshock is a
significant conclusion.
4.5 Earthquake storms
Between September 1156 and May 1159, there
was an intensive sequence of six destructive earth-
quakes in northwestern Syria, each followed by
many smaller events, some of which were strongly
felt (Ambraseys 2004; Guidoboni et al. 2004a).
This series, which was termed “paroxysm” by
Ambraseys (2004) and “destructive seismic crises”
by Guidoboni et al. (2004a), can also be called a
seismic storm according to Utsu’s (2002) terminol-
ogy. The mainshocks of the major sequences oc-
curred on September 27, 1156; October 13, 1156;
December 9, 1156; April 2, 1157; July 5 1157;
and August 9–September 7, 1157, the last being
the largest one. Close sequences also occurred
0 50 100 150 200
Delay time (in days)
Estimated mainshock magnitude
1927
1995
1927
L
M-L
M
Fig. 4 Delay time–estimated mainshock magnitude rela-
tions. Diamonds denote historic events and stars denote
modern ones
JSeismol
Table 7 Unspecified clusters of historical earthquakes
Unspecified sequences
199–198 BC (GCT) “but the number of victims was limited, because it did not happen in a single shock.”
419 (GCT) Great earthquakes in the East
March 9, 757 (Am6) Four events, or a single one followed by three significant aftershocks
January 5–December 25 835 (GCT) The earth shook for 40 days. Possibly a duplication of 713?
November 24, 847 (GCT) A dreadful earthquake at Damascus...Theearthquake reached Antioch...Thenit
reached Mawsil. . .
December 30, 859–January 29, 860 Earthquakes which.. .
(GCT)
January 9, 951–May 28, 952 (GCT) “...thereweremanyearthquakes tremors in Aleppo and other cities. They lasted for
fortydays,...”
July 30–August 27, 1063 (GC) There were earthquakes.
May 19–June 18, 1094 (GC) In that month, there was a series of many earthquakes in the Syrian territories lasting
for a long time
September 28, 1151 (GC) Earth shook three times
February 1, 1152 (GC) Earth shook three times
February 1287, 2nd half (GC) A series of earthquakes
December 1795 (AF2, SDM) Two shocks
References abbreviations in Appendix D
between 526 and 528 around Antioch and in 1759
in southern Lebanon and northern Israel, but
these were not as notable as the 1156–1159 storms.
4.6 Unspecified clusters
There are 13 reports such as “numerous earth-
quakes,” “great earthquakes,” “earth shook three
times,” or just “there were earthquakes,” with no
mention of a specific event (Table 7). This is insuf-
ficient to distinguish the various categories used
here, and they could be any type of a sequence, in-
cluding a regular mainshock–aftershock sequence,
a swarm, or a storm. Moreover, they could also
be a compilation of several different earthquakes
from remote locations into a single cosmic event,
such as the “circum-Mediterranean” 881 A.D.
earthquake: “there was a strong earthquake in
Syria, Egypt, some parts of Mesopotamia, North
Africa and Andalusia,” and this is why it was not
included in this study.
4.7 Missing sequences
A swarm-like sequence, in which the magnitude
slowly increases to a maximum that is not much
stronger than the preceding or following events,
was not identified. This is not to say that swarms
have not occurred, but only that the resolution
and magnitude determination of historical data is
insufficient to recognize such clusters.
5 Conclusions
Historical and modern accounts report about a
hundred damaging earthquakes in the Levant dur-
ing the last 22 centuries (Appendices A and B).
Obviously, the complex nature of the historical
data may bias the classification of earthquake
sequences, and therefore, reinspection of the
original data is desired in order to improve
the reliability of the present listings (tables and
Appendices A and B). This is an extremely large
task, certainly beyond the scope of the present
work. Yet, although subjective, interpretative, and
incomplete, past descriptions enable consider-
able expansion of the seismic experience, almost
20-fold the limited and short time span of the
instrumental period. Of course, as new informa-
tion arises and better understanding develops,
reevaluation of the original sources will be well
worth doing. Overall, comparing the pattern of
seismic sequences as it appears in the now existing
JSeismol
historical record with that of today, it is possible
to distinguish several modes:
(a) Single events: Forty six accounts (about a
half) reported single strong earthquakes with
no additional shocks mentioned (Table 2).
Such events are not known in modern seis-
mology and may just be the result of incom-
plete reports.
(b) Mainshocks: Thirty five accounts reported
a “multievent” sequence that included one
predominant earthquake, the mainshock,
which caused most of the damage (Table 1).
Some shakings, mostly in recent centuries,
were reported to last for several tens of sec-
onds, which is reasonable. Other past events
were said to last a few hours, a duration
that is unexplainable, unless the mainshock
was merged with its immediate strong after-
shocks (Table 3). Commonly, the earthquake
surprised the inhabitants “such as has not
occurred before,” sometimes with a violent
onset, “everything collapsed with the first
shock” (e.g., 1796), and occasionally with
the later phase being the destructive one,
allowing invaluable spare time for evacua-
tion (in 1822) (Table 4).
(c) Foreshocks: Earthquakes that preceded the
major shock were reported in four sequences
only (11%), ranging from minutes to several
weeks (Table 5). In some cases, people re-
acted on time, evacuated, and saved their
lives. This may indicate a true fear, a good
instinct, and even an educated experience,
rather than a clear understanding of this
natural phenomenon.
(d) Aftershocks: Thirty five mainshocks were
followed by additional earthquakes, lasting
for hours, days, weeks, months, and some-
times more than a year (Table 6). Several
sequences included tens of felt tremors, some
of which added to the damage. Since, in
a given seismotectonic area, larger events,
supposedly, produce more numerous and
stronger aftershocks (Bath and Omori lows),
it is reasonable to assume that longer his-
torical sequences followed larger mainshocks
(Fig. 2). Hence, historical events associ-
ated with a long sequence of aftershocks
(e.g., A.D. 341, 526, 1212) could have been
stronger than previously assumed.
(e) Largest aftershocks: Close to half (eight of 18
events) of the notable aftershocks appeared
within a week after the mainshock, but some
were delayed by as much as a month (5/18
events), half a year (4/18 events), and even
more (Figs. 3 and 4).
(f) Seismic storms: A series of six successive
mainshock–aftershock sequences appeared
between September 1156 and May 1159 in
northwestern Syria. Smaller series may have
occurred there also in 526–528, and in 1759 in
southern Lebanon and northern Israel.
(g) Unspecified clusters: The data regarding 13
sequences is insufficient to further classify
them (Table 7).
(h) Swarms: This was identified only in the
modern activity; however, though they were
unidentifiable because of the lack of data and
details, they cannot be excluded from the
historical record.
Acknowledgements Thanks are due to Gadi Shamir,
Vladimir Lyakhovsky, and Yariv Hamiel, all from the Geo-
logical Survey of Israel (GSI), for fruitful discussions. Ben-
nie Begin and Rivka Amit, GSI, are appreciated for helpful
comments. Bevie Katz helped with editing the text. Critical
reading and constructive suggestions by two anonymous
reviewers helped in improving the manuscript significantly
and are much appreciated. This work was partly funded by
the Inter-ministerial Steering Committee for Earthquake
Readiness in Israel.
Appendix A
Table 8
Appendix B
Table 9
JSeismol
Table 8 Seismic sequences associated with historical damaging earthquakes in the Levant
Date Affected area Size Ref. Type of Pattern of seismicity
sequence Foreshocks Main Aftershocks Noted after
760–750 B.C. Jerusalem ? Am6, GCT, ZAB S 1
199–198 B.C. Sidon M AW, GCT, Ka2 C Several
Mid 2nd cent. B.C.? Nearby Sidon M AW, GCT, GMDH, Ka2 S 1
February 21, 148, Antioch M AW, GCT, Ka2 S 1
afternoon B.C.
(or 130 B.C.)
c. (69–) 65 B.C. Syria and Antioch M AW, GCT, Ka2 S 1
31 B.C. early spring Judea M Ka1, Ka2, GCT, RH S 1
A.D. March 23, Antioch S (–M) GCT S 1
37 early in the morning
c. 47 Antioch S (–M) GCT S 1
December 13, 115, morning Antioch M (–L) AJ, GCT, Ka1, MGS S 1
c. 127–130 Nicopolis, Caesarea M Ka1, GCT S 1
303 or 304 Tyre and Sidon M GCT S 1
341 Antioch (S–) M GCT A 1 1 year 3 days?
348 or 349 Berytus (Beirut) S (–M) GCT S 1
May 18–19, 363, night Judea and Samaria M–L Am8, GCT A 1 Tremors 1, after 6 h
(and possibly Petra)
419 Palestine, Jerusalem M GCT, Ka1 C Several
450–457 night Tripolis S (–M) GCT S 1
September 13–14, 458, night Antioch S – M GCT A 1 Unknown
length
August 22, 502, night Ptolomais, Tyre, Sidon M GCT S 1
May 20 or 29, 526, midday Antioch and Seleucia M GCT A 1 1–1.5 year 6 days
November 29, 528 Antioch and Laodicea M GCT A 1 (lasted Unknown Lasted 1 h?
1h?) length
July 9, 551, day Beirut, off coast Lebanon M–L AJ, Am2, Am8, AMA, S 1 (or 2)
GCT, DEK, ETSK, DSM
c. 570 Around Antioch M GCT S 1
580 or 581 noon Daphne and Antioch S (–M) GCT S 1 Tremors?
587 or 588 late October, night Antioch S (–M) GCT A 1 Unknown
length
JSeismol
Table 8 (continued)
Date Affected area Size Ref. Type of Pattern of seismicity
sequence Foreshocks Main Aftershocks Noted after
601–602 day Syria, Rum (Cilicia) M GCT S 1
September 634 Palestine, Jerusalem S–M GCT A 1 1 month
634 Aleppo S GCT S 1
June 659 Jerusalem M GCT S 1
September 659–August 660 Jericho S–M GCT S 1
February 28 or March 10, 713, Antioch, Allepo, M GCT A 1 40 days
middle of the night Qenneshrin
January 18, 746, morning Between Jerusalem M–L AAT, Am7, Am8, A 1 40 days
and Tiberias GCT, Ka2, MHH,
RH
March 9, 757, midnight Jerusalem S–M AAT, Am6, GCT C 1 3 events? 3 events?
(or 749, or 750)
January 5–December 25, 835 Antioch S GCT C 40 days
November 24, 847 morning a. Damascus a, b, c.: GCT C Several events,
b. Antioch S–M a: lasted 3 h
c. Mawsil
June 12, 853–June 1, 854, night Tiberias S GCT S 1
December 30, 859– Laodicea–Antioch M–L AAKY, Am8, AMA, C 1 (several?)
January 29, 860 An, GCT
June 9, 951–May 28, 952 Aleppo M GCT A 40 days
972 Around Antioch S–M GCT S 1
April 5, 991, night Damascus and M GCT A 1 1 month
Ba’albek
November 10, 1002– Syria at the time M AJ, AMA, GC S 1
October 29, 1003
December 5, 1033, before sunset Jericho–Nablus– M–L GC A 1 8 days December 5–6;
Akko December 6,
December 6–7,
all 8 days?
August 21, 1042–August 9, 1043 Tudmur, Palmyra S–M GC S 1
July 30–August 27, 1063 Tripoli M GC C Several
March 18–August 30, 1068 Ayla, or Hejaz L AJ, AZP, GC, ZAP A 1 lasted 2.5 h 2 events in 2.5 h?
May 29, 1068 Ramla? M GC S 1
September 26, 1091, night Antioch M GC A 1 Numerous
JSeismol
May 19–June 18, 1094 Syrian territories S GC C Many
in 1 month
June 26, 1117, night Tyre (Scandelion) S GC S 1
Sequence from: Around Azrab, M–L Am5, GBC, GC A 1 (3 80 events in October 11: 3
October 11, 1138, south of Triple subevents?) 8 months events,
afternoon, to June 1139 Junction of east October 14,
Anatolian fault 3 in October 27,
several on
October 29
and October 31,
terrifying June 21
August 17, 1140–August 6, Sheizar S SDM S 1
1141
c. 1150 (1160?) Jerusalem S–M GC S 1
and Jericho
September 28, 1151, night Busra S Am5, GC C 3 shakes
February 1, 1152, shortly Busra S Am5, GC C 3 shakes
before dawn
September 27, 1156 Damascus? S Am5, GBC, GC A, St 1 Several
October 13, 1156, night Afamiyah S–M Am5, GBC, GC 1 Many
(between Allepo
and Hamat)
December 9, 1156, night Allepo S–M GBC, GC 1 Many
April 2, 1157, night Afamiyah S–M GBC, GC 1 Several
towards dawn
July 5, 1157, morning Hamat M Am5, GBC, GC 1 Several
August 9–September Aleppo, Hamat, L Am5, GBC, GC 1 Many
7, 1157 Hims
August 1163 Antioch M GC S 1
June 29, 1170, 3:45 UT Antioch, Tripoli L Am5, GBCB, A 1–2 >3–4 months
Bakaa Valley, GC, MGS
Allepo
May 20, 1202, 2:40 UT Lebanon, Syria, L AB, AJ, AMA, A 2 shock or 4 days 4 days?
Israel: Akko, AM1, AM2, Am8, 2 subevents
Tyre, Baniyas, DKT1, DKT2,
Safad. . . EMA, GC, MAE
JSeismol
Table 8 (continued)
Date Affected area Size Ref. Type of Pattern of seismicity
sequence Foreshocks Main Aftershocks Noted after
May 1, 1212, night/dawn Ayla, Shubak– M (–L) AMA, Am8, GC, F + A1,∼12 h 1 1 year
Karak KAD before
March 22, 1259, night Damascus S GC A 1 Numerous
October 13, 1284 Damascus S GC S 1
February 1287, 2nd half a, between Zafad S GC C Series
and Hims; or b, near
Hims; or c,
near Zafad
March 8, 1287 Near Hims S GC S 1
March 22, 1287 Near Laodicea S GC S 1
January 11–February 8, 1293 Karak, Tafila M (–L) GC S 1
January 13–February 11, 1339 Tripoli S GC S 1
January 3, 1344 Aleppo M–L GC A 2 closed >1 month
February 20, 1404 Aleppo and Tripoli M–L AB, GC F + A 1, on 1403 12 18 1 5 months 10 days?
April 9–May 8, 1407 Antioch S GC S 1
December 29, 1408 Shughr Bakas M AAKY, S 1
and Balatunus AB, GC
November 8 or 16, 1458 Karak M Am8, AMA, GC, KAD S 1
March 9, 1537 Antioch S AK S 1
January 14, 1546, afternoon Jordan Valley, M AK A 1 >4 months March 13,
Jerusalem to Nablus May 13
February 1557 Jerusalem S AK S 1
September 13, 1563, dawn Damascus S AK, SDM S 1
January 4, 1588, 13:00 Aila and Tabuk M AMA S 1
January 21, 1626 Gaziantep, Aleppo, M AF2, SDM S 1
Hama
November 24, 1705, night Yabrud, Al-Qastal, M AF1, GMDS, F + A 1 1 1 month
Damascus SDM
March 1719 Aleppo S AF2 S 1
1722–1723 Aleppo S AF2 S 1
April 15, 1726 Jum, NW of Aleppo S–M AF2 S 1
September 25, 1738 Amanus M AF2, SDM S 1
JSeismol
October 30, 1759, 03:45 LT Zafad and Qunaitra M AB, DKT1, A 1 Series
Ka1, MRH
November 25, 1759, 19:23 LT Litani and Bakaa L AB, AJ, Am8, A 1 >9 months November 26,
DKT1, GMDS, Ka1 December 5,
December 12,
December 30
December ??, 1795, 14:10 Aleppo S AF2, SDM C 2
April 26, 1796, 9:05 Latakia L AB, AF1, AF2, AJ, A 1, for 1 min 2 months
Am3, SDM
August 13, 1822, 20:40 Gaziantep, Aleppo, L AB, AJ, Am3, SDM F + A 8 days slight shocks, 1 in 3 phases 2.5 years 30 in 8 min
Han Shekhum 1 strong half hour in 40 s
before
May 26, 1834, 4:00 Jerusalem and S–M AAT, BM, Shal A 1 10 days
Bethlehem
January 1, 1837, 14:34 Southern Lebanon M–L AJ, Am4, Am8 A 2 in 5 min, 4–5 months January 16,
10–30 s ea. January 22,
January 25,
D
May 20
April 3, 1872, 7:40 Antakya L AAKY, AB, AJ, A 1 in 2 phases 10 months 2 months
Am3, Am8
This table shows a list of historical earthquakes that caused damage in the Levant, most of which possibly originated from the DSFS and nearby structures. See detailed
explanation in the text, Section 3. Under the “date” column, events are marked by time of occurrence, as detailed as known, by the year, month, day, and part of
the day or night. The “affected area” column shows locality of the most severely affected area. This is not necessarily the epicenter zone. Under the “size” column,
estimated size of the earthquake is given in the broad categories as suggested by Ambraseys and Jackson (1998). Estimations were taken from historical, geological, and
paleoseismic studies and, if not available, were made by personal judgment. The “ref.” (references) column shows studies used for that event. Most of these works are
based on primary sources; some others rely on secondary sources or paleoseismic studies. See abbreviations in Appendix . The “pattern of seismicity” column shows
the sequence of events. Foreshocks are noted by number of events and period of time they preceded the main event; main(shocks) are given by number of events and
the time lasted; aftershocks are described by number and length of time they lasted; and noted after(shocks) are mentioned by number, date of occurrence, and the
time delayed.
L large (7.8 > Ms ≥ 7.0), M moderate (7.0 > Ms ≥ 6.0), S small (Ms < 6.0), A aftershock(s), C cluster, F foreshock(s), S single, St storm
JSeismol
Table 9 Pattern of seismicity associated with significant modern earthquakes in the Levant
Date Affected Size, Ref. Type of Pattern of seismicity
area Mag. sequence Foreshocks Main Aftershocks Noted after
July 11, 1927, 13:04 Northern M, M
L
= 6.2 Av, ABSN, 1A 1 12 in 7 months, 2 more M
L
4.5 on July 17, M
L
Dead Sea SAN in next 29 months? 5.5 on February 22, 1928
March 16, 1956, 19:32 and Southern S M
L
= 5.2, 5.5 ISC, PK A 2 30 weak events, lasted
March 16, 1956, 19:43 Lebanon until November
March 31, 1969, 7:16 Shadwan, M, M
L
= 6.6 BMA, Ke, P + A 35 in 2 weeks, 3 1 >2,000;19M
L
≥ 4 in M
L
≥ 5onApril8and
(Gulf of Suez) Gulf of Suez Sa1 M
L
≥ 4 half a year; 4 in next September 26
15 months
February 3, 1983, 23:30 Gulf of Elat S, M
L
= 5.3 Ho, ISC, P + A 2 weeks 1 in swarm 8 months, 28 events
(Aqaba) Sa1 M
L
≥ 4; 94 events
M
L
> 3
August 24, 1984, 6:02 Haifa S M
L
= 5.3 GII, HvES A 1 Five 2 < M
L
< 3
aftershock within
10 days
August 3, 1993, 12:43 Gulf of Elat S, M
L
= 5.8, 5.6 Ho, GII, P + A M
L
= 3.5,2.5h 2,ina 420ofM
L
> 35
(Aqaba) ISC before the swarm? in 5 months
mainshock
November 22, 1995, 4:15 Gulf of Elat L, M
L
= 6.2 AT, Ho A 1 > 5,000, most
D
in February 26, 1996,
(Aqaba) Mw = 7.2 100 days, a few Mw = 5.6
M
L
> 4 continued
for 2 years
February 11, 2004, 8:15 Northern S, M
L
= 5.2 GII, Sa2 A 1 A few tens in half M
L
3.7 on February 13
Dead Sea a year and July 9
This table shows a list of significant modern (twentieth century) earthquakes that occurred in the Levant and were associated with fore- and aftershocks. Origin time,
affected area, magnitude, and pattern of seismicity were taken from the references mentioned (see abbreviations in Appendix ). Sizes of the events areinterms
suggested by Ambraseys and Jackson (1998)
JSeismol
Appendix C
Comments regarding some selected events
that were considered in this study
In general, only historical events that were re-
ported to damage at least one site or cause sig-
nificant effects were included. Earthquakes that
were reported to have caused no damage, based
only on archaeological data, found dubious and in
need of further investigation, or seen only in de-
formed layers (KAE, MAB), were not included.
The area of interest is the Levant, in and around
the DSFS. See the text, Section 3, for explanation.
The list is partly based on Salamon et al. (2007,
electronic supplementary) and should be consid-
ered as preliminary.
760–750 B.C. Also known as Amos’s,
Zechariah’s or Uzziah’s earthquake. It was
possibly associated with notable effects. Am6:
“...anevent the date, location and magnitude
of which cannot be assessed...” relating, for
example, to AFF and BM (October 11, 759
B.C., evening, M
L
= 7.3).
525 B.C. This was listed by Am2 and BM as
M
L
= 7.5 that destroyed Sur and Sidon. How-
ever, we could not trace back the ancient
sources that reported this event.
Second century B.C. Following AW, GCT and
Ka2, the following interpretation is suggested:
1. 199–198 B.C., AW, GCT: Earthquake in
Sidon, Phoenicia and Syria. Ka2: “famed
emergence of the island of Hiera in 198
B.C.”
2. Mid-second century B.C., We follow
Ka2, although AW and GCT suggested
the 199–198 B.C earthquake in Phoenicia
and Syria.
3. February 21, 148 or 130 B.C., AW, GCT:
Earthquake in Antioch. Ka2: possibly
146 B.C. or 140 B.C.
4. February 28, 92 B.C., Ka2: False
earthquake, “imported into the Israeli
catalogues from elsewhere in Eastern
Mediterranean” (e.g., by Shal, BM,
AAT). It was probably taken from the
mid-second century B.C. earthquake.
5. Ca. 90 B.C., AW, Ka2: Earthquake in
Apamea Kibotos (Phrygia, Asia Minor).
GCT: The Apamea Kibotos (Turkey)
occurred before 88 B.C.
69–64 B.C Ka2: Occurred in Antioch and later
“...imported into the Israeli catalogues...”
and placed at 64 B.C. in Jerusalem (e.g., by
AAT, BM). GCT date this event to 65 B.C.
and AW to 69 B.C.
31 B.C. Ka2: This was a moderate event rather
than a strong one, as was previously suggested
by others (e.g., AAT, BM: for September 2,
31).
27–20 B.C. GCT, Ka1: Affected eastern
Mediterranean, possibly Lydia, Phrygia,
and the Aegean.
A.D. 19 Listed by several authors (e.g., AAT,
BM, Si, Wi) to affect Sidon and the Lebanese
coast. However, the earliest reference we
could find for this event is Ar, who does not
refer back to any specific source. It is possible
that Ar mixes up the earthquake of 17 A.D.,
which hit several cities in present-day western
Turkey (and not of Bithinia), with another
earthquake that occurred in Bithinia in 29
A.D. (or perhaps in 32 A.D.) and was also
suggested by Ar to affect Judea and Jerusalem
(see also GCT).
33 AAT: Slight damage, local effect, and ap-
parently small. Wi discusses it in light of the
Crucifixion event. We follow Am6: “...the
earthquake at the Crucifixion is a spurious
physical event.”
90 This is one of the several events that are seen
only in deformed layers (e.g., MAB). Princi-
pally, a remote and large earthquake could
have produced the same deformed layer like a
close and small (M < 6) one. Such events were
not included in the list.
112 AMA: “Archaeological evidence suggests
early second century destruction at...,” how-
ever, we did not find written sources for this
event.
303–306 AAT and BM report an earthquake in
306. We follow GCT, who mention an earth-
quake in 303/304 in Sidon, Tyre, and Syria.
348/9 GCT: Probably between September 1,
348, and August 31, 349.
JSeismol
362–363 Several events are mentioned in this
time span. Following GCT, only the May 18–
19, 363, event is introduced.
447 Based only on archaeological evidence from
Hammat Gader (AAT). GCT: Earthquake on
the night of January 26, 447, in Constantino-
ple, Turkey, and other places.
475 GCT: Gabala, Syria, no mention of damage.
Ca. 500 Ka1: “Verification of time and spread
across Palestine should be attempted.”
532 GCT: Antioch, no damage.
565 Mentioned by BM, who relies on Si and
Ws1. Ws1 made a systematic error as regards
the chronology and did not convert the dates
of the Egira into the Julian calendar, so 30
entries of his catalogue are brought forward
by about six centuries. Willis himself (Ws2)
and, subsequently, Am1 had pointed out this
error, but these corrections were evidently
not assimilated by the subsequent catalogue
makers. In this case, the year 565 of the Egira
corresponds to the period September 25,
1169–September 13, 1170; thus, it is the large
earthquake of June 29, 1170.
634 GCT: “Although the tremors which struck
Jerusalem and Aleppo presumably occurred
in the same year (634 A.D.), there may have
been two separate earthquakes, since the two
cities are a great distance apart.”
717 GCT: in Mesopotamia.
746–750 For this period: BM and Shal, an earth-
quake on January 18, 746; AMA, an earth-
quake on January 18, 747; and AAT, an
earthquake on January 18, 749. We follow Ka2
and Am7, who suggest the occurrence of two
earthquakes (at least) of which “The second
earthquake, which occurred in 749 or early in
750, affected only Mesopotamia and presum-
ably the adjacent part of northern Syria.”
May 3, 765 (756/758–775) Ka1: Needs further
investigation.
808 (775–780) Ka1: Needs further investigation.
January 5–December 25, 835 GCT: “The earth
shook for forty days...”;however, there is no
mention of a single notable event, and it is
therefore defined as a cluster.
June 12, 853–June 1, 854 AAT: 853+
May 16, 881 The affected area is too large to
conclude of a specific event.
1016 Ka1: Needs further investigation.
March 6, 1032 GC: confused with March 6, 1033,
which occurred in Constantinople.
January 4, 1034 GC: Arabic sources gave the
date of December 5, 1033, as January 4, 1034.
AAT: One of the main shocks during 1033/4
winter earthquake swarm.
March 12, 1036–March 11, 1037 GC: The earth-
quake occurred in Cilicia, southern Turkey,
west of the Iskendrun Bay.
1060 Ka1: “It is possible that felt reports for
years 1060–1063 and possibly even later events
were fused together” regarding damage in
Jerusalem. Needs further investigation.
1063+ AAT mention epicenter in Antioch and
destruction in Elat and may have confused
reports from the event of 1063 in Tripoli,
Lebanon, with reports from 1068 in Elat.
1067–1070 Several earthquakes are mentioned
during this time period (e.g., AAT, BM), and
they are mostly attributed to the event that
occurred on March 18, 1068, in southern Is-
rael. AMA, GC: The dates of April 20, 1067;
November 11, 1067; April 20, 1068; February
25, 1070; March 18, 1168; and 1169 are dupli-
cations and misreports of the earthquakes that
actually occurred on March 18 and May 29,
1068.
April 18, 1086–April 7, 1087 Eastern Syria, Iraq,
Mesopotamia, probably outside the region of
this study.
May 19–June 18 1094 GC: “there were earth-
quakes night and day...”; however, there is
no mention of a single notable event, and it
is therefore defined as a cluster.
December 30, 1097 GC: No explicit mention of
damage.
December 24, 1105 GC: Jerusalem, no damage.
1113–1117 Several events are mentioned to have
occurred in southern Turkey and northern
Syria and around Jerusalem in this time period
(e.g., Am5, AAT, BM, GC):
1. The two events of July 18, 1113, and
August 9, 1113, as mentioned by AAT,
are below the damage threshold consid-
ered for this work and were not included.
2. According to Am5, the strongest earth-
quake occurred on November 29, 1114;
JSeismol
it was preceded by foreshocks on August
10, 1114, and November 13, 1114, and
they “...are clearly associated with the
EastAnatolianfaultzone...”Therefore,
they were not included.
3. GC (as well as AAT) suggest that the
August 10, 1114, earthquake may have
occurred in the region of Jerusalem caus-
ing no damage and the other two in
southern Turkey (GC event of Novem-
ber 29, 1115, is November 29, 1114, of
Am5, and possibly of December 25, 1115,
of AAT).
4. GC: On June 26, 1117, an earthquake oc-
curred in southern Lebanon, causing lim-
ited damage only. AAT considered this
event to have happened in Jerusalem.
We follow GC, who refer to original
sources.
1156–1159 Detailed discussion in Am5, GBC,
and GC.
1201+ Is excluded since it “. . . may be identical
to the following one,” i.e., 1202 (AAT).
October 1, 1261–September 30, 1262 GC: Earth-
quake in Syria, no mention of damage.
April 17, 1269 GC: Occurred in Cilicia, central
and southern Turkey, probably northwest of
the Iskendrun Bay, not included here.
May 1, 1312 AMA: This event was taken by BM
from May 1, 1212.
January 20, 1322 GC: in Damascus, no damage.
September 7, 1366–August 27, 1367 GC: Safad,
Israel, no damage.
September 20, 1399 GC: Damascus, slight earth-
quake, no damage.
November 16, 1402/3 AM2: erroneous location
in Syria. It actually occurred in the Gulf of
Corinth, Greece.
December 18, 1403 GC: In Aleppo, northwest-
ern Syria, but did not cause any damage.
November 5–December 4, 1404 GC: In Aleppo,
northwestern Syria, but did not cause any
damage.
1456–59 Between one and three events. We fol-
low GC and include only the event of Novem-
ber 8 or 16, 1458.
May 1481 AAT: An earthquake in Syria and
Palestine. We follow GC, who mention a large
earthquake in the southern Aegean on May 3,
1481.
March 29–April 28, 1484 GC: In Aleppo, no
mention of damage.
1534 This is very probably a duplication of
the earthquake of January 14, 1546. AK:
“Arvanitakis (1904), on the authority of
Dositheos (1715), dates the event to 1534,
and Willis (1928) copies the earthquakes of
1534 and 1546 from Arvanitakis (1904)and
Perrey (1850), respectively, thus duplicating
the event. Sieberg (1932) and later authors
[...]addnothingbutconfusion.”
January 7, 1537 AMA, SDM: Eastern Mediter-
ranean (Damietta in Egypt, and Antioch in
Syria), possibly an association of two events,
needs further investigation.
October 10, 1568 AF2, SDM: Damage in
Latakia, Syria, also felt in Cyprus. A possible
location between Syria and Cyprus. It
therefore seems to be too small and outside
the region of interest.
February 1656 AMA: “...strong earthquake in
Tripoli in Libya destroyed almost half its
houses...” Later authors (BM, AAT) place
this event in Tripoli in Syria.
1712 AAT: Limited damage in Jerusalem only,
needs further investigation.
July 21, 1752 Earthquake is mentioned by many
writers, probably starting from Si; however, as
noted by Am2, “no authority is quoted.” This
event should be further studied.
1801–1802 Ka1: Needs further investigation.
1896 SDM, AAT: Several earthquakes in Syria
(February 20, May 12, May 14, June 29), no
damage.
Appendix D
Reference abbreviations
AAKY Akyuz et al. (2006)
AAT Amiran et al. (1994)
AB Ambraseys and Barazangi (1989)
ABSN Avni et al. (2002)
AF1 Ambraseys and Finkel (1993)
AF2 Ambraseys and Finkel (1995)
AJ Ambraseys and Jackson (1998)
JSeismol
AK Ambraseys and Karcz (1992)
AM1 Ambraseys and Melville (1988)
AM2 Ambraseys and Melville (1995)
AMA Ambraseys et al. (1994)
Am1 Ambraseys (1962a)
Am2 Ambraseys (1962b)
Am3 Ambraseys (1989)
Am4 Ambraseys (1997)
Am5 Ambraseys (2004)
Am6 Ambraseys (2005a)
Am7 Ambraseys (2005b)
Am8 Ambraseys (2006b)
An Antonopoulos (1980)
Ar Arvanitakis (1904)
AT Al-Tarazi (2000)
Av Avni (1999)
AW Ambraseys and White (1997)
AZP Amit et al. (1999)
BM Ben-Menahem (1991)
BMA Ben-Menahem and Aboodi (1971)
DEK Daëron et al. (2004)
DKT1 Daëron et al. (2005)
DKT2 Daëron et al. (2007)
DSMP Darawcheh et al. (2000)
EM El Mrabat (2005)
EMA Ellenblum et al. (1998)
ETSK Elias et al. (2007)
GBC Guidoboni et al. (2004a)
GBCB Guidoboni et al. (2004b)
GC Guidoboni and Comastri (2005)
GCT Guidoboni et al. (1994)
GII Geophysical Institute of Israel (2007)
GMDH Gomez et al. (2003)
GMDS Gomez et al. (2001
)
Ho Hofstetter (2003)
HvES Hofstetter et al. (1996)
ISC International Seismological Centre
(2001)
Ka1 Karcz (1987)
Ka2 Karcz (2004)
KAD Klinger et al. (2000)
KAE Ken-Tor et al. (2001)
Ke Kebeasy (1990)
MAB Migowski et al. (2004)
MAE Marco et al. (1997)
MGS Meghraoui et al. (2003)
MHH Marco et al. (2003)
MRH Marco et al. (2005)
PK Plassard and Kogoj (1968)
RH Reches and Hoexter (1981)
Sal Salamon (1993)
Sa2 Salamon (2005)
SAN Shapira et al. (1993)
Shal Shalem (1956)
Si Sieberg (1932)
SDM Sbeinati et al. (2004)
Wi Williams (2004)
Ws1 Willis (1928)
Ws2 Willis (1933)
ZAB Zilberman et al. (2004)
ZAP Zilberman et al. (2005)
References
Akyuz HS, Altunel E, Karabacak V, Yalciner CC (2006)
Historical earthquake activity of the northern part
of the Dead Sea Fault Zone, southern Turkey.
Tectonophysics 426(1–2):281–293. doi:10.1016/j.tecto.
2006.08.005
Al-Tarazi E (2000) The major Gulf of the Aqaba
earthquake, 22 November 1995—maximum inten-
sity distribution. Nat Hazards 22:17–27. doi:10.1023/
A:1008109810031
Ambraseys NN (1962a) A note on the chronology of
Willis’s list of earthquakes in Palestine and Syria. Bull
Seismol Soc Am 52:77–80
Ambraseys NN (1962b) Data for the investigation of the
seismic sea-waves in the Eastern Mediterranean. Bull
Seismol Soc Am 52:895–913
Ambraseys NN (1989) Temporary seismic quiescence:
SE Turkey. Geophys J 96:311–331. doi:10.1111/j.1365-
246X.1989.tb04453.x
Ambraseys NN (1997) The earthquake of 1 January 1837
in Southern Lebanon and Northern Israel. Ann Geofis
XL:923–935
Ambraseys NN (2004) The 12th century seismic parox-
ysm in the Middle East: a historical perspective. Ann
Geophys 47:733–758
Ambraseys NN (2005a) Historical earthquakes in
Jerusalem—a methodological discussion. J Seismol
9:329–340. doi:10.1007/s10950-005-8183-8
Ambraseys NN (2005b) The seismic activity in Syria and
Palestine during the middle of the 8th century; an
amalgamation of historical earthquakes. J Seismol
9:115–125. doi:10.1007/s10950-005-7743-2
Ambraseys NN (2006a) Earthquakes and archaeol-
ogy. J Archaeol Sci 33:1008–1016. doi:10.1016/j.jas.
2005.11.006
Ambraseys NN (2006b) Comparison of frequency of occur-
rence of earthquakes with slip rates from long-term
seismicity data: the case of Gulf of Corinth, Sea of
Marmara and Dead Sea Zone. Geophys J Int 165:516–
526. doi:10.1111/j.1365-246X.2006.02858.x
JSeismol
Ambraseys NN, Barazangi M (1989) The 1759 earthquake
in the Bekaa Valley: implications for earthquake
hazard assessment in the Eastern Mediterranean
region. J Geophys Res 94:4007–4013. doi:10.1029/
JB094iB04p04007
Ambraseys NN, Finkel CF (1993) Material for the investi-
gation of the seismicity of the Eastern Mediterranean
region during the period 1690–1710. In: Stucci M (ed)
Materials of the CEC project ‘review of historical
seismicity in Europe’, vol 1. CNR, Milan, pp 173–
194
Ambraseys NN, Finkel C (1995) The seismicity of Turkey
and adjacent areas, a historical review, 1500–1800.
Eren Yayincilik, Istanbul
Ambraseys NN, Jackson JA (1998) Faulting associ-
ated with historical and recent earthquakes in the
Eastern Mediterranean region. Geophys J Int 133:
390–406. doi:10.1046/j.1365-246X.1998.00508.x
Ambraseys NN, Karcz I (1992) The earthquake of
1546 in the Holy Land. Terra Nova 4:253–262.
doi:10.1111/j.1365-3121.1992.tb00480.x
Ambraseys NN, Melville CP (1988) An analysis of the
Eastern Mediterranean earthquake of 20 May 1202.
In: Lee WHK, Meyers H, Shimazaki K (eds) Historical
seismograms and earthquakes of the world. Academic,
London, pp 181–200
Ambraseys NN, Melville CP (1995) Historical evidence of
faulting in Eastern Anatolia and Northern Syria. Ann
Geofis 38:337–343
Ambraseys NN, White D (1997) The seismicity of
the Eastern Mediterranean region 550–1 BC: a re-
appraisal. J Earthquake Eng 1:603–632. doi:10.1142/
S1363246997000234
Ambraseys NN, Jackson JA, Melville CP (2002) Histori-
cal seismicity and tectonics: the case of the Eastern
Mediterranean and the Middle East. In: Lee WHK,
Kanamori H, Jennings PC, Kisslinger C (eds) Interna-
tional handbook of earthquake and engineering seis-
mology, part A. Academic, Amsterdam, pp 747–763
Ambraseys NN, Melville CP, Adams RD (1994) The seis-
micity of Egypt, Arabia and the Red Sea: a historical
review. Cambridge University Press, Cambridge
Amiran DHK, Arieh E, Turcotte T (1994) Earthquakes in
Israel and adjacent areas: macroseismic observations
since 100 BCE. Isr Explor J 44:260–305
Amit R, Zilberman E, Porat N, Enzel Y (1999) Re-
lief inversion in the Avrona playa as evidence
of large-magnitude historical earthquakes, southern
Arava Valley, Dead Sea rift. Quat Res 52:76–91.
doi:10.1006/qres.1999.2050
Antonopoulos J (1980) Data from investigation on seismic
sea-waves events in the Eastern Mediterranean from
1000 to 1500 AD. Ann Geofis 33:179–198
Arvanitakis GL (1904) Essai sur le climat de Jérusalem,
Bulletin de l’Institut Égyptien, ser 4, 4 (1903) 128–189
[Essai d’une statistique des tremblements de terre en
Palestine et Syrie 178–183]
Avni R (1999) The 1927 Jericho earthquake, comprehen-
sive macroseismic analysis based on contemporary
sources. PhD thesis, Ben Gurion Univ Negev, Beer-
Sheva, Israel (in Hebrew with English abstract)
Avni R, Bowman D, Shapira A, Nur A (2002) Erroneous
interpretations of historical documents related to the
epicenter of the Jericho earthquake in the Holy Land.
J Seismol 6:469–476. doi:10.1023/A:1021191824396
Bakun WH, Wentworth CM (1997) Estimating earthquake
location and magnitude from seismic intensity data.
Bull Seismol Soc Am 87(6):1502–1521
Båth M (1965) Lateral inhomogeneities in the upper
mantle. Tectonophysics 2:483–514. doi:10.1016/0040-
1951(65)90003-X
Ben-Menahem A (1991) Four thousand years of seismicity
along the Dead Sea Rift. J Geophys Res 91:20195–
20216. doi:10.1029/91JB01936
Ben-Menahem A, Aboodi E (1971) Tectonic patterns in
the northern Red-Sea region. J Geophys Res 76:2674–
2689. doi:10.1029/JB076i011p02674
Daëron M, Elias A, Klinger Y, Tapponnier P, Jacques E,
Sursock A (2004) Sources of the AD 551, 1202
and 1759 earthquakes (Lebanon and Syria). In: Am
Geophys Union, Fall Meeting 2004, abstract #T41F-
1294
Daëron M, Klinger Y, Tapponnier P, Elias A, Jacques E,
Sursock A (2005) Sources of the large AD 1202 and
1759 Near East earthquakes. Geology 33:529–532.
doi:10.1130/G21352.1
Daëron M, Klinger Y, Tapponnier P, Elias A, Jacques E,
Sursock A (2007) 12,000-year-long record of 10
to 13 paleo-earthquakes on the Yammoûneh fault,
Levant fault system, Lebanon. Bull Seismol Soc Am
97(3):749–771. doi:10.1785/0120060106
Darawcheh R, Sbeinati MR, Margottini C, Paolini S
(2000) The 9 July 551 AD Beirut earthquake, Eastern
Mediterranean region. J Earthquake Eng 4:403–414.
doi:10.1142/S1363246900000229
Dositheos B (1715) Istoria peri ton en Ierosolymois patri-
archisanton, lib xi, cpt 7, par 3. Bucharest (in Greek)
Elias A, Tapponnier P, Singh SC, King GCP, Briais
A, Daëron M, Carton H, Sursock A, Jacques E,
Jomaa R, Klinger Y (2007) Active thrusting offshore
Mount Lebanon: source of the tsunamigenic AD 551
Beirut–Tripoli earthquake. Geology 35(8):755–758.
doi:10.1130/G23631A.1
Ellenblum R, Marco S, Agnon A, Rockwell T, Boas A
(1998) Crusader castle torn apart by earthquake at
dawn, 20 May 1202. Geology 26:303–306. doi:10.1130/
0091-7613(1998)026<0303:CCTABE>2.3.CO;2
El Mrabat T (2005) The great earthquakes in the Maghreb
region and their consequences on man and environ-
ment. Centre National pour la Recherche Scientifique
et Technique-Laboratoire de Géophysique (CNRST-
LAG), Rabat
Enzel Y, Kadan G, Eyal Y (2000) Holocene earthquakes in
the Dead Sea graben from a fan-delta sequence. Quat
Res 53:34–48. doi:10.1006/qres.1999.2096
Frankel A (1994) Implications of felt area–magnitude rela-
tions for earthquake scaling and the average frequency
of perceptible ground motion. Bull Seismol Soc Am
84(2):462–465
Gasperini P, Bernardini F, Valensise G, Boschi E (1999)
Defining seismogenic sources from historical felt re-
ports. Bull Seismol Soc Am 89:94–110
JSeismol
Geophysical Institute of Israel (2007) Search for infor-
mation on earthquakes in Israel and adjacent areas.
http://www.gii.co.il/html/seis/seis_fs.html. Accessed 7
August 2008
Gomez F, Meghraoui M, Darkal AN, Sbeinati R,
Darawcheh R, Tabet C, Khawlie M, Charabe M,
Khair K, Barazangi M (2001) Coseismic displacements
along the Serghaya fault: an active branch of the Dead
Sea fault system in Syria and Lebanon. J Geol Soc
London 158:405–408
Gomez F, Meghraoui M, Darkal A, Hijazi F, Mouty
M, Sulaiman Y, Sbeinati R, Darawcheh R, Al-
Ghazzi R, Barazangi M (2003) Holocene fault-
ing and earthquake recurrence along the Serghaya
branch of the Dead Sea fault system in Syria and
Lebanon. Geophys J Int 153:658–674. doi:10.1046/
j.1365-246X.2003.01933.x
Guidoboni E, Comastri A (2005) Catalogue of earthquakes
and tsunamis in the Mediterranean area from the 11th
to the 15th century. INGV-SGA, Bologna
Guidoboni E, Comastri A, Traina G (1994) Catalogue of
ancient earthquakes in the Mediterranean area up to
the 10th century. ING-SGA, Bologna
Guidoboni E, Bernardini F, Comastri A (2004a) The
1138–1139 and 1156–1159 destructive seismic crises
in Syria, south-eastern Turkey and northern
Lebanon. J Seismol 8:105–127. doi:10.1023/B:JOSE.
0000009502.58351.06
Guidoboni E, Bernardini F, Comastri A, Boschi E
(2004b) The large earthquake on 29 June 1170 (Syria,
Lebanon, and central southern Turkey). J Geophys
Res 109:B07304. doi:10.1029/2003JB002523
Gutenberg B, Richter CF (1954) Seismicity of the earth
and associated phenomenon, 2nd edn. Princeton
University Press, Princeton
Hofstetter A (2003) Seismic observations of the 22/11/1995
Gulf of Aqaba earthquake sequence. Tectonophysics
369:21–36. doi:10.1016/S0040-1951(03)00129-X
Hofstetter A, van Eck T, Shapira A (1996) Seismic
activity along fault branch of the Dead Sea–
Jordan Transform System: the Carmel-Tirtza fault
system. Tectonophysics 267:317–330. doi:10.1016/
S0040-1951(96)00108-4
Hough SE, Jones LM (1997) Aftershocks: are they earth-
quakes or afterthoughts? Eos Trans Am Geophys U
78:505–508. doi:10.1029/97EO00306
International Seismological Centre (2001) On-line bulletin.
Thatcham, United Kingdom. http://www.isc.ac.uk.
Accessed 7 August 2008
Karcz I (1987) Bibliographic reliability of catalogues of his-
toric earthquakes in and around Israel, II Catalogue of
Turcott and Arie (1986). Geol Surv Isr Rep GSI/10/87
Karcz I (2004) Implications of some early Jewish sources
for estimates of earthquake hazard in the Holy Land.
Ann Geophys 47:759–792
Karcz I, Lom P, (1987) Bibliographic reliability of cata-
logues of historic earthquakes in and around Israel,
I Methodology and background. Geol Surv Isr Rep
GSI/9/87
Karcz I, Kafri U, Meshel Z (1977) Archaeological
evidence for subrecent seismic activity along the
Dead Sea–Jordan Rift. Nature 269(5625):234–235.
doi:10.1038/269234a0
Kebeasy RM (1990) Seismicity. In: Said R (ed) The geol-
ogy of Egypt. Taylor & Francis, Rotterdam
Ken-Tor R, Agnon A, Enzel Y, Stein M, Marco S,
Negendank JFW (2001) High-resolution geological
record of historic earthquakes in the Dead Sea
basin. J Geophys Res 106:2221–2234. doi:10.1029/
2000JB900313
Khair K, Karakaisis GF, Papadimitriou EE (2000) Seismic
zonation of the Dead Sea Transform Fault area. Ann
Geofis 43:61–79
Kisslinger C (1996) Aftershocks and fault-zone properties.
Adv Geophys 38:1–36
Klinger Y, Avouac JP, Dorbath L, Abou Karaki N,
Tisnerat N (2000) Seismic behavior of the Dead Sea
fault along Araba valley (Jordan). Geophys J Int
142:769–782. doi:10.1046/j.1365-246x.2000.00166.x
Marco S, Stein M, Agnon A, Ron H (1996) Long-
term earthquake clustering: a 50,000-year paleoseis-
mic record in the Dead Sea Graben. J Geophys Res
101B:6179–6191. doi:10.1029/95JB01587
Marco S, Agnon A, Ellenblum R, Eidelman A, Basson U,
Boas A (1997) 817-year-old walls offset sinistrally
2.1 m by the Dead Sea Transform, Israel. J Geodyn
24:11–20. doi:10.1016/S0264-3707(96)00041-5
Marco S, Hartal M, Hazan N, Lev L, Stein M (2003) Ar-
chaeology, history, and geology of the 749 AD earth-
quake. Dead Sea Transform. Geol 31:665–668
MarcoS,RockwellTK,HeimannA,FrieslanderU,
Agnon A (2005) Late Holocene slip of the Dead
Sea Transform revealed in 3D palaeoseismic trenches
on the Jordan Gorge segment. Earth Planet Sci Lett
234:189–205. doi:10.1016/j.epsl.2005.01.017
Meghraoui M, Gomez F, Sbeinati R, Van der Woerd
J, Mouty M, Darkal AN, Radwan Y, Layyous I,
Al Najjar H, Darawcheh R, Hijazi F, Al-Ghazzi
R, Barazangi M (2003) Evidence for 830 years of
seismic quiescence from palaeoseismology, archaeo-
seismology and historical seismicity along the Dead
Sea fault in Syria. Earth Planet Sci Lett 210:35–52.
doi:10.1016/S0012-821X(03)00144-4
Migowski C, Agnon A, Bookman R, Negendank JFW,
Stein M (2004) Recurrence pattern of Holocene
earthquakes along the Dead Sea transform revealed
by varve-counting and radiocarbon dating of lacus-
trine sediments. Earth Planet Sci Lett 222:301–314.
doi:10.1016/j.epsl.2004.02.015
Neimi TM, Zhang H, Atallah M, Harrison BJ (2001)
Late Pleistocene and Holocene slip rate of the North-
ern Wadi Araba fault, Dead Sea Transform, Jordan.
J Seismol 5:449–474. doi:10.1023/A:1011487912054
Nemer T, Meghraoui M (2006) Evidence of coseismic rup-
tures along the Roum fault (Lebanon): a possible
source for the AD 1837 earthquake. J Struct Geol
28:1483–1495. doi:10.1016/j.jsg.2006.03.038
Omori F (1894) On the aftershocks of earthquakes. J Coll
Sci Imp Univ Tokyo 7:111–120
Perrey A (1850) Mémoire sur les tremblements de terre
ressentis dans la Peninsule Turco-Hellénique et en
Syrie. Mémoires de l’Académie de Belgique 23:1–73
JSeismol
Plassard J, Kogoj B (1968) Catalogue des seisms ressentis
au Liban. Ann Mém Obs, Ksara
Poirier JP, Taher MA (1980) Historical seismicity in
the near and Middle East, North Africa, and Spain
from arabic documents (VIIth–XVIIIth Century).
Bull Seismol Soc Am 70:2185–2201
Reches Z, Hoexter DF (1981) Holocene seismic and tec-
tonic activity in the Dead Sea area. Tectonophysics
80:235–254. doi:10.1016/0040-1951(81)90151-7
Salamon A (1993) Seismotectonic analysis of earthquakes
in Israel and adjacent areas. PhD thesis, Hebrew Uni-
versity, Jerusalem
Salamon A (2005) Natural seismogenic effects of the
February 11, 2004, M
L
= 5.2, Dead Sea earthquake.
Isr J Earth Sci 54:145–169. doi:10.1560/2FG7-8KDH-
UPYB-D17X
Salamon A, Rockwell T, Ward SN, Guidoboni E, Comastri
A (2007) Tsunami hazard evaluation of the Eastern
Mediterranean: historical analysis and selected mod-
eling. Bull Seismol Soc Am 97:705–724
Sbeinati MR, Darawcheh R, Mouty M (2004) The his-
torical earthquakes of Syria: an analysis of large and
moderate earthquakes from 1365 BC to 1900 AD. Ann
Geophys 47:733–758
Shalem N (1956) Seismic tidal waves (tsunamis) in the
Eastern Mediterranean. Soc Explor Eretz Isr 20:159–
170 (in Hebrew)
Shapira A, Avni R, Nur A (1993) Note: a new estimate
for the epicenter of the Jericho earthquake of 11 July
1927. Isr J Earth Sci 42:93–96
Shcherbakov R, Turcotte DL, Rundle JB (2004) Gener-
alized Omori’s law for earthquake aftershock decay.
Geophys Res Lett 31:L11613. doi:1029/2004GL19808
Shebalin NV (1973) Macroseismic data as informa-
tion on source parameters of large earthquakes.
Phys Earth Planet Inter 6:316–323. doi:10.1016/0031-
9201(72)90016-7
Sieberg A (1932) Untersuchungen über Erdbeben und
Bruchschollenbau im östlichen Mittelmeergebiet.
Denkschriften der medizinsch-naturwissenschaft-
lichen. Gesellschaft zu Jena 18:161–273
Sirovich L, Pettenati F (2001) Test of source–parameter
inversion of the intensities of a 54,000-death shock
of the seventeenth century in Sourtheast Sicily.
Bull Seismol Soc Am 91:792–811. doi:10.1785/
0120000037
Utsu T (1961) A statistical study on the occurrence of
aftershocks. Geophys Mag 30:521–605
Utsu T (2002) Statistical features of seismicity. In: Lee
WHK, Kanamori H, Jennings PC, Kisslinger C (eds)
International handbook of earthquake and engi-
neering seismology, part A. Academic, Amsterdam,
pp 719–732
Williams JB (2004) Estimation of earthquake source
parameters from soft sediment deformation layers
present in Dead Sea muds. Report presented to
Depart Civil Engineer, California State University,
Long Beach
Willis B (1928) Earthquakes in the Holy Land. Bull
Seismol Soc Am 18:73–103
Willis B (1933) Earthquakes in the Holy Land—a correc-
tion. Bull Seismol Soc Am 23:88–89
Zilberman E, Amit R, Bruner I, Nachmias Y (2004) Neo-
tectonic and paleoseismic study—Bet Shean Valley.
Isr Geol Surv Rep GSI/15/2004
Zilberman E, Amit R, Porat N, Enzel Y, Avner U
(2005) Surface ruptures induced by the devastating
1068 AD earthquake in the southern Arava val-
ley, Dead Sea Rift, Israel. Tectonophysics 408:79–99.
doi:10.1016/j.tecto.2005.05.030