faulting and seismicity: eastern mediterranean

Geophys. J. Int. (1998) 133, 390–406

Faulting associated with historical and recent earthquakes in theEastern Mediterranean region

N. N. Ambraseys1 and J. A. Jackson21Department of Civil Engineering, Imperial College of Science, T echnology and Medicine, L ondon, SW7 2BU, UK. E-mail: [email protected]

2Bullard L aboratories, University of Cambridge, Cambridge CB3 0EZ, UK

Accepted 1997 November 17. Received in original form 1997 June 4

SUMMARYThis paper summarizes evidence for surface faulting in historical and recent earthquakesin the Eastern Mediterranean region and in the Middle East. Such information isparticularly important for studies of active tectonics and for palaeoseismology. Wehave found 78 cases of faulting pre-1900 and 72 post-1900, some of which show thatfaults that have apparently been inactive this century had already ruptured before1900. For some cases faulting could not have been predicted from 20th century activity,and in others it could have been expected, but has not been observed during theinstrumental period. The data are sufficient to allow the derivation of relationshipsbetween magnitude and rupture length.

The purpose of this paper is to present the cases of coseismic1 INTRODUCTION

surface faulting known to us at present, both historical andmodern, to show that faults in the region which appear to beEvidence for surface faulting in historical earthquakes in thequiescent today have been active in historical times, sometimesEastern Mediterranean and the Middle East is of importancemore than once, and to identify hitherto unknown active faults.to all modern studies of tectonics and seismicity. Such evidenceThis compilation thus updates the last attempt to documentnot only confirms that known tectonic structures are active,coseismic surface ruptures in the region by Ambraseys (1975),but can also identify new ones. Despite shortcomings in thewith almost double the number of cases in this new study.documentary evidence, information about surface faulting can

be found in contemporary accounts and this provides a

valuable reference point in the palaeoseismological record of

faults. Such knowledge is particularly important when, for DATAexample, the activity of a fault is to be researched by trenching

The data used have been culled from a variety of publishedmethods, as it allows the completeness of the palaeoseismologi-

and unpublished sources and field investigations, in a numbercal investigation to be assessed.

of cases carried out by the first author. Because of spaceObviously, the most interesting cases are those which have

limitation for events before this century, only a few referenceshappened where their occurrence could not be predicted from

are given, and these are chiefly collections of literary sources.20th century seismicity alone or, alternatively, where surface

For the later period we have selected references which coverfaulting could be expected from the 20th century seismicity

both field data and seismological or engineering studies. It isbut until now is not known to have happened. Since surface

somewhat embarrassing but also unavoidable that one-quarterfaulting is associated with large earthquakes, evidence of

of the works quoted are by the first author, which stems fromfaulting can also be used to assess their size, even when

necessity rather than from other motives.historical macroseismic sources do not provide enough direct

evidence for magnitude estimates.

The area of our investigations, shown in Fig. 1(a), is withinPre-instrumental periodlatitude 25° and 45° north and longitude 18° and 70° east. It

comprises the Balkans, Turkey, the Caucasus and the Middle Historical sources record large surface fault ruptures, smallEast up to west Pakistan, a region of active tectonics and with ruptures not being spectacular enough to attract attention.a history which is amply, but not uniformly, documented Descriptions from which one can deduce faulting are relativelythroughout the period of our interest of the past two millennia. few and hard to verify, particularly when the sources areFig. 1(b) shows the distribution of medium and large earth- secondary and the recorded ground deformation is not wellquakes during this century, and Fig. 1(c) is a location map described. It follows, therefore, that for the early historical

period the information presented here is incomplete, but it isshowing some of the major fault zones referred to in the text.

390 © 1998 RAS

by guest on June 29, 2014http://gji.oxfordjournals.org/

Dow

nloaded from

http://gji.oxfordjournals.org/

Faulting in the Eastern Mediterranean 391

information about faulting is implicit as, for instance, in the

case of the #280 BC earthquake in Iran. Some examples ofthe descriptions found for this category are given in AppendixA. Other accounts of faulting are more explicit but quite a few

are only very brief, and yield no further reliable informationby being read into.

(B) Cases for which surface faulting is not supported by

clear evidence but can be inferred from the association of anarrow and long epicentral region of a large-magnitude earth-quake aligning with, or close to, a known fault. Occasionally

the length of a break can be reckoned from the length of thelong axis of the epicentral region which contains an assumedrupture. Clearly this would not tell us exactly how far the fault

rupture extended, as it may have continued for a greaterdistance into sparsely populated areas, which we are unlikelyto find reported in historical sources, but it will tell us that

the shock was probably associated with a surface rupture thatcan be investigated today in the field. In these cases historicalinformation will not reveal the exact location and rupture

length, but it can help to define the time and the segment ofthe zone that was probably ruptured.

(C) Faulting assumed because of the large size (Ms≥7.0) of

the associated earthquake and its proximity to a known activefault zone. This category is more tenuous than category B, but

it was included to guide further studies. There are many eventsof Ms≥7.0 that might have been associated with faulting, suchas those in and around the Marmara Sea area, in Eastern

Anatolia and Iran, but these are omitted as their epicentralarea is ill defined.

Of these three categories (A) and (a) involve some ruptures

which may not previously have been associated with knownactive or Quaternary faults. Categories (B) and (C) merelydate probable breaks of segments of known faults, and helpassign size to these events. All these cases indicate recent fault

Figure 1. (a) Area of our investigations, showing earthquakes of mb>4activity because the proximity of these earthquakes to known

during the period 1964–1990. (b) Distribution of significant shallowfaults was part of the evidence assigning them to these

earthquakes during this century: open circles, Ms between 6 and 6.9;categories.solid circles, Ms≥7.0. The largest symbols are Ms≥8. (c) Location

Figs 2, 4 and 5 show the distribution of the epicentres inmap of the main fault area referred to in the text: NAF NorthTable 1 for the whole period of observation, before 1894 andAnatolian Fault, EAF East Anatolian Fault, DSF Dead Sea Fault, CFafter 1893.Chaman Fault.

Instrumental periodput on record so that others can improve upon it by refiningit and adding new case histories. During the instrumental period information about both the

faulting and the seismological parameters of the associatedOne of the problems in these early and later descriptions of

surface faulting is that one cannot always be certain whether earthquake improves: there are more detailed field observa-tions and better instrumental data allowing the uniformground deformation associated with an earthquake was of

tectonic origin or due to landslides, liquefaction or slumping re-assessment of instrumental Ms magnitudes.

However, during the first half of this century this improve-of the ground. In some cases ground deformation genuinely oftectonic origin can be identified from descriptions of ground ment was very slow and surface faulting continued to be

imperfectly reported. For example, the fault ruptures associatedruptures which extended continuously or discontinuouslyalong considerable distances, but relative displacements are with the Locris earthquakes of 1894 in Greece were not

properly mapped and their tectonic origin was not generallyseldom given for vertical, and never for horizontal, slip. The

information which is usually available for this period may accepted by geologists, who until relatively recently consideredthis feature to have been a superficial effect of sliding. Also, oftherefore be classified into three broad categories according to

the following criteria. the 360 km long fault break associated with the 1939 Erzincan

earthquake in Turkey, only its western half was visited after(A) or (a) Strong evidence for surface faulting explicitly (A)or implicitly (a) described in the sources. The length of the the event, only part of the break was sketched rather than

mapped on a one-to-one-million scale, and measurements ofrupture is rarely given, and only in few cases can it be reckoned

from the distances between the localities which it traversed. fault displacement were made at a single location. The sameapplies to other major surface fault ruptures during that periodTo avoid any misinterpretation of the source material we have

indicated in Tables 1 and 2 by small (a) cases for which in Anatolia, Iran, and Greece.

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from


392 N. N. Ambraseys and J. A. Jackson

Table 1. List of earthquakes associated with surface fault break.

Date Epicentre Ms Az Mec L H V Q Location RefN E deg. km cm cm

1 −464 – – 37.0–22.4 m 340 N 20 – 350 Cf Sparta GR2 −426 – – 38.9–22.7 m – – – – – AAm Maliac G. GR3 −280 – – 35.6–51.4 L – – – – – aA Sh. Rey IR4 17 – – 38.5–27.8 L 270 N – – – APf Gediz R. TR5 32 – – 40.5–31.5 L 080 R – – – Cf Gerede TR6 37 – – 36.0–36.0 m 030 L – – – Cf Antioch TR7 – – 37.3–36.5 m – – – – – C E.Anatol. TR8 110 – – 39.5–33.5 m – – – – – C Galatia TR9 115 Dec 13 35.8–36.3 L 010 L – – – aAf Oront. R. SY

10 155 – – 40.1–27.5 m 100 R – – – aAf Manyas TR11 181 May 3 40.5–31.0 L – – – – – Bf Mudurnu TR12 236 – – 40.9–36.0 m 110 R – – – aPf Amasya TR13 368 Oct 11 40.5–29.5 L – – – – – Cf Iznik TR14 460 Apr 7 40.3–27.8 m 060 R – – – aAf Manyas TR15 499 Sep – 40.5–37.0 m 110 R – – – APf Niksar TR16 518 – – 42.0–21.0 m – – 43 – – AP Macedonia MC17 551 – – 38.5–22.7 m 290 N – – – APfm Chaeron GR18 554 Aug 15 40.8–29.5 L – – – – – Bf Izmit TR19 601 Apr – 37.0–36.5 - – – – – – aA E.Anatol. TR20 750 – – 37.0–38.0 - – – – – – aA Mesopot. SY21 856 Dec 22 36.0–54.3 L 250 T – – – aAf Qumis IR22 926 Aug – 38.5–27.5 m 270 N – – – aAf Manisa TR23 967 Sep – 40.8–32.0 L 080 R – – – Bf Gerede TR24 995 – – 38.7–40.0 L 060 L – – – Cf Palu TR25 1033 Dec 5 32.5–35.5 L 000 L – – – Cmf Jordan IS26 1035 May – 40.8–33.0 m 070 R – – – aAf Cerkes TR27 1045 Apr 5 40.0–38.0 L 120 R – – – aAmf Erzinc. TR28 1050 Aug 5 41.0–33.5 L 080 L – – – aAf Cankiri TR29 1068 Mar 18 28.5–36.7 L – – – – – aA Hejaz SA30 1114 Nov 29 37.5–37.5 V 040 L – – – af Maras TR31 1157 Aug 12 35.0–36.5 V 000 L – – – af Hama SY32 1170 Jun 29 35.5–36.5 L 000 L – – – Cf Afamiya SY33 1202 May 20 33.7–35.9 L 020 L – – – Bfm Bekaa LE34 1254 Oct 11 40.0–39.0 L 110 R 150 – – APf Susehri TR35 1296 Jul 17 39.2–27.4 m 050 N – – – aAf Soma TR36 1336 Oct 21 34.7–59.7 7.6* 155 – 100 – – BPf Kwaf IR37 1408 Dec 29 36.0–36.4 m 010 L 20 – – APf Orontes SY38 1419 Mar 15 40.5–30.5 L – – – – – Cf Mudurnu? TR39 1493 Jan 10 33.0–59.8 7.0* 120 T 30 – – APf Birjand IR40 1505 Jul 6 34.8–69.1 7.4* 010 – 56 – 300 AP Kabul AF41 1544 Jan 22 38.0–37.0 m 090 L – – – BAf Elbistan TR42 1595 Sep 22 38.5–27.9 m 270 N – – – aP Ahmetli TR43 1646 Apr 7 38.3–43.7 L 070 – – – – aP Van TR44 1651 Jun 7 37.8–29.3 m 120 N – – – Bf Honaz TR45 1653 Feb 22 37.9–28.5 7.1* 090 N 70 – 300 APd Menderes TR46 1661 Mar 15 42.2–24.0 L – – – – – aA Maritza BU47 1666 Sep 23 36.7–43.5 L – – – – – C N.Mosul IQ48 1668 Aug 17 40.5–36.0 7.9* 090 – 400 – – APf Amasya TR49 1721 Apr 26 37.9–46.7 7.7* 125 – 50+ – – AP Tabriz IR50 1740 Oct 5 38.7–22.4 6.6* – – 20 – – aP Lamia GR51 1752 Jul 29 41.3–26.5 L – – – – – aA Evros TR52 1759 Nov 25 33.7–35.9 7.4* 020 L 100 – – Bekaa LE53 1780 Jan 8 38.2–46.0 7.7* 120 RN 60+ – 600 AP Tabriz IR54 1784 Jul 18 39.5–40.2 7.6* 110 R 150 – – Bfm Elmali TR55 1789 May 28 38.8–39.5 L – – – – – BP Elazig TR56 1796 Apr 26 35.5–36.0 6.6* – – – – – aA Latakia SY57 1822 Aug 13 36.7–36.5 7.5* 020 L 200 – – APd Antakya TR58a 1825 – – 36.1–52.6 6.7* – – – – – aPd Harhaz IR58b 1829 May 5 41.2–25.1 7.2* – – 50 – – aPk Xanthi GR59 1837 Jan 1 33.2–35.5 7.4* 000 – 80 – – BPkm Bshara LE60 1838 – – 29.6–59.9 7.0* 170 – 70 – – APf Nasratab IR61 1840 Jul 2 39.5–43.8 7.3* 140 R 80 – – BPf Kazlgol TR62 1855 Feb 28 40.0–28.5 7.4* 270 – 70 – – APm Ulubat TR63 1855 Apr 11 40.3–29.1 6.6* – – – – – aA Gemlik TR64 1861 Dec 26 38.2–22.2 6.6* 280 N 13 – 220 APk Vostiza GR65 1864 Dec 7 33.2–45.9 6.4* – – 2+ – 50 aPk Zorbatia IQ

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



Table 1. (Continued.)


66 1866 May 12 39.2–41.0 7.2* 230 L 45 – – APm Gonek TR67 1870 Aug 1 38.5–22.6 6.7* 010 N 6+ – 200 aPm Fokis GR68 1872 Apr 3 36.4–36.4 7.2* 030 – 20 – – APk Amik Gol TR69 1874 May 3 38.5–39.5 7.1* 250 L 45 – 200 AM Golcuk 1 TR70 1875 Mar 27 38.5–39.5 6.7* 250 – 20 – 200 aP Golcuk 2 TR71 1875 May 3 38.3–29.9 6.5* 040 N 10 – 110 aP Civril TR72 1880 Jul 29 38.6–27.2 6.5* 120 N 10 – 40 aPk Emiralan TR73 1887 Sep 30 38.7–29.8 6.3* 290 N 10 – 50 aPd Banaz TR74 1889 Jan 17 37.7–30.5 - – – – – – aPk Isparta TR75 1892 Dec 19 30.9–66.5 6.9* 020 L 30 80 30 AMi Chaman PK76 1893 Mar 2 38.0–38.3 7.1* 270 L – – – Bf Malatya TR77 1894 Apr 27 38.6–23.2 6.9* 300 N 40 – 100 AMdm Martin #GR78 1899 Sep 20 37.9–28.8 6.9 090 N 40 – 100 AMd Mender #TR79 1904 Apr 4 41.8–23.1 7.2 230 N 25 – 200 aMim Struma #BU80 1905 Jun 1 42.0–19.5 6.3 040 N 10 – 100 aPk Scutari #AL81 1905 Dec 4 38.1–38.6 6.8 240 L – – – aPkf Malat. TR82 1909 Jan 23 33.4–49.3 7.4 315 R 45 – 250 AMd Silakhor #IR83 1909 Oct 20 28.9–68.3 7.1 130 L* 50 – – B Baluch #PK84 1909 Feb 9 40.2–37.8 6.4 280 R* 15 – – APdm Ender. #TR*85 1912 Aug 9 40.7–27.2 7.4 065 NR 50 – 300 AMd Marmara #TR86 1914 Oct 3 37.6–30.1 7.0 230 NR 23 – 150 aPk Burdur #TR87 1916 Jan 24 40.8–37.5 7.2 110 RL – – – Cf Samsun TR88 1917 Jul 15 33.5–45.8 5.6 140 T* 2 – – APk Tursaq IQ89 1928 Apr 14 42.1–25.2 6.8 290 N 64 – 50 AMU Plovdiv #BU90 1928 Apr 18 42.2–24.9 7.0 300 N 50 – 350 AM Plovdiv *BU91 1929 May 1 37.7–57.8 7.3 330 T 70 – 210 AG Kop. Dagh #TU92 1930 May 6 38.2–44.6 7.2 305 RN 30 400 500 AMU Salmas #IR93 1932 Sep 26 40.5–23.9 6.9 090 N 15 25 180 AMdi Ieriss #GR94 1933 Nov 28 32.0–55.9 6.2 140 T 5 – 50 AMk Buhabad #IR95 1935 May 30 29.8–66.8 7.6 015 T – – – Bdk Quetta PK96 1938 Apr 19 39.5–34.0 6.8 120 R 14 100 60 AMd Kirsehir *TR97 1939 Dec 26 39.7–39.7 7.8 110 R 340 650 250 AM Erzincan *TR98 1941 Feb 16 33.4–58.9 6.1 005 RT 12 – 50 AGd Muham/ad #IR99 1942 Dec 20 40.7–36.5 7.1 300 R 47 180 AMk Erba-Niks *TR

100 1943 Nov 26 41.0–35.5 7.4 275 R 270 200 100 AMd Ladik *TR101 1944 Feb 1 40.9–32.6 7.3 255 R 160 370 100 AM Ger-Bolu *TR102 1944 Jun 25 39.0–29.4 6.0 140 NR 18 – 30 APdU Saphane #TR103 1946 May 31 39.3–41.2 5.7 300 R 10 30 30 APkd Ustukr. *TR104 1946 Jul 27 35.6–45.8 5.5 145 T* 2 – – aPk Penjwin IQ105 1947 Sep 23 33.7–58.7 6.8 180 RT 20 100 80 AG Dustab. #IR106 1948 Oct 5 37.9–58.5 7.2 260 T – – – Bdk Ashkhab. TU107 1949 Aug 17 39.4–40.8 6.9 100 R 38 150 30 AMd Elmalid. *TR108 1951 Aug 13 40.7–33.3 6.9 260 R 32 60 30 APd Kursunlu #TR109 1953 Feb 12 35.4–54.9 6.5 070 T 8 – 140 APdi Turud #IR110 1953 Mar 18 39.9–27.4 7.3 240 R 58 430 50 AMd Gonen *TR111 1954 Apr 30 39.2–22.2 6.7 300 N 30 20 90 AMd Sofades #GR112 1957 Mar 8 39.3–22.7 6.6 100 NL 1 20 20 APi Velestin GR113 1957 May 26 40.6–31.0 7.0 260 R 40 160 45 AP Abant *TR114 1958 Aug 16 34.3–48.2 6.6 300 T 28 – 50 AMdk Firuz. *IR115 1962 Sep 1 35.7–49.8 7.2 105 L 85 60 80 AGd B. Zahra *IR116 1964 Oct 6 40.0–28.0 6.8 100 NR 40 – 10 AMkU Manyas #TR117 1966 Aug 19 39.2–41.4 6.8 120 RN 34 30 25 AGdU Varto *TR118 1966 Aug 20 39.3–41.2 6.2 110 RN 7 5 20 AMdm Varto TR119 1966 Sep 1 37.4–22.1 5.6 155 N* 2 – 5 aMk Megalop. GR120 1966 Oct 29 38.8–21.1 5.8 150 N 4 – 40 AMd Acarnan. #GR121 1967 Jul 22 40.7–30.7 7.1 280 R 80 190 130 AGd Mudurnu *TR122 1967 Jul 26 39.5–40.3 6.0 120 R 4 20 10 APk Tunceli #TR123 1967 Jul 30 40.7–30.4 5.5 300 R 3 20 40 AGd Mudurnu TR124 1967 Nov 30 41.4–20.4 6.6 030 NL 10 – 50 AMd Debar *AL125 1968 Feb 19 39.5–24.9 7.3 040 RN 3 – 50 AMi Ag. Efstr *GR126 1968 Aug 31 34.0–58.9 7.4 275 L 80 450 250 AG D. Bayaz *IR127 1969 Mar 28 38.3–28.5 6.5 290 NL 35 20 80 AMd Alasehir *TR128 1970 Mar 28 39.1–29.4 7.1 310 NL 45 30 230 AGUm Gediz *TR129 1971 May 12 37.6–30.1 6.2 230 N 4 – 30 AMim Burdur #TR130 1971 May 22 39.0–40.7 6.8 050 L 38 60 10 AGd Bingol *TR131 1975 Sep 6 38.5–40.7 6.6 270 T 28 – 60 AGd Lice *TR

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



Table 1. (Continued.)


132 1975 Oct 3 30.3–66.1 6.5 025 L 5 4 – AMi Baluch PK133 1976 Nov 24 39.1–43.9 7.3 110 R 48 350 50 AG Chaldiran *TR134 1977 Dec 19 30.9–56.6 5.8 320 RT 7 15 7 AGd Gisk *IR135 1978 Jun 20 40.6–23.2 6.4 300 N 32 2 20 AGdU Volvi *GR136 1978 Sep 16 33.4–57.1 7.4 330 T 80 260 180 AMdU Tabas IR137 1979 Nov 14 33.9–59.8 6.6 345 RT 18 90 60 AMdU Karizan *IR138 1979 Nov 27 34.0–59.6 7.1 080 LT 68 260 250 AMd Khuli *IR139 1980 Jul 9 39.3–22.8 6.4 090 N 8 5 20 AGd Almyros *GR140 1981 Feb 25 38.1–23.1 6.4 250 N 15 80 AGdm Alkyon *GR141 +1981 Mar 4 38.2–23.2 6.3 070 N 12 50 AGd Alkyon *GR142 1981 Jun 11 29.8–57.8 6.7 340 RT 15 3 5 AGU Golbaf *IR143 1981 Jul 28 30.2–57.6 7.1 340 RT 65 43 40 AGdU Sirch *IR144 1983 Oct 30 40.4–42.2 6.7 050 L 12 100 60 AGdU Panisler *TR145 1986 Sep 13 37.0–22.0 5.7 190 N 6 15 AG Kalamata *GR146 1988 Dec 7 40.8–44.2 6.7 290 RT 33 150 50 AGdm Spitak *AR147 1990 Jun 20 36.8–49.4 7.3 110 LT 80 60 90 AMdk Manjil *IR148 1994 Feb 23 30.9–60.6 6.0 320 TL 4 30 AMm Lut IR149 1995 May 13 40.0–21.7 6.5 240 N 15 0 5 AMdk Kozani GR150 1995 Oct 1 38.2–30.3 6.2 330 NR 10 10 30 AMk Dinar TR

References (see Appendix B): [1] 39, 71; [2] 31, 71; [3] 21, 31; [4] 4, 32, 71; [5] 4, 32, 71; [6] 4, 32, 71; [7–9] 4, 32, 71; [10] 4, 32; [11–13] 4, 32,

71; [14–15] 4, 32; [16–17] 4, 5, 32, 71; [18] 4, 32; [19–24] 4, 32, 71; [25–28] 4; [29] 4, 23; [30–32] 4; [33] 4, 22; [34] 4,24; [35] 4; [36] 21; [37]

24; [38] 4; [39] 21; [40] 20, 37; [41] 10; [42] 17; [43] 21; [44] 4; [45] 8, 17; [46–47] 4; [48] 16; [49] 21; [50] 10; [51] 17; [52] 13; [53] 21;

[54–55] 17; [56–57] 9; [58] 21; [59] 12; [60–61] 21; [62–63] 4; [64] 19b, 123; [65–66] 4; [67] 28; [68–70] 9; [71–73] 4; [74] 9; [75] 70; [76]

9; [77] 19; [78] 8, 14; [79] 147; [80] 89; [81] 14; [82] 21; [83] 76; [84] 8; [85] 15; [86] 8; [87] 4; [88] 21; [89] 88, 97; [90] 6, 88, 97, 119, 150;

[91] 21; [92] 136; [93] 114; [94] 21; [95] 146; [96] 8, 62, 113, 116, 122; [97] 8, 46, 48, 62, 80; [98] 21; [99] 8, 48, 59, 60, 62, 84, 108; [100] 8, 48,

60, 62, 80, 84, 90; [101] 8, 62, 84, 132; [102] 8; [103] 8, 62, 133; [104–105] 21; [106] 121; [107] 8, 48, 80; [108] 8, 118; [109] 27; [110] 48, 62, 65,

80, 85; [111] 109; [112] 19; [113] 8, 34, 48, 65, 106; [114] 26; [115] 1, 21, 98, 117; [116] 64, 83; [117] 8, 33, 48, 91, 104, 144; [118] 33; [119–120]

2; [121] 34, 48, 65, 80, 104; [122–123] 4; [124] 4, 45, 104, 130; [125] 114, 134; [126] 22, 29, 74, 77, 94, 100, 101, 135, 137; [127] 8, 30, 42, 66, 77;

[128] 8, 30, 66, 77, 131, [129] 8; [130] 8, 40, 43, 82, 124; [131] 8, 41, 79, 138; [132] 68; [133] 8, 44, 48, 72, 138, 139; [134] 22, 38, 55; [135] 92, 95,

96, 110, 126, 127; [136] 50, 51, 54, 102; [137] 22, 73, 102; [138] 73, 102; [139] 19, 112; [140] 58, 78, 86, 87, 129, [141] 58, 78, 86, 87, 129; [142] 56,

141; [143] 56, 105; [144] 8, 48; [145] 93, 128; [146] 61, 81; [147] 57, 99, 103, 148; [148] 149; [149] 75, 115; [150] 63, 67.

Notes

All events are assumed to have focal depths in the crust.

Magnitude: magnitudes for the instrumental period are recalculated Ms values derived from the Prague formula. For early events of the pre-

instrumental period, magnitudes (starred) have been derived from macroseismic information calibrated against instrumental Ms values. The size of

historical events under investigation has been classified under three broad categories: V, very large event M≥7.8; L, large event 7.0≤Ms<7.8; M,

medium event 6.0≤Ms<7.0.

Fault attitude and mechanism: T=thrust; L=left-lateral strike-slip; R=right-lateral strike-slip; N=normal, with a combination of these symbols

for oblique motion. *=Assumed from regional fault pattern.

Length of faulting: L=total length of surface rupture, including intermediate unfractured segments in km.

Relative displacements: H=maximum observed lateral offset in cm; V=maximum observed vertical offset in cm; s=small displacements of

imperceptible sense of motion; –=no data.

Quality of evidence of faulting, Q (first column of Q): (A) surface faulting explicitly or (a) implicitly, deduced from the sources or field investigations;

(B) no evidence for faulting in the sources; surface faulting inferred from the elongated shape of the epicentral region; and (C) faulting assumed

because of the large size of the earthquake and its proximity to a known active fault zone.

Location evidence (second column of Q) for quality categories A and a is subdivided into: G=good, derived from detailed field studies; M=moderate, based on cursory field survey of the fracture zone; P=poor, deduced from historical data or, for more recent events, from field evidence

in need of authentication; A=very poor, exact location of fault break unknown.

Nature of fault zone (third column of Q): d=Trace discontinuous or eroded; total length of rupture deduced from few and widely spaced reported

observations; U=arcuate trace, graben, or complex fault zone; k=some of the observed or reported ground deformations probably not of tectonic

origin; i=only part of the break was accessible or mapped; actual rupture length is probably longer than reported; n=reported ground effects, to

the best of our judgement, not of tectonic origin or associated with a known earthquake; m=multiple shock; observed deformations and rupture

length probably enhanced by more than one earthquake.

For quality A, B and C (in any column of Q): f=assumed association of historical event with known Quaternary or recent fault-break.

The name of the location where the event took place is given in the penultimate column, and the last column gives the country. AF: Afghanistan;

AL: Albania; BU: Bulgaria; GR: Greece; IQ: Iraq; IR: Iran; IS: Israel; LE: Lebanon; MC: Macedonia; PK: Pakistan; SA: Saudi Arabia; SY: Syria;

TR: Turkey; TU: Turkmenistan.

* or #before the country designation indicates that the event was used/not used by Wells & Coppersmith (1994) in the derivation of their

calibration formulae.

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



Table 2. Uncertain and spurious cases of surface faulting.

FaultLoc Ms Az Mec L H V Q Location

1 1862 Nov 3 38.5–30.3 6.5* – – 3 – 50 aAn Suhut TR

2 1870 Feb 22 36.6–29.0 – – – 2 – 30 aPn Fethiye TR

3 1879 Mar 22 37.8–47.9 6.6* 170 T 2+ – – Bn Buzqush IR

4 1890 Jul 11 36.5–54.6 7.2* 060 – 10 – – aPn Tash IR

5 1911 Apr 18 31.2–57.0 6.2 155 – 15 – 50 aPn Ravar #IR

6 1927 Jul 11 32.0–35.5 6.0 – – – – 200 an Jordan IS

7 1929 Jul 15 32.1–49.5 6.0 150 – 1 – 100 Bkn Londeh IR

8 1943 Jun 20 40.7–30.5 6.4 Hendek TR

9 1957 Jul 2 36.1–52.4 6.8 120 – 3 – 10 akn Elburz IR

10 1957 Dec 13 34.6–47.8 6.7 315 T 10 – 100 akn Farsinaj #IR

11 1963 Jul 26 42.1–21.4 6.1 115 – 6 – 10 akn Skopje *MC

12 1966 Feb 5 39.1–21.6 6.2 230 N 2 – 30 kn Kremas GR

13 1968 Sep 3 41.8–32.3 6.5 160 – 2 20 30 akn Bartim #TR

14 1968 Sep 24 39.2–40.2 5.1 150 – 6 – 25 akn Kigi #TR

15 1969 Mar 25 39.2–28.5 6.1 100 N 5 10 10 akn Demirci TR

16 1972 Apr 10 28.4–53.0 6.9 120 – 20 5 25 akn Qir *IR

17 1972 Jul 2 30.0–50.9 5.3 290 N 10 – 400 AGn Mishan #IR

18 1975 Mar 7 27.5–56.4 6.1 Sarkhun IR

19 1976 Nov 7 33.8–59.2 6.4 140 – 9 s s akn Vandik IR

20 1977 Apr 1 27.6–56.3 6.3 n Khurgu #IR

21 1977 Apr 6 31.9–50.8 6.1 n Naghan #IR

22 1977 Jun 5 32.6–48.1 5.7 n Dizful #IR

23 1983 Aug 6 40.1–24.7 6.8 R n N. Aegean #GR

24 1992 Mar 13 39.6–39.5 6.8 – – 30 20 – akn Erzincan *TR

25 1995 Jun 15 38.4–22.3 6.5 280 N 7 – 3 adn Egio GR

References (see Appendix B): [1–2] 6; [3–5] 21; [4–5] 21; [6] 4; [7] 21; [8] 140; [9] 21; [10] 36; [11] 3, 6; [12] 19; [13–14] 6; [15] 42; [16]

6, 21, 35; [17] 52; [18] 145; [19] 69; [20] 53; [21] 7; [22] 145; [23] 145; [24] 47, 49, 142; [25] 120.

ASSESSMENT OF MAGNITUDES

It is important to know the magnitude of the causativeearthquake, not only for the development of predictivemoment–magnitude relations as a function of the length, slip

and attitude of a surface break, but also for hazard analysis.For the pre-instrumental period, surface-wave magnitudes, Ms,can be assessed using a calibration formula which can be

derived from regional, shallow, 20th century earthquakes interms of their radii of isoseismals, r, and corresponding intensit-ies, I, in the MSK (Medvedev–Sponheuer–Karnik) scale. InFigure 2. Locations of earthquakes associated with surface faultingthe present case the calibration formula we used was derivedfor the whole period of observation.from intensity data and isoseismals culled from a variety ofpublished sources, including Shebalin et al. (1974), Papazachos

et al. (1982) and Ambraseys & Jackson (1990), variables whichOccasionally, surface fault ruptures were wrongly attributedwere correlated with uniformly recalculated Ms (Ambraseys &to landslides and slumping of the ground, and pre-existingFree 1997). From 488 isoseismals coming from about 9000Quaternary fault scarps were often associated with recentintensity points which were associated with 123 shallowearthquakes. An example is a 10 km long Quaternary normal(h<26 km) earthquakes of the period 1905–1990 and fromfault showing a throw of 4 m, which was attributed to thetheir corresponding Ms values, which have been recalculatedearthquake of 1972 July 2 (Ms=5.3) in southwestern Iranin this study, the predictive relationship is(Berberian & Tchalenko 1976). A site visit in 1976 confirmed

that this scarp, averaging about 2 m, was clearly an old feature,Ms=−1.54+0.65 (Ii)+0.0029 (Ri)+2.14 log(Ri)+0.32p ,

certainly pre-dating the 1972 earthquake and controlling the(1)

course of the seasonal streams and various old tracks acrossit which were not dislocated by the 1972 earthquake. Old local where Ri=(ri2+9.72)0.5 and r, in kilometres, is the mean

isoseismal radius of intensity I, and p is zero for mean valuesfarmers remembered the scarp from their early days, and 1955aerial photos show it clearly. and one for 84 percentile values (Ambraseys 1992).

With few exceptions, macroseismic data for the historicalThere is no doubt that in the last two decades the situation

has improved: sites of historical faulting have been revisited, period are scanty and the magnitudes that can be calculatedfrom eq. (1) are rather uncertain. In such cases we grouptrenched and mapped, and faulting due to recent earthquakes

properly recorded. earthquakes into three broad categories: V, very large events

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



with Ms values probably exceeding 8.0; L, large shocks of variety of data sets, have been derived for different parts ofmagnitude between 7.0 and 8.0; and M, medium events with the world, and reviews of these relationships are availableMs ranging between 6.0 and 7.0. (Wells & Coppersmith 1994).

For the late pre-instrumental period, starting with the 18th For 62 of the 150 earthquakes in Table 1, we have bothcentury, macroseismic data improve in quality and quantity well-determined surface-wave magnitudes (Ms) from instrumen-and this allows the use of eq. (1) for the assessment of tal data, and reasonably reliable rupture lengths from fieldmagnitudes. observations. These events are all in the instrumental period,

with 55 per cent of the data coming from strike-slip, 30 percent from normal and 15 per cent from thrust faults, excludingRESULTScases of quality (B) and others for which the rupture length is

Table 1 summarizes all the events, 150 in all, that we know,imperfectly known [marked (i) in column (Q)]. A straightfor-

or suspect, to have been associated with coseismic surfaceward orthogonal regression between Ms and log (L) gives

faulting, and Fig. 2 shows their location. Table 2 lists anotherMs=5.13+1.14 log(L ) , (2)25 cases of faulting which we believe to be uncertain or

spurious. The values of the various parameters listed for eachwith L in kilometres, with a standard deviation of 0.15 in Ms.event have been culled from a variety of sources, and theAlternatively, regressions of Ms on log (L ) and of log (L ) on

sources of information given for each entry have been chosenMs give

chiefly because they give up-to-date cross-references for theevent. The values for the various parameters in these Tables Ms=5.27+1.04 log(L ) (3)supercede and correct previous estimates made by the authors

andand by other writers.Each entry in these tables gives the date and time of the log(L )=−4.09+0.82Ms , (4)

event in the New Style (Gregorian calendar), the geographicalrespectively, with almost the same standard deviation of 0.22coordinates of the location of the middle point of the rupture,in Ms for both cases, while a non-linear fit results inand the size of the associated event in terms of its surface-

wave magnitude Ms. For the instrumental period, Ms values Ms=5.06+1.42 log(L ) −0.14[log(L )]2 , (5)have been recalculated uniformly using surface-wave ampli-

with a slightly larger standard deviation. Fig. 3 shows eqs (2),tudes and periods and the original Prague formula, which does(3) and (4) together with the data points.not restrict the period to the specific range 18–22 s, and allows

For 58 of the 62 earthquakes used to derive eqs (2) to (5)the use of data in the range 3–25 s (Vanek et al. 1962;we also have horizontal (H ) and vertical (V ) maximum surfaceAmbraseys & Free 1997).displacements, but the fit improves little, given byNext, the azimuth of the strike of the break (Az), measured

from north to east, is given, when known, from field obser-Ms=5.11+0.86 log(L )+0.21 log(R) , (6)

vations, or marked by (f ) in the quality column Q, if its valuehas been assumed from regional tectonics. Slip type is desig- with a standard deviation of 0.20 in Ms, in which R is thenated by (S) for strike-slip, (N) for normal, (T) for thrust and resultant displacement from H and V in centimetres.by a combination of these notations for oblique slip. The In terms of resultant displacement R alone, Ms may beobserved length of surface rupture (L), in kilometres, is given approximated byas deduced from the sources or as obtained from field studies.

Ms=5.21+0.78 log(R) , (7)A plus sign indicates that the actual length was probablygreater than shown. The horizontal relative displacement (H ), with a rather large standard deviation of 0.36 in Ms.in centimetres, is the maximum value observed on the fault We find that the resultant displacement R is about 5.0break or across principal displacement zones. The vertical (±4.0)×10−5L , regardless of mechanism, and a number closerelative displacement (V ), in centimetres, represents the maxi- to compilations by Scholz 1982) and Scholz et al. 1986).mum throw across principal displacement zones, excluding However, the size of the sample is insufficient and the scattermeasurements affected by ground deformations, which are too large to allow a better estimate of eqs (2)–(5) and R as aprobably superficial, due to slumping or liquefaction. A factor function of mechanism.(Q) adds more coded information regarding the nature of the The predictive relationship between magnitude and faultfault and quality of measurements (see note at the end of

length for the instrumental period, eq. (3), is almost identicalTable 1). The location of the earthquake is given by the

to that derived by Wells & Coppersmith (1994) from a globalmodern name of the area affected. The last column identifies

data set, their Fig. 8, in which their magnitude is momentthe country in which the event took place.

magnitude, Mw. Their data set consists of 244 earthquakesOf the 150 entries in Table 1, 52 per cent are for the period

worldwide, of which 127 are associated with surface ruptures,before, and 48 per cent for after 1900. For the first period 31

and 117 with calculated subsurface ruptures. Of the 127 casesper cent of the entries are of category A, 40 per cent of ‘a’, 15

in their first data set only 35 are included in our Table 1,per cent of B and 14 per cent of C. For the present century,which in addition contains another 115 cases not used by86 per cent of the entries are of category A, only 8 per cent ofWells & Coppersmith (1994).‘a’, and the remaining 6 per cent of B and C.

It is interesting to compare magnitudes of the pre-instrumental period, estimated from macroseismic data from

RELATIONS BETWEEN MAGNITUDE ANDeq. (1) (marked with an asterisk in Table 1) with magnitudes

RUPTURE LENGTHpredicted from observed rupture lengths from eq. (3). Thecomparison of these two methods for 26 events shows thatA considerable number of relationships between magnitude,

rupture length, surface displacements and mechanism, using a magnitudes derived from rupture length are on average smaller

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



Figure 3. Results of regression between Ms and log (L ). Curve 1 is eq. (2); curve 2 is eq. (3); curve 3 is eq. (4); curve 4 is eq. (11). L is the length

of faulting in kilometres. Note the effect of the sample distribution on the dependent variable for orthogonal and non-orthogonal regression.

by 0.2 (±0.3) in Ms than macroseismic magnitudes, probably with L in kilometres, which is similar to the empirical relation-

ships given above and in Wells & Coppersmith (1994) and isbecause rupture lengths were actually longer than reported,which is reasonable. a reasonable fit to the earthquakes of M≥6.0 in Fig. 3.

The advantage of this approach over some global empiricalIndeed, it is important, particularly for palaeoseismological

investigations, to have some indication of whether the rupture relationship is that it is more explicit where the assumptionsare: A is known to vary regionally (Ekstrom & Dziewonskilength and offset estimated from historical sources are likely

to be seriously under- or overestimated, given the magnitude of 1988) and so is d. Moreover, for earthquakes in which the

fault length is small compared with the seismogenic thickness,the event. This is a principal use of magnitude–length relation-ships. For an assessment of individual events or particular regions, the relationships between moment and magnitude and between

moment and fault length are both known to be different fromit may be more informative to make such estimates from a

combination of first principles and more closely constrained those given above, such that B#1.0 (Ekstrom & Dziewonski1988) and Mo3L3. Thus a single relationship over the wholeempirical relationships, along the following lines:magnitude range of Fig. 3 (and over the magnitude ranges

(1) for earthquakes that rupture the entire thickness (d) ofdiscussed by Wells & Coppersmith 1994) is not likely to be

the seismogenic upper crust, the downdip width of the fault isvalid anyway. The explicit approach illustrated here is there-

d/sinh, where h is the fault dip, and the moment is thenfore more likely to be useful for detailed palaeoseismological

investigation of specific events.Mo=(mcd/sin h)L2 , (8)

where m is the rigidity modulus and c is the ratio of averageGENERAL OBSERVATIONSdisplacement (u) to fault length (L ), which is observed to be

close to 5×10−5 for intracontinental earthquakes (Scholz 1982;We have attempted to associate the earthquakes in Table 1

Scholz et al. 1986);with a probable style (strike-slip, normal or thrust/reverse) of

(2) both observationally and theoretically it is known thatfaulting. This is often a judgement based on knowledge of the

for such earthquakes the relationship between moment andknown style of faulting in the epicentral region, as the historical

magnitude (M, whether Ms or Mw) is of the formsources are rarely explicit enough to be unequivocal, especially

with horizontal displacements on strike-slip faults. As anlog(Mo)=A+BM , (9)illustration, the earthquake of 1780 in Tabriz (No. 53) was

where A and B are constants, with B #1.5 (e.g. Kanamori &reported with only a vertical displacement, though it almost

Anderson 1975; Ekstrom & Dziewonski 1988);certainly occurred on a right-lateral strike-slip fault, a style

(3) combining these expressions gives a relationship betweenwhich dominates that part of NW Iran (see e.g. Jackson &

moment and fault length of the formAmbraseys 1997).

Of the events listed in Table 1, 35 per cent are associatedM=(1/B) log(mcd/sin h)−A/B+(2/B) ( log L ) . (10)with strike-slip faulting, 28 per cent with normal faulting, and

For illustration, if we take m=3×1010 N m−2, c=5×10−5,only 14 per cent with thrust or reverse faulting (22 per cent

A=9.0 (for Mo in units of N m, see Ekstrom & Dziewonskiare of unknown type). The relatively low number of thrust/

1988), and B=1.5, then for a seismogenic layer of thicknessreverse faults is at variance with compilations of modern fault-

d=15 km and a vertical strike-slip fault (h=90°), the relation-plane solutions in the region, which show many thrust and

ship isreverse mechanisms in western Greece, eastern Turkey, theCaucasus and Iran. The reasons for their under-representationMw=4.9+1.33L , (11)

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



here are probably that: (1) even steep reverse faults may fail distribution of cases depicted in these figures. For instance, in

Fig. 5 the faulting pattern in the modern period shows noto break the surface, where they produce folding instead;(2) thrust faults with a shallow dip rarely break the surface ruptures along much of the East Anatolian fault zone, the

Dead Sea fault zone and Northern Iran.anyway, even in large events. Both effects are known in this

region, for example in the Zagros (e.g. Jackson & McKenzie In contrast, for the historical period, Fig. 4 shows that thesezones had already been ruptured in places before this century,1984) and the Caucasus (Triep et al. 1995), and are probably

responsible for some of the large historical events for which and that the two sets of figures complement each other, with

historical cases often forming a negative or mirror image ofthere is no reported evidence of surface faulting.From the preceding paragraphs it will be seen that not only the distribution of modern cases, and apparent gaps in the

20th century being filled in by historical cases.is historical information regarding surface faulting not always

clear, and is in many cases inconclusive, but also even for a Like the North Anatolian fault zone, which was delineatedby a series of surface fault ruptures during this century fromnumber of earthquakes in the first half of this century evidence

for surface faulting is poor and occasionally insufficient. In east to west, in the last century the conjugate eastern Anatolian

fault and its Levantine extension into the Dead Sea fault zonealmost none of the historical cases do documentary sources,even up to the end of the 19th century, provide more than a were also delineated by a succession of fault breaks.

Truly great earthquakes of M>8.0+ are not easy to identifyminimum of information about faulting, and neither the length

nor the attitude of the break can be deduced with certainty. from historical evidence. The chief difficulty is that it is notalways possible to establish reliably the simultaneity of theirThe benefit of being able to have observations over a period

of almost 20 times longer than this century, however, is obvious. destructive effects at distances of hundreds of kilometres with-

out running the risk of amalgamating two or more separateThe locations of coseismic ruptures in Table 1 are shown inFig. 2. Most are in categories (A) and (a), and are associated events into a great earthquake. A glaring example of such an

amalgamation is the earthquake of 365 July 21 in the Easternwith well-known major fault zones such as the North and

Eastern Anatolian fault zones, the Dead Sea fault system, the Mediterranean (Guidoboni et al. 1989; Ambraseys 1994), themisassociation of which with other earthquakes stretched itsNorthern and Eastern deformation belts of Iran and the

Chaman zone in Pakistan, confirming the long-term and size to 8.3 (e.g. Papazachos & Papazachou 1997) and has leadto speculation and to the development of ‘catastrophe’ theoriesalmost continuous activity of these zones.

Figs 4 and 5 show the data for all categories A to C plotted (e.g. Jacques & Bousquet 1984). Where good evidence exists,

as for instance for Nos. 30, 31, and possibly 19 in Table 1, asseparately for the historical, pre-1894, and modern, post-1893,periods, respectively. The most interesting historical faulting is well as for a few other not yet fully studied events not listed

in Table 1, the historical record does in only very few casesthat which has happened where its occurrence could not be

predicted from 20th century activity or, alternatively, where it suggest magnitudes reaching or exceeding M 8.0.The data in Table 1 are only a fraction of the total numbercould be expected from 20th century seismicity but has not

been observed this century. of events (Ms≥6.0) identified so far for the study area and

they are listed in this table only because there is some evidenceThe importance of Figs 2, 4 and 5 lies therefore not so muchin the similarities but rather in the differences between the of their association with surface faulting. However, although

Table 1 presents a regionally limited and most certainly incom-

plete set of data that cannot, and should not, be used alone toassess long-term seismicity, these data demonstrate an interes-ting pattern in the gross time sequence of surface faulting of

principal fault zones in the region.

SPECIFIC OBSERVATIONS

Some specific observations from this compilation that are

worth highlighting include the following.

(1) The destructive earthquake of 518 AD in Macedonia,Figure 4. Locations of earthquakes associated with surface faulting

which allegedly caused a surface rupture about 40 km long.for the historical pre-instrumental period before 1894.

Exactly where this happened is difficult to ascertain. Althoughthe location of the sites affected cannot be identified today,

the most likely site is the valley of the upper reaches of theVardar river, north-northeast of Gostivar (Ambraseys 1970).(For the locations of place-names given here and in the

following, see the references cited.)(2) The earthquake of 551 AD in central Greece, one of a

series that year, ruptured, in all probability, the eastward

extension of the Delphi fault, which has been quiescent forhundreds of years.

(3) Also in central Greece, the earthquake of 1740 October

5 that ruined Regini, Lamia and Ypati produced small grounddeformations which run south of Lamia towards Regini butFigure 5. Locations of earthquakes associated with surface faulting

for the modern instrumental period 1893–1996. neither their exact location nor their nature are known.

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



(4) The case of faulting in Thrace on 29 July 1752 is were ruptured by the earthquakes of 115 December 13,

interesting because it is located in an area which is considered 1408 December 29, 1759 November 25, 1796 April 26,to be relatively inactive. Equally interesting, for the same 1822 August 13 and possibly of 1837 January 1. Other large-reason, is the location of the large earthquake of 1829 May 5 magnitude events, for which we have no literary evidence forin the region of Drama–Xanthi in Thrace. faulting, confirm the high seismic potential of that region.

(5) We find that many segments of the North Anatolian (9) A major event in 1068 in the Hejaz in northwesternfault zone, including the coastal area of the Sea of Marmara, Arabia is unusual not only because of its location but alsowhich today show only minor activity, were ruptured before because of the evidence, admittedly slight, suggesting a surfacethis century, and that some of the events in central northern rupture, the location of which must be sought in the regionAnatolia were truly large and probably multiple events. of Tabuk.

(6) An earthquake of 110 AD in central Anatolia seems to (10) There are a few cases of faulting in northern Syria inhave been associated with the Tuzgulu fault, south of the 601, 750 and later, but their location is very uncertain.North Anatolian zone, and that of 1544 with the Surgu fault, (11) For the Zagros suture zone in western Iran evidencewest of the East Anatolian zone, but details are lacking. for surface faulting is lacking. This may be due to a lack of

(7) In spite of the large number of relatively small earth- information, or to a lack of large-magnitude earthquakes,quakes (Ms<6.0), surface faulting in Asia Minor and on which is typical of the zone, or to both. However, evenmainland Greece is often poorly expressed and is not often moderately large earthquakes of modern times, such as thereported in the literature. 1972 April 10 Ghir earthquake of Ms 6.9, have failed to

(8) Much of the Eastern Anatolian fault and its southward produce surface faulting in the Zagros, whereas events of theextension into the Ghab, Yammouneh and Roum faults of the same size often produce coseismic surface ruptures in NE Iran.Levantine system, which have been inactive during this century, This may be related to the large thickness of salt above the

basement in the Zagros, preventing ruptures reaching the

surface (see Jackson & McKenzie 1984).

(12) In contrast, in northern and eastern Iran, where 20th

century earthquake faulting is well known, there is literary

evidence for major ruptures, such as that of #280 BC.

Further east, historical information for faulting becomes

scarce and the few cases identified, such as that of the earth-

quake of 1505 July 6 north of Kabul in Afghanistan, are

probably a small sample of the number of cases that actually

involved surface faulting.

Most of the very large (V) and large (L) earthquakes inFigure 6. Locations of very large (Ms≥7.9, solid) and large

(7.0≤Ms≤7.8) earthquakes in Table 1. Table 1 are associated with large strike-slip faults, such as the

Figure 7. Location of Nauzad and Mask on the fault trace associated with the 1493 earthquake, Table 1 no. 39.

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



Figure 8. The 50 km long surface fault break associated with the earthquake of 1912 August 9 between the Gulf of Saros and the Sea of Marmara

(Table 1, no. 85) remained, until recently, imperfectly known. Arrows indicate the fault break; distance between arrows is 50 km. G: the site of

Ganos (modern Gazikoy) on the Marmara coast; white square: the location of Miseli (modern Murselli) near which a strand of the rupture is

shown in the accompanying photographs.

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



Figure 9. Fault zone of interconnected en echelon fractures with offset stream-beds associated with the Dasht-i Bayaz earthquake in eastern Iran

(Table 1, no. 126). Two lines of qanats (underground tunnels, indicated by their access shafts) cross the zone, another abandoned line runs parallel

to the east, part of which follows the 1968 fault break. Numerous abandoned shaft lineaments are evidence that previous ground movements

damaged the underground aqueduct system in the vicinity of the fault zone.

Ambraseys, N., 1975. Studies in historical seismicity and tectonics,North and East Anatolian, the Dead Sea and the ChamanGeodyn. T oday, 7–16, Pub. R. Soc. London.faults, shown in Figs 1(c) and 6, a feature that was expected

Ambraseys, N., 1992. Soil mechanics and engineering seismology,since large earthquakes are generated by long faults.Invited Lecture, Proc. 2nd Natl. Conf. Geotechn. Eng., pp. xxi–xlii,

Thessaloniki.CONCLUSIONS Ambraseys, N., 1994. Material for the investigation of the seismicity

of Libya, L ibyan Studies, 25, 7–22.In conclusion, we may observe that any seismologist at the

Ambraseys, N. & Free, M., 1997. Surface wave magnitude calibrationturn of this century, or any scholar much earlier, could have

for European region earthquakes, J. Earthq. Eng., 1, 1–22.accessed the historical data before his time that we used in Ambraseys, N. & Jackson, J., 1990. Seismicity and associated strain inthis paper. Had it occurred to him to do so he would have central Greece between 1890 and 1988, Geophys J. Int., 101, 663–708.discovered almost all the main deforming belts in the region Berberian, M. & Tchalenko, J., 1976. Earthquakes of southern Zagros

we know today as well as the overall distribution of seismic (Iran): Bushehr region, in Contribution to the Seismotectonics of Iran,

Part II, ed. Berberian, M., Geol. Surv. Iran Rept, 39, 346–358.hazard.Ekstrom, G. & Dziewonski, A., 1988. Evidence of bias in estimation

of earthquake size, Nature, 332, 319–323.ACKNOWLEDGMENTS Guidoboni, E., Ferrari, G. & Margottini, C., 1989. Una chiave di

lettura per la sismicita antica: la ricerca dei gemelli frl terremoto delThis research was supported by the Climatology Programme365 d.C., in I T erremoti Prima del Mille, pp. 552–73, ed.of CEC (DGXII) and is currently supported by a NaturalE. Guidoboni, Ist. Naz. Geof., Rome.

Environment Research Council grant for the study of long-Jackson, J.A. & Ambraseys, N.N., 1997. Convergence between Eurasia

term seismicity and continental tectonics in the Easternand Arabia in Eastern Turkey and the Caucasus, in Historical and

Mediterranean region and the Middle East. It is Imperial Prehistorical Earthquakes in the Caucasus, eds Giardini, D. &College ESEE contribution no. 97/30, and Cambridge Earth Balassanian, S., NATO ASI series 2, 28, 79–90.Sciences contribution no. 5121. Jackson, J. & McKenzie, D., 1984. Active tectonics of the Alpine –

Himalayan Belt between western Turkey and Pakistan, Geophys.

J. R. astr. Soc., 77, 185–264.REFERENCESJacques, F. & Bousquet, B., 1984. La raz de maree du 21 juillet 365

du cataclysme local et la catastrophe consmique, Melanges EcoleAmbraseys, N., 1970. A note on an early earthquake in Macedonia,

Proc. 3rd Europ. Conf. Earthq. Eng., Sofia, 73–78. Franc. de Rome, 96, 423–461.

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



Kanamori, H. & Anderson, D., 1975. Theoretical basis of some $ An eyewitness, describing the effects of the earthquake ofempirical relations in seismology, Bull. seism. Soc. Am., 65, 1254 in central northern Anatolia, informs us that: ‘... As1073–1095. we rode along for three days (between Susehri and Erzincan)

Papazachos, B., Comninakis, P., Hatzidimitriou, P., Kiriakidis, E., we saw a fault in the earth, exactly as it had been split openKyratzi, A., Panagiotopoulos, D., Papadimitriou, E., Papaioannou, C.

in the earthquake, and piles of earth that slid down from& Pavlides, S., 1982. Atlas of isoseismal maps for earthquakes in

the mountains and filled the valleys ... We passed throughGreece 1902–81, Publ. Geoph. L ab. Univ. T hessaloniki, no. 4,

the valley where ... a great lake had welled up in the courseThessaloniki.of the earthquake ...’ [34].Papazachos, B. & Papazachou, C., 1997. I Seismoi T is Elladas, p. 182,

$ From a contemporary history we learn that in the earth-Ziti, Thessaloniki.

Scholz, C.H., 1982. Scaling laws for large earthquakes: consequences quake of 1408 in the Orontes valley in Syria: ‘... The groundfor physical models, Bull. seism. Soc. Am., 72, 1–14. fissured and was thrown down over the distance of one

Scholz, C.H., Aviles, C.A. & Wesnousky, S.G., 1986. Scaling differences barid (20 km), from the town of Qusair to Saltuham...’ [37].between large interplate and intraplate earthquakes, Bull. seism Soc.

$ In the earthquake of 1493 near Birjand, in Iran: ‘... For twoAm., 76, 65–70. farsakhs (12 km) between Nauzad and Mask the ground

Shebalin, N., Karnik, V. & Hadzijevski, D., 1974. Catalogue ofwas fissured to such a depth that the bottom of the crack

earthquakes Part I, 1901–70; Part II, prior to 1901; Part III, Atlaswas invisible...’ [39].of isoseismal maps, UNDP/UNESCO Survey Seismicity of the Balkan

$ A traveller describing the effects of the earthquake of 1825Region, Skopje.in western Mazanderan in Iran, not far from the site of theTriep, R.G., Abers, G.A., Lerner-Lam, A.L., Mishatkin, V.,recent Manjil earthquake of 1990 June 20, says: ‘... BetweenZacharchenko, N. & Starovit O., 1995. Active thrust front of the

Greater Caucasus: the April 29, 1991, Racha earthquake sequence Kuhrud and Bul Qalam there is some evidence that in thisand its tectonic implications, J. Geophys. Res., 100, 4011–4033. locality the shock was associated with permanent ground

Vanek, J., Zatopek, A., Karnik, V., Kondorskaya, N., Riznichenko, Y., deformations. The piers of a masonry bridge built on solidSavarenski, E., Soloviev, S. & Shebalin, N., 1962. Standardization rock and destroyed by the earthquake seemed as if theyof magnitude scales, Izvest. Akad. Nauk., Ser. Geofiz., no. 2,

could never have been intended to support the same arch,pp. 153–158, Moscow.

so different was their parallel ... and the opposite sides ofWells, D. & Coppersmith, K., 1994. New empirical relationships among

the ravine had no doubt suffered displacement.’ [58].magnitude, rupture length, rupture width, rupture area and surface$ Another eyewitness account about the earthquake ofdisplacement, Bull. seism. Soc. Am., 84, 974–1002.

1866 May 12 in eastern Anatolia in Turkey says that: ‘... As

a result of the earthquake the ground was rent; the earth-APPENDIX A: quake fracture led from the village of Halipan in the south,

to the border of the district of Varto, running uninterruptedIn order to give the reader some idea of the substance offor a distance of eight hours’ journey (#30 km) ...’[66].information that can be found in historical documents that

$ A more explicit account about the earthquake of 1874 Maymay refer to surface faulting, we present in this appendix the3 in eastern Anatolia, written by a mining engineer, saystranslation of pertinent parts, taken at random and out ofthat in this earthquake, ‘... The south side of lake Golcukcontext, from original sources. Numbers in brackets refer towas uplifted by a metre or two. The valley at the southeastentries in Table 1.end of the lake, near Kizin and Burnus Han, through which

$ The earthquake of #280 in Rhagae (modern Shahr-e the lake empties itself by a stream running into the TigrisRey in north central Iran) is described by a near- river, was upheaved. Because of this, the stream ceased tocontemporary sources as follows: ‘... Rhagae, in Media, has flow and the lake began to rise. Roads and tracks that ranreceived its name because the earth about the Caspian along its shore were submerged and villages on its marginsGates had been rent by earthquakes to such an extent that were swamped and had to be abandoned. By the end of themany cities and villages were destroyed and the rivers year the water had almost reached the level of the upliftedunderwent changes of various kinds...’ [3]. valley.

$ For the earthquake of 518 AD in Macedonia, a contempor- ‘The valley southeast of Sarikamis was ‘‘rent’’ all the way toary account about the effects of the earthquake includes the Haraba with the southeast of Lake Golcuk uplifted by onefollowing information: ‘... Many mountains throughout the to two metres along a length of about 45 km ...’ [69,70].province (of Dardania) were rent asunder; rocks and forest

trees were torn from their sockets and a yawning chasm 12feet in breadth and 30 000 Roman feet (43 km) in extent APPENDIX B: REFERENCES TO TABLESintercepted and entombed many of the fugitive citizens...’

1 Ambraseys, N., 1963. The Buyin-Zara earthquake of[16].September 1962, Bull. seism. Soc. Am., 53, 705–740.$ The evidence for the earthquake of 551 AD in Greece comes2 Ambraseys, N., 1967. The earthquakes of 1965–66 in thefrom a contemporary source which, in the long narrativePeloponnesus, Greece, Bull. seism. Soc. Am., 57, 1025–1046.adds that: ‘... In that locality where the so-called Schisma3 Ambraseys, N., 1968. An engineering seismology study of(Cleft) is located there was a tremendous earthquake ... Andthe Skopje earthquake of 26 July 1963, in T he Skopjethe earth was rent asunder in many places and formedEarthquake, pp. 35–88, UNESCO, Paris.chasms. Now some of these openings came together again4 Ambraseys, N., 1970a. Early earthquakes in the Near and..., but in other places they remained open, with the conse-Middle East 17–1699 AD; Part I: Documentation of historicalquence that the people in such places are not able toearthquakes in the Middle East; Part II: Historical earthquakesintermingle with each other except by making use of many

detours ...’ [17]; after 17 AD; Part III: ‘North Africa and South-east Europe,

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



UNESCO Repts, Nos SC/1473/69 and SC/2129/70, 450, Paris, 26 Ambraseys, N. & Moinfar, A., 1974. The Firuzabad earth-

quake of 16 August 1958, Ann. Geofis., 27, 1–21.and unpublished data.

5 Ambraseys, N., 1970b. A note on an early earthquake in 27 Ambraseys, N. & Moinfar, A., 1977. The Torud earthquake

of 12 February 1953, Ann. Geofis., 30, 185–200.Macedonia, Proc. 3rd European Conf. Earthq. Eng., 73–78,

Sofia. 28 Ambraseys, N. & Pantelopoulos, P., 1989. The Fokis-Greece

earthquake of 1 August 1870, J. Euro. Earthq. Eng, 3, 10–18.6 Ambraseys, N., 1975. Studies in historical seismicity and

tectonics, Geodyn. T oday, 7–16. 29 Ambraseys, N. & Tchalenko, J., 1969. Dasht-e BayazEarthquake of 31 August 1968, UNESCO Publ. no.7 Ambraseys, N., 1979. A test case of historical seismicity:

Isfahan and Chahar Mahal, Iran, Geogr. J., 145, 56–71. 1214/BMS, Paris.

30 Ambraseys, N. & Tchalenko, J., 1972. Seismotectonic aspects8 Ambraseys, N., 1988. Engineering seismology, J. Earthq. Eng.Struct. Dyn., 17, 1–106. of the Gediz earthquake of March 1970, Geophys. J. R. astr.

Soc., 30, 229–252.9 Ambraseys, N., 1989. Temporary seismic quiescence: SE

Turkey, Geophys. J., 96, 311–331. 31 Ambraseys, N. & White, D., 1996. The seismicity of the

eastern Meditrranean region before the Christian era, ESEE10 Ambraseys, N., 1990. Two little-known 16–17th century

earthquakes in central Greece, Communic. Natl. Grec. Etudes Rep. no. 96–4, 97, Imperial College of Science, London, and

extended summary in Seismology in Europe, pp. 674–679,Sud-est Europ., 2, 75–82, Athens.

11 Ambraseys, N., 1992. Soil mechanics and engineering seis- Icelandic Met. Office, Reykjavik, J.Earthq. Eng., 1.

32 Ambraseys, N. & White, D., 1997. The seismicity of themology Invited Lecture, Proc. 2nd Natl. Conf. Geotechn. Eng.,

Thessaloniki, pp. xxi–xlii. eastern Mediterranean region during the first millennium of

our era, J. Earthq. Eng., 1, 603–632.12 Ambraseys, N., 1997. The earthquake of 1 January 1837 in

southern Lebanon and northern Israel, Ann. Geofis., 40, 33 Ambraseys, N. & Zatopek, A., 1968. The Varyto Ustukran

earthquake of 19 August 1996, Bull. seism. Soc. Am., 58, 47–102.923–936.

13 Ambraseys, N. & Barazangi, M., 1989. The 1759 earthquake 34 Ambraseys, N. & Zatopek, A., 1969. The Mudurnu Valley

earthquake of July 22 1967, Bull. seism. Soc. Am.,59, 521–589.in the Bekaa Valley, J. geophys. Res., 94, 4007–4013.

14 Ambraseys, N. & Finkel C., 1987. Seismicity of Turkey and 35 Ambraseys, N., Moinfar, A. & Tchalenko J., 1972. Ghir

earthquake of 10 April 1972, UNESCO Pub. no. 2789, Paris.neighbouring regions 1899–1915, Ann. Geophys., 5B, 701–726.

15 Ambraseys, N. & Finkel, C., 1987. The Saros-Marmara 36 Ambraseys, N., Moinfar, A. & Peronaci F., 1973. The

Farsinaj earthquake of 13 December 1957, Ann. Geofis., 26,earthquake of 9 August 1912, J. Earthq. Eng. Struct. Dyn.,

15, 189–211. 679–692.

37 Ambraseys, N., Lensen, G. & Moinfar A., 1978. The Patan16 Ambraseys, N. & Finkel, C., 1988. The Anatolian earth-

quake of 17 August 1668, in Historical Seismograms and earthquake of 28 December 1974, UNESCO Publ. no.

SC/GEO/75/134, Paris.Earthquakes of the World, pp. 173–180, ed. Lee, W., Academic

Press, San Diego. 38 Ambraseys, N., Arsovski, M. & Moinfar, A., 1979. The Gisk

earthquake of 19 December 1977 and the seismicity of the17 Ambraseys, N. & Finkel, C., 1995. T he Seismicity of T urkey

and Adjacent Areas 1500–1800, Eren, Istanbul. Kuhbanan fault-zone, UNESCO Publ. no. SC/80/2.1614, Paris.

39 Armijo, R. & Lyon-Caen, H., 1991. E–W extension and18 Ambraseys, N. & Free M., 1997. Surface wave magnitude

calibration for European region earthquakes, J. Earthq. Eng., Holocene normal fault scarps in the Hellenic Arc: possible

rupture of the 464 BC Sparta earthquake, Nature, 351,1, 1–22.

19 Ambraseys, N. & Jackson, J., 1990. Seismicity and associated 137–139.

40 Arpat, E., 1971. 22 Mayis 1971 Bingol depremi; Olu Denizstrain in central Greece between 1890 and 1988, Geophys.

J. Int, 101, 663–708. fay sisteminin Karliova ilicesi ile Hazar Golu arasindaki bolumu,

Report MTA, Ankara.19b Ambraseys, N. & Jackson, J., 1997. Seismicity and strain

in the Gulf of Corinth since 1694, J. Earthq. Eng., 1, 433–474. 41 Arpat, E., 1977. 1975 Lice depremi, Yeryuvari ve Insan, 2,15–27, Ankara.20 Ambraseys, N., Lensen, G., Moinfar, A. & Pennington W.,

1981. The Pattan (Pakistan) earthquake of 28 December 1974: 42 Arpat, E. & Bingol, E., 1969. The rift system of western

Turkey, Bull. Min. Res. Explor. Inst., 73, 1–9, Ankara.field observations, Q. J. eng. Geol., 14, 1–16.

21 Ambraseys, N. & Melville, C., 1982. A History of Persian 43 Arpat, E., Saroglu, F., 1972. The East Anatolian Fault

system: thoughts on its development, Bull. Mineral. Explor.Earthquakes, Cambridge University Press, Cambridge.

22 Ambraseys, N. & Melville C., 1988. An analysis of the Inst., no. 78, 33–39, Ankara.

44 Arpat, E., Saroglu, F. & Iz, H., 1977. 1976 Caldiran depremi,eastern Mediterranean earthquake of 20 May 1202, in

Historical Seismograms and Earthquakes of the World, Yeryuvari ve Insan, 2, 29–41, Ankara.

45 Arsovski, M., 1970. Contemporary tectonic properties ofpp. 181–200, ed. Lee, W., Academic Press, San Diego.

23 Ambraseys, N. & Melville C., 1989. Evidence for intraplate some seismic active zones in Yugoslavia, Proc. 3rd Europ. Conf.

Earthq. Eng, pp. 181–88, Sofia.earthquakes in north-west Arabia, Bull. seism. Soc. Am., 79,1279–1281. 46 Barka, A., 1992a. The North Anatolian fault-zone, Ann.

T ect., 6, 164–195.24 Ambraseys, N. & Melville, C., 1995. Historical evidence of

faulting in eastern Anatolia and northern Syria, Ann. Geofis., 47 Barka, A., 1992b. Surface cracks of the March 13 1992

Erzincan earthquake, Bogazici, Univ. Spec. Publ., 68–79.38, 337–43.

25 Ambraseys, N. & Melville, C., 1996. A parametric earth- 48 Barka, A. & Kadinski-Cade, K., 1988. Strike-slip fault

geometry and its influence on earthquake activity in Turkey,quake catalogue of Iran 700–1996, ESEE Research Report

96/4, London. T ectonics, 3, 663–684.

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



49 Barka, A. & Eyidogan, H., 1993. The Erzincan earthquake 67 Eyidogan, H. & Barka, A., 1996. The 1 October 1995 Dinar

of 13 March 1992 in eastern Turkey, T erra Nova, 5, 190–194. earthquake, SW Turkey, T erra Nova, 8, 479–485.50 Berberian, M., 1979. Earthquake faulting and bedding 68 Fara, A., 1976. Study of recent seismotectonics in Pakistan,thrust associated with the Tabas-e Golshan (Iran) earhquake Rept CENT O Working Group on Recent T ectonics, Istanbul.of September 16 1978, Bull. seism. Soc. Am., 69, 1861–1887. 69 Goudarzi, K. & Ghaderi-Tafreshi, M., 1976. The Qaen51 Berberian, M., 1982. Aftershock tectonics of the Tabas-e earthquake of November 71976, Proc. CENTO Earthq. HazardGolshan (Iran) earthquake sequence: a documented active Minim. Seminar, 301–309, Tehran.thin-and thick-skinned tectonic case, Geophys. J. R. astr. Soc., 70 Griesbach, C.L., 1893. Notes on the earthquake in68, 499–530. Baluchistan on the 20th December 1892, Rec. Geol. Surv. India,52 Berberian, M. & Tchalenko, J., 1976. Earthquakes of Pt.2, 57–61, & Pt.1, p.27 (1898).southern Zagros (Iran): Bushehr region, ed. M. Berberian 71 Guidoboni, E., Comastri, A. & Traina, G., 1994. CatalogueContribution to Seismotect. Iran, Part II, Geol. Surv. Iran Rep., of Ancient Earthquakes in the Mediterranean Area up to the39, 346–358. 10th Century, Istituto Naz. di Geofisica, Rome.53 Berberian, M. & Papastamatiou, D., 1978. Khurgu (north 72 Gulkan, P., Gurpinar, A., Celebi, M., Arpat, E. &Bandar Abbas, Iran) earthquake of March 21 1977, Bull. seism. Gencoglu, S., 1978. Engineering report on the Muradiye-Soc. Am., 68, 411–428,. Caldiran, Turkey, earthquake of 24 November 1976, Publ.54 Berberian, M., Asdudeh, I., Bilham, R., Scholtz, C. & Natl. Res. Counc. Natl. Akad. Sci., Washington, DC.Soufleris, C., 1979. Mechanism of the main shock and the 73 Haghipour, A. & Amidi, M., 1979. Geotectonics of theaftershock study of the Tabas – Golshan (Iran) earthquake of Ghaenat earthquakes of NE Iran, Nov. 14 to Dec. 9, 1979,September 16 1978; a preliminary report, Bull. seism. Soc. Am., Rept Geol. Surv. Iran, Tehran.69, 1851–1859. 74 Hanks, T. & Wyss, M., 1972. The use of body-wave spectra55 Berberian, M., Asudeh, I. & Arshadi, S., 1979. Surface in the determination of seismic source parameters, Bull. seism.rupture and mechanism of the Bob-Tangol (SE Iran) earth- Soc. Am., 62, 561–589.quake of 19 December 1977, Earth planet. Sci. L ett., 42, 75 Hatzfeld, D., Nord, J. et al., 1995. The Kozani-Grevena456–462. (Greece) earthquake of May 13 1995, Seism. Res. L ett., 66,56 Berberian, M., Jackson, J., Ghorashi, M. & Kadjar, M., 61–70.1984. Field and teleseismic observations of the 1981 Golbaf- 76 Heron, A.M., 1911. The Baluchistan earthquake of the 21stSirch earthquakes in SE Iran, Geophys. J. R. astr. Soc, 77, October 1909, Rec. Geol. Surv. India, 41, 22–35.809–838.

77 Jackson, J. & Fitch, A., 1979. Seismotectonic implications57 Berberian, M., Qorashi, M., Jackson, J., Priestley, K. &

of relocated aftershock sequence in Iran and Turkey, Geophys.Wallace, T., 1992. The Rudbar-Tarom earthquake of 20 June

J. R. astr. Soc., 57, 209–229.1990 in NW Persia: preliminary field and seismological obser-

78 Jackson, J., Gagnepain, J., Houseman, G., King, G.,vations, and its tectonic significance, Bull. seism. Soc. Am., 82,

Papadimitriou, D., Soufleris, C. & Spencer, C., 1982. Seismicity,1726–1755.

normal faulting and geomorphological development of the58 Bezzeghoud, M., Deschamps, A. & Madariago, R., 1986.

Gulf of Corinth – Greece, Earth planet. Sci. L t., 57, 377–397.Broad-band modelling of the Corinth, Greece earthquakes of

79 Jackson, J. & McKenzie, D., 1984. Active tectonics of theFebruary and March 1981, Ann. Geophys., 4, 295–304.

Alpine – Himalayan Belt between western Turkey and59 Blumenthal, M., 1943. Zur Geologie der Landstrecken der

Pakistan, Geophys. J. R. astr. Soc., 77, 185–264.Erdbeben von Ende 1942 in Nord-Anatolien und dortselbst

80 Kadinsky-Cade, K. & Barka, A., 1989. Relationship betweenausgefuhrte makroseismische Beobachtungen, Maden T etkik

restraining bends and earthquake magnitude: large earth-ve Arama Enst., no. 1/29, 33–58, Ankara.quakes in strike-slip zones, USGS Workshop Fault Segment &60 Blumenthal, M., 1945. Ladik deprem hatti Samsun ili MadenControls, Open File Rept no. 89/315, 181–192.T etkik ve Arama Enst., no. 1/33, 153–179, Ankara.81 Kadinsky-Cade, K., Rouhban, B., eds, 1989. Proc. Int61 Bommer, J. & Ambraseys, N., 1989. The Spitak (Armenia)Seminar on Spitak-88 Earthquake, Yerivan.earthquake of 7 December 1988, J. Earthq. Eng. Struct. Dyn.,82 Keightly, W., 1975. Destructive earthquakes in Burdur and18, 921–925.Bingol, Turkey, Publ. Natl. Acad. Sci., Washington, DC.62 Dewey, J.W., 1976. Seismicity of Northern Anatolia, Bull.83 Ketin, I., 1966. 6 ekim 1964 Manyas depremi esnasindaseism. Soc. Am, 66, 843–868.zeminde meydana gelen tansiyon catlaklari, T urk. Jeoloj.63 Erdik, M., Aydinoglu, N., Pinar, A. & Kalafat, D., 1995.Kurumu Bult., 10, 44–51, Ankara.October 11995 Dinar (Turkey) earthquake, Rept Bogazici84 Ketin, I., 1969. Uber die nordanatolische horizontal verschi-University, Istanbul.ebung, Bull. Mineral. Res. Explor. Inst., 72, 1–28, Ankara.64 Erentoz, C. & Kurtman, F., 1964. Rapport sur le tremble-85 Ketin, I. & Roesli, F., 1954. Macroseismischement de terre de Manyas survenu en 1964 Bull. Mineral. Res.Untersuchungen ueber das nordwestanatolische Beben vomExplor. Inst. T urkey, no. 63, 1–6, Ankara.18 Marz 1953, Eclogae Helvet., 46, 187–208.65 Eyidogan H., 1988. Rates of crustal deformation in western86 Kim, W., Kulhanek, O. & Meyer, K., 1984. Source processesTurkey as deduced from major earthquakes, T ectonophysics,of the 1981 Gulf of Corinth earthquake sequence from body-148, 83–92.wave analysis, Bull. seism. Soc. Am., 74, 459–477.66 Eyidogan, H. & Jackson, J., 1985. A seismological study of87 King, G. & Nabelek, J., 1985. Role of the fault bends in thenormal faulting in the Demirci, Alashehir and Gediz earth-initiation and termination of earthquake rupture, Science,quakes of 1979–7 in western Turkey: implications for the228, 984–987.nature and geometry of deformation in continental crust,

Geophys. J. R. astr. Soc., 81, 569–607. 88 Kiroff, K.T., 1935. Studien der Erdbeben in Sud-Bulgarien

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



von 14 und 18 April 1928, Sborn. Biulg. Akad. Nauk., 29, of April 30 1954, in Sophades, central Greece, Geophys. J. R.astr. Soc, 87, 885–895.1–116, Sofia.

89 Kociaj, S. & Sulstarove, E., 1980. The earthquake of June 110 Papazachos, B., Moutrakis, A., Psilovikos, A. &Leventakis, G., 1979. Surface fault traces and fault plane1, 1905, Skodra, Albania, T ectonophysics, 67, 319–32.

90 Kocyigit, A., 1988. Basic geological characteristics and total solution of the May-June 1978 shocks in the Thessaloniki area,North Greece, T ectonophysics, 53, 171–183.offset of the North Anatolian fault zone in the Sushehri area

NE Turkey, J. Pure Appl. Sci., no. 228, 984–97, Ankara. 111 Papazachos, B., Comninakis, P. et al., 1982. Atlas ofIsoseismal Maps for Earthquakes in Greece 1902–81, Publ.91 Kudo, K., 1983. Seismic source characteristics of recent

major earthquakes in Turkey, Comprehensive Study of Geoph. Lab. University of Thessaloniki, no. 4, Thessaloniki.112 Papazachos, B., Panagiotopoulos, D., Tsapanos, T.,Earthquakes in Turkey, Publ. Eng. Faculty, Hokkaido

University, Sapporo. Mountrakis, D. & Dimopoulos, G., 1983. A study of the 1980summer seismic sequence in Magnesia region of Central92 Kulhanek, O. & Meyer, K., 1979. Source parameters of the

Volvi-Langadhas earthquake of June 20 1978 deduced from Greece, Geophys. J. R. astr. Soc., 75, 155–168.

113 Parejas E., Pamir H., 1939. Le tremblement de terre dubody-wave spectra at stations UPP and KIR, Bull. seism. Soc.Am., 69, 1289–1294. 19 avril 1938 en Anatolie centrale, Istanb. University, Fen

Fakult. Mecmuas., 4, 183–193, Istanbul.93 Lyon-Caen, H., Armijo, R. et al., 1988. The 1986 Kalamata

(South Peloponnesus) earthquake: detailed study of a normal 114 Pavlides S. & Tranos M., 1991. Structural characteristicsof two strong earthquakes in the North Aegean: Ierissos 1932fault evidence for east–west extension in the Hellenic Arc,

J. geophys. Res., 93, 14 967–15 000. and Agios Efstratios 1968, J. struct. Geol., 13, 205–214.

115 Pavlides, S., Zouros, N., Chatzipetros, A., Kostopoulos, D.94 McEvilly, T.V. & Niazi, M., 1975. Post-earthquake obser-vations at Dasht-e Bayaz, Iran, T ectonophysics, 26, 267–280. & Mountrakis, D., 1995. The 13 May 1995 western Macedonia,

Greece, earthquake: preliminary results, T erra Nova, 7,95 Mercier, J., Mouyaris, N., Simeakis, C, Roundoyannis, T.

& Angelidhis, C., 1979. Intraplate deformations: a quantitative 544–549.116 Perejas, E. & Pamir, H., 1939. 19.4.1938 Orta Anadolustudy of the faults activated by the 1978 Thessaloniki earth-

quakes, Nature, 278, 45–48. yerdepremi, Istambul University Fakult. Mec., B4, Istanbul.117 Petrescu, G. & Purcaru, G., 1964. The mechanism and96 Mercier, J., Garey-Gailhardis, E., Mouyaris, N., Simeakis,

C., Roundoyannis, T. & Anghelidis, C., 1983. Structural analy- stress pattern at the focus of the September 1, 1962, Buyin-

Zara (Iran) earthquake, Ann. Geophys., 20, 1–6.sis of recent and active faults and regional state of stress inthe epicentral area of the 1978 Thessaloniki earthquake, 118 Pinar, N., 1953. 31 agustos 1951 Kursunlu depreminin

jeolojik ve makrosismik etudu, Rev. Facult. Sci. Univ. Istanbul,T ectonics, 2, 577–600.

97 Mirkov, M., 1930. Pretsizni nivelachii izmirvania v yuzhno Ser.A, 18, 131–141.119 Richter, C., 1958. Elementary Seismology, Freeman, Sanbiulgarskata zemetriasna oblast, Doklad. Geograf. Inst., no.

750, Sofia. Francisco, CA.

120 Roberts, G. & Koukouvelas, I., 1966. Structural and98 Mohajer, G.A. & Pierce, G.R., 1963. Qazvin, Iran earth-quake, Bull. Am. Petrol. Geol., 47, 1878–83. seismological segmentation of the Gulf of Corinth fault system:

implications for models of fault growth, Ann. Geofis., 39,99 Moinfar, A. & Naderzadeh, A., 1990. T he Manjil, Iran,earthquake of 20 June 1990, no. 119, Ministry of Housing & 619–646.

121 Rustanovich, D. & Shirokova, E., 1964. Some results ofUrban Devel., Tehran.100 Niazi, M., 1968. Fault rupture in the Iranian Dasht-e the study of the Ashkhabad earthquake of 1948, Izvest. Akad.

Nauk, Ser.Geoph., 12, 1077–1080.Bayaz earthquake of August 1968, Nature, 220, 569–570.101 Niazi, M., 1969. Source dynamics of the Dasht-e Bayaz 122 Salomon-Calvi, W., 1940. Das Erdbeben von Kirsehir von

19 April 1938, Maden T etkik ve Arama Enst., no. 1/31, 34–38,earthquake of August 31 1968, Bull. seism. Soc. Am., 59,1843–1861. Ankara.

123 Schmidt, J., 1879. Studien ueber Erdbeben, pp. 68–83,102 Niazi, M. & Kanamori, H., 1981. Source parameters of1978 Tabas and 1979 Qainat, Iran, earthquakes from long- 196–197, Leipzig.

124 Seymen, I. & Aydin, A., 1972. The Bingol earthquake faultperiod surface waves, Bull. seism. Soc. Am., 71, 1201–1213.103 Niazi, M. & Bozorgnia, Y., 1992. The 1990 Manjil, Iran and its relation to the north Anatolian fault zone, Bull. Mineral.

Res. Explor. Inst., no. 79, 1–8, Ankara.earthquake and seismology overview, PGA attenuation and

observed damage, Bull. seism. Soc. Am., 82, 774–799. 125 Shebalin, N., Karnik, V. & Hadzijevski, D., 1974. Catalogueof earthquakes Part I 1901–70; Part II, prior to 1901; Part III,104 North, R.G., 1977. Seismic moment, source dimensions

and stress associated with earthquakes in the Mediterranean Atlas of isoseismal maps, UNDP/UNESCO Survey Seismicityof the Balkan Region, Skopje.and Middle East, Geophys. J. R. astr. Soc., 48, 137–161.

105 Nowroozi, A. & Mohajer, A., 1985. Fault movements and 126 Soufleris, C. & Stewart, G., 1981. A source study of the

Thessaloniki (northern Greece) earthquake sequence, Geophys.tectonics of eastern Iran: boundaries of the Lut plate, Geophys.J. R. astr. Soc., 83, 215–237. J. R. astr. Soc., 67, 343–358.

127 Soufleris, C., Jackson, J., King, G., Spencer, C. & Scholz,106 Ocal, N., 1959. 26 mayis 1957 Abant zelzelesi, Sism.Yayinlari, no. 4, Istanbul Kandilli Rasathan., Istanbul. C., 1982. The 1978 earthquake sequence near Thessaloniki

(northern) Greece, Geophys. J. R. astr. Soc., 68, 429–458.107 Pamir, H.N., 1943. Corum ve Erbaa depremleri, T urkCograf. Dergis., 1, no. 2, 1–7, Ankara. 128 Stavrakakis, G., Drakopoulos, J., Latoussakis, J.,

Papanastassiou, D. & Drakatos, G., 1989. Spectral character-108 Pamir, H.N. & Akyol, I., 1943. Corum ve Erbaa depremleri,T urk Cograf. Derg., 1, no. 2, 1–7, Ankara. istics of the 1986 September 13 Kalamata (southern Greece)

earthquake, Geophys. J. Int., 98, 149–157.109 Papastamatiou, D. & Mouyaris, N., 1986. The earthquake

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from



129 Stavrakakis, G., Blionas, S. & Goutis, C., 1991. Dynamic migration of earthquakes along the North Anatolian Fault

Zone and seismic gaps, Geophysics, 117, 1258–70.source parameters of the 1981 Gulf of Corinth (central Greece)earthquake sequence on FFT and iterative maximum entropy 141 Tomblin, J., 1981. The Kerman, Iran earthquake of 28 July

1981, Mission Reports 12.8.81, United Nations Disast. Relieftechniques, T ectonophysics, 185, 261–275.

130 Sulstarove, E. & Kociaj, S., 1969. Termeti i 30 Nendorit Org. (UNDRO), Geneva.142 Trifonov, V., Bayractutan, M., Karakhanian, A. & Ivanova,1967 dhe brezi sizmogjen Vlore-Debar, Bull. Univ. Shtet.

T iranes, Ser.Shken.Nat., 2, 65–94, Tirana. T., 1993. The Erzincan earthquake: tectonic aspects, T erraNova, 5, 184–189.131 Tasdemiroglu, M., 1971. The Gediz earthquake in Western

Anatolia Turkey, Bull. seism. Soc. Am., 61, 1507–1527. 143 Vanek, J., Zatopek, A., Karnik, V., Kondorskaya, N.,Riznichenko, Y., Savarenski, E., Soloviev, S. & Shebalin, N.,132 Tasman, Y., 1944. Gerede-Bolu depremi, Maden T etkik ve

Arama Enst., no. 1/31, 34–38, Ankara. 1962. Standardization of magnitude scales, Izvest. Akad. Nauk.,Ser. Geofiz., 153–158, Moscow.133 Tasman, Y., 1946. Varto ve Van depremleri, Maden T etkik

ve Arama Enst., no. 2/36, 287–291, Ankara. 144 Wallace, R., 1968. Earthquake of August 19, 1996, Varto

area, Eastern Turkey, Bull. seism. Soc. Am., 58, 11–56.134 Taymaz, T., Eydogan, H. & Jackson, J., 1991. Sourceparameters of large earthquakes in the East Anatolian fault 145 Wells, D. & Coppersmith, K., 1994. New empirical relation-

ships among magnitude, rupture length, rupture width, rupturezone (Turkey), Geophys. J. Int., 106, 537–550.

135 Tchalenko, J. & Ambraseys, N., 1970. Structural analysis area and surface displacement, Bull. seism. Soc. Am., 84,974–1002.of the Dasht-e Bayaz earthquake fractures, Bull. Geol. Soc.

Am., 81, 41–60. 146 West, W.D., 1945. Preliminary geological report on the

Baluchistan earthquake of May 31st 1935, Rec. Geol. Surv.136 Tchalenko, J. & Berberian, M., 1974. The Salmas (Iran)earthquake of May 6th 1930, Ann. Geofis, 27, 151–212. India, 69, 203–240.

147 Yates, R., Shieh, K. & Allen, C., 1996. Geology of137 Tchalenko, J.S. & Berberian, M., 1975. Dasht-e Bayaz

fault, Iran; earthquake and related structures in bed rock, Bull. Earthquakes, Oxford University Press, Oxford.148 Zare, M. & Moinfar, A., 1994. Comment on the Rudbar-Geol. Soc. Am., 86, 703–709.

138 Toksoz, M.N., Arpat, E. & Saroglu, F., 1977. East Tarom earthquake of 20 June 1990, Bull. seism. Soc. Am.,84, 484–485.Anatolian earthquake of 24 Nov. 1976; field observations,

Nature, 270, 423–25. 149 Zare, N., 1995. Earthquake faulting and seismotectonics

of the Lut area; an observation on the Sefidabeh 1994 earth-139 Toksoz, M.N., Nabelek, N. & Arpat, E., 1978. Sourceproperties of the 1976 earthquake in east Turkey; a comparison quake fault, Proc. 2nd Intern. Conf. Seism. Earthq. Eng.,

179–186, Tehran.of field data and teleseismic results, T ectonophysics, 49,199–205. 150 Zelkov, Ya., 1929. Svedeniya po zemledielneto, Byulet. na

Infrom. Sluzhba, 10, pts 1 and 2, Sofia.140 Toksoz, N., Shakal, A. & Michael, A., 1979. Space-time

© 1998 RAS, GJI 133, 390–406


Dow

nloaded from


faulting and seismicity: eastern mediterranean

Education

cases faulting

cases of faulting pre

historical earthquakes

surface faulting canbe

number of cases

interesting cases

century activity

studies of active tectonics