Bulletin of the Geological Survey of Japan，voL41（3），p．99－114，1990 539．128．4551．243．1（560－17）

A恥恥融龍麗k職膿躍騰灘e鵬s f暇fa翻醗睡恥⑪酋恥e置醜丁眼艦ey

Hirokazu KATo＊，：Kan KAToH：＊＊，Aykut BARKA＊＊＊and Ismail Ku婁gu†

KATo，H．，KAToH，K．，BARKA，A．and Ku婁gu伊1．（1990） Alpha track measurements for faults in northem Turkey．B％ll．0601．S％7∂．如醐，vol．41（3），p．99－114．

Abs色盟c重：Two kinds of methods of alpha track measurement were carried out for faults inclu（1ing earthquake faults，active faults and non－active geological faults in northern Turkey．One is the“cup method”for detecting the fault trace and the other is the“pipe method”for evaluating the fault activity。， The results by the cup method show that the alpha track measurement is useful to detect not only active faults including earthquake ones but also non－active ones even in the region containing a little radon．They also show the good reproducibility of the alpha track measurement if they are measured in nearly the same temperature condition． The results by the pipe method show that the seasonal change depended on the temperature mainly．But in both of two methods，the underground water supply must be taken into consideration before the evaluation of fault activity． In Mekece region，northwestem Turkey，the seismic gap area of one branch of the North Anatolian Fault Zone，the maximum alpha track density obtained by the both methods is abnormaly high．Furthermore，the result by the pipe method shows the highest value even in winter．These facts indicate that this branch fault of the North Anatolian fault is very active．

The authors carried out the geological 夏． 亙醜tro西盟e髄o醜 investigation of some faults including earth－ quake faults．and active faults in the North The North Anatolian Fault Zone，about1， Anatolian Fault Zone in northem Turkey， 200 kilometers Iong，is a right lateral employing the alpha track etching method． intracontinental transform bomdary between In recent years，measurements of subsur－ the Eurasian and the Anatolian Plates in face ra（10n concentration have been used to Turkey．This fault zone trends E－W in gen・ detect faults，particularly active faults and to eral with a slight convex to the north in its predict earthquake occurrence （e，9．KING， central part．The seismic activity has been 1980）． intense along this fault zone，a．nd the we11－ Measurements of changes in radon concen－ known earthquake sequence between 1939 tration by the alpha track etching method and1967produced almost continuous surface were carried out in the search for the follow－ breaks，shifting from east to west with age ing faults． along the main part of this fault zone． a） surface breaks resulted from the 1939 Erzincan Earthquake at the northwestem ＊Geology Department margin of the Erzincan Basin， ＊＊Environment Geology Department b）surface breaks resulted from the1944 榊（former）General Directorate of Mineral Gerede－Bolu Earthquake to the west of Bolu Research and Exploration（MTA） †General Directorate of Mineral Research and and the active fault segment showing the Exploration（MTA） creep movement at Ismetpa多a，

一99一 Bz61」6ガ％ （ゾ 孟h6 G60Jo9ぢ6αl Sz67z76ッ ρプノ41）ごz％， Vてol．41，〈ろo．3

EURASIAN（BL継＿SEA）PLATE l of Turkey（MTA），Ministry of Energy and 葬…華欝1織！昌 Natural I之esources， Turkey， entitled in “Research on earthquake faults，active faults and earthquake prediction” （1982－1984fiScal ～乱、、、滞K匿 years）and“Research on Quatema．ry crustal 雇ア movement and earthquake prediction”（1985－ ／ ARABIAN 1987fiscal years）．These projects have been 博R、c慧詫翼ノκ10／ △PLATE conducted within the framework of“the Insti－ tute for Transfer of Industrial Technology PL醒E ・ （ITIT）Project”organized by Intemational Fig．1Map showing investigated areas．A black 幽 Research and Dovelopment Cooperation Divi－ arrow indicates slip sense of transform sion，AIST，MITI，Japan． fault and a white arrow indicates that of plate motion．NAFZ：North Anatolian The authors are indebted to Mr．E．Tsu－ Fault Zone，EAF：East Anatolian Fault． kuda of the Geological Survey of Japan for the setting the pipes of alpha track measure－ ments in Turkey．We also wish to thank c）newest surface breaks resulted from the Messrs．1．gHAKMAK and H．Sov of MTA for 1983Horasan－Narman Earthquake to the their help，kindness and warm welcome． east of Erzurum in eastem Turkey， d）act圭ve fault segment at the one branch of 麗：。 Pr直醜c亘瞬e o£ a恥臨匙 tr我ek etch亘ng the NorthAnatolian Fault Zone near Mekece 朧e蝕憾 in northwest Turkey，where an earthquake occurrence is afraid in the near future， Radon measurements are carried out by e）non－active reverse faults in the SinoP detecting alpha particles which are emitted region facing the Black Sea，100kilometers from radioactive decay of radon and its north of the North Anatolian Fault Zone． daughters．As the can。ier fluid including those These surveyed areas are indicated by alpha particles flows up through faults and squares in Figure1． associated fractures in the basement rocks Most results of alpha track measurements and covered sediments，the radon concentra－ in Turkey have been already published frag－ tion is expected to attain a maximum value mentarily（KATo，1984，1986；KAToαα1．， on the fault trace． Radon concentration 1985，1989），however enough consideration changes in response to the vertical flow of was not made．The authors outline some subsurface fluid occurred in association with faults described previously as metioned above earthquake faulting and is expected to in northem Turkey，compile those prelimi－ increase markedly just before an earthquake． nary results of alpha track measurements by The authors tried two methods of alpha the cup method，and important remarks track measurements as shown in Figure2． obtained by the pipe method and discuss Both methods use small strips of cellulose better way for earthquake prediction． nitrate film which is sensitive to alpha radia－ This study is a part of the joint research tion，and explained as follows，here temporar・ projects operated by the Geological Sur▽ey of ily called the cup method and pipe method． Japan（GSJ），Agency of Industrial Science Cup method：This is an easy and useful and Technology（AIST），Ministry of Intema－ way to detect the fault trace．Small plastic tional Trade and Industry（MITI），Japan，and cups to which inside bottom film strip is Mineral Research and Exploration Institute attached with adhesive tape，are placed

一100一 ノ1々）hごz渉名αo左 ηz6㏄銘γε盟z6η孟sノわ7カ％」云s わz7zo7孟h67％ Tz6名たツ （κごz∫0 6オごzl．）

Pl麗睡丁開OD CUP繭妊丁開OD particles generated by radon isotopes．The rub，ber5曾◎pper alpha tracks are counted under the micro－ scope and calculated by T．D．＝N／（S×T） where T．D．is alpha track density，N is the number of tracks，S（cm2）is area of the film and T is the exposure time（day）．Obtained track density represents the average concen－ tration of radon isotopes at the measuring point．

麗．：瞭z甑a囎：Regi樋

亘盟一l Geo恥gic島璽 我脳丑 se飽朧o且09亘e翻 set回 崩囎gs The Erzincan Basin is one of the Neogene－ Quatemary intramomtain basins along the North Anatolian Fault Zone．The Erzincan Basin measures about50km in length and10 皆6c謂 km in width．The basin is situated at1，150m－ Fig．2 Schematic illustration of two kinds 1，130mhighabovesealeveL of alpha track measurement（pipe Miocene marine sedimentary rocks mcon－ method and cup method）。 formably overlies the basement rocks of pre－ Neogene age． upside－down in shallow holes at a depth of Volcanoes of、Pliocene－Pleistocene age with about a few tens centimeters．They are well preser▽ed conical shapes are arranged spaced at intervals of about a few meters on sporadically on the subsidiary faults which a line which is normal to probable fault trace， mn nearly parallel to the North Anatolian and then are covered with soil and rock frag－ Fault Zone．These volcanoes are composed of ments whose thicknesses are about10centi－ a wide variety of rocks facies such as meters，and are retrieved about one week or rhyolite，dacite and andesite． more later． The basin is filled with Holocene gravels in Pipe method：At one site radon concentra－ which minor faults develop associated with tion is to be measured repeatedly in such a the active movement of the North Anatolian long term as half year to about a year by this Fault Zone．The Erzincan Earthquake whose method because retriving and changing films magnitude was7．90n a Richter scale occur－ is much easier than the cup method．Plastic red on26December，1939aromd theErzincan pipes with a diameter of about6centimeters Basin in eastem Turkey，although its precise and a length of about50centimeters are epicentral situation may not be reliable．This placed on and arom（1a fault line。The film earthquake produced a350km long fault strip is suspended from the rubber cap put on break between Erzincan and Amasya．The the top of the pipe and is replaced to new one maximum horizontal displacement of right for each measurement。 Iateral sense is about350cm and the maxi－ Retrieved films are processed in 20％ mum vertical displacement of south side up is NaOH solution for100minutes at60。C to about200cm（KETIN，1948）． produce visible track－1ike images of alpha The Erzincan Basin is situated along or

一101一 Bz61」θ’1％ （～プ 孟h6 （セoJo9／641S％7z／6ッ （ゾノ4ψθn， レリ乙4ヱ，．八石o．3

2⑯⑪m・

臨1曜 眉5

負恥観

唱騰 冨 3。3：：：：言。39：匿。馨23● 百 ・ ㊧鞠趣聡 5⑪

｝恕 臨1回r 、ヴ＝ F 注 昼 竈騒層⑪ 唱3 5m 』 胸k鵬

3⑭。45’醐 ．輪鞭欝縫・・ 鷺レ㌧嬉レレ、誓、 レ 唱5⑪ 工D． ／’ 3雪も⑭，匿騰撫鋸、 賜睡r ロ 髄⑪ 鵡’、！，， ’趨 ノド ぎ嬢”・ンーマ＝臼 魯 幅！ 5 奮 2 5 囁 7 IF 皿皿圏 幽幽 團 冒 圃 占鼠 層 魑 5鵬 躯 琶 Fig。3 Simplified geologic map of the Erzincan region with results of the alpha track measure－ ments（modified from KATo，1984）．1：Pre－Neogene basement rocks，2：Pliocene sedi－ mentary rocks，3：Pliocene－to Pleistocene sedimentary rocks，4：Quatemary vo王canic rocks，5：Fan deposits including terrace deposits，6：Salt playa，7：Debris flow deposits， 8：Alluvium，9：strike slip fault，10：thmst，T．D．：track density（track numbers／m2・ day），F：10cation of fault，Numbers on the axis of abscissa show the sites of measure・ ments starting with l at the northem end to the south．Solid line shows the result in1983 and dashed line shows the result in1984．

around the eastem terminal of this fault a remarkable reverse displacement in’the break． In detail the earthquake fault is west of Mihar Village．Small hills line up divided into several segments such as the intermittently just on the south of this seg－ Mihar，Bahik，H：111r，Altintepe and Tanyeri ment．These hills were resulted from the segments from northwest to southeast in and cumulation of vertical displacements along around the basin as shown in Figure3． The this segment． Mihar（A），Bahik（B）and H111r（C） seg一 The Bahik segment，about16km long， ments gave markable results． strikesN55。Wanddips84。NE．Thissegment The Mihar segment，about17km long， αnd subsidiary faults can be observed in a strikesN80。Wandhasanearlyverticaldip． small dried－up valley where fractured serpen－ This segment displaces numerous ridges and tine and colluvial deposits are exposed to the valleys dextrally．Therefore，this segment east of Bahik Village．Horizontal striations shows the right lateral sense as a whole，but it indicating the right lateral sense of this seg－ seems to be a high angle strike－slip fault with ment develop on the fault plane accompanied

一102一 ・4ゆhα加盈耀αε膨耀錫云sか加1云s初％碗h醐丁％名勿（κα孟oα41．） by fault clay． At the Mihar segτnent，although the earth・ TheH111rsegmentstrikesN65－80。Wand quake fault is believed to run near the has a nearly vertical dip．To the southwest of measuring point No．110f the first measure－ Hlhr，it can be observed that serpentinite on ment in1983，the remarkable peak of alpha the northem side of the segment is much track density was not recognized at the point． crashed but the gravel bed of Pliocene－ It seems that the track density rather meven－ Quatemary age is little fractured． ly，increases from north to south especially The alingment of these segments does not from the point No．11to the point No．18． always show a right lateral echelon pattem These points are located on the fault scarp． although each segnlent and the North The result of the second measurement in1984 Anatolian Fault Zone show the right lateral is similar to that圭n1983in genera1． movement at present time as a whole． At the Bahik segment，two peaks of track 購一2 A亘轟a t聡ekl職eas狸e瞳e虚s㎞蝕e density are at points No．6and No．9in both Erz亘齢蹴Reg量o盟 of two measurements．T、herefore the first Measurement was carried out twice to result is in a general agreement with the examine the reproducibility of measured data second one． by the cup method for the Mihar and Bahik At the HIIlr segment，the track density segments in July，1983and July，1984，and also shows the maximtm value at point No．10 for the H111r segnlent in July，1983． where the fault is considered to pass through． At the Mihar segment，measuring cups Another high value of track density at point were buried at the intervals of a few meters No．1where a subsidiary fault may pass from north to south crossing the E－W trend－ through．The track density at point No．9 ing fault scarp to the west of Mihar Village． shows a relatively low value because the film At and around this segment loose gravel bed surface was covered with drops of water． of presumably Pliocene to Pleistocene age is distributed． 且V。K岬瓶jk旦且・亘s騰e腹聡墨a・Ye謡ee Reg亘樋 At the Bahik segment，the measuring cups were mostly also buried with a space of a few 旦V－l Geo旦09量e我且 ＆聾胡 se童s朧o亘09童c我且 se琶四 meters from north to south crossing the fault t員聡gs trace on the banks about5meters high where In this region the surface break of Bolu－ the fault plane is obser▽able．Furthernlore，at Gerede （Gerede－gerk：e婁）Earthquake occur－ the first measurement two measuring cups red along the North Anatolian Fault Zone． were placed just on the fault plane and This earthquake with magnitude about7．40n another one measuring cup was placed on the a Richter scale occurred on l February，1944 subsidiary fault at the bottom of the ravine． in northwestem Turkey．The total length of Aromd these measuring points，weathered the earthquake fault is about160－190km． soil of colluvial origin and serpentinite frag－ The right lateral strike－slip displacement is ments are distributed． 350－360cm and the dip－slip displacement At the H111r segment，the cups were placed （north side down）is100cm（KETIN，1948； with a space of5meters from north to the AMBRAssEYs，1970；DEwEY，1976；ToKs6z south crossing the fault trace on the gentle θ勧1．，1979）． slope covered with weathered soil and ser・ In the Kapakll Region，the Black Sea Plate pentinite fragments． （a part of the Eurasian Plate）on the northern Results in Erzincan region are shown in side of this fault，is underlain by Pre－Eocene Figure3． ophiolitic melange containing blocks of Per一

一103一 B％JJ6ガ％ （～プ 云h6 0εolo9♂αzJ Sz67z78＝γ ρプノψα7z， Vlol．41，ムb．3

e 3230’旺 410 ℃ Gered6～ O lO2030如5Qm

らら ら 5． ゐ’9ゐ距oン 6 6 一簸霧 婁cε．．霞 ◎ ら 血 ◎ 直 島 ゐ “ ら 幽 “ 直 、鎚P々e声 島 ・ム％」6 6◎ 購kll ㈱菅P，．ち幽ら A 6＾ ＾ ，ハ c 量5me費 Po，q ’e．ゴ7．落’ ム ロびノ あ ハ《αin奮enαn‘e ※1▲※ムムc B。c △＾＾（（ S奮dion

¶．A．r． ＾ 。 凸∴ム∴6 ら ム ム ム ム あ あ ハ ヂ 糧ム＾轟ム6凸ム※1122※◎・1合＾＾1ム1※1※1＾ ’が C14 誇

2＾繰2※2※急※2※：ム轡 ▲ 4』 A ム 6 A A A ム る ム 4 ♂ 7 T．D． 20 凹1凹2 羅3囲4匿璽5－6柳 B T、D・ c

100 T．D． A 100 ’

製 o S 1 2 3 4 5 6 7 8 9 10 11 12 13 14 O 2 3 4 6 7 8 晶 曲 曲 Fg．4 Simplified geologic map of the Kapakl1－lsmetpa婁a region with results of the alpha track measurements．1：（Jurassic一）Gretaceous rocks（c），2：Eocene rocks（e），3：late Miocene sedimentary rocks，4：late Miocenevolcanicrocks，5：Quatemarysediments，6：faults，7： railway．Dashed line indicates the expected tendency though the alpha track density value of that measuring Point is lacked．

mian to Jurassic limestone，w’hich are over－ Miocene volcanic rocks．As Holocene 1ain by Eocene turbidite and Miocene con－ deposits are extensively （玉istributed around glomerate，and Miocenevolcanicrocks．Inthe Ismetpaga，physiographic features ofthe fault Anatolian Plate on the southem side of the are not clear．On the southern side Jurassic to fault，there occur Cretaceous turbidite，Juras－ early Cretaceous massive limestone is dis－ sic to early Cretaceous massive limestone and tributed． late Miocene volcanic rocks． Yenice to the southwest of Bolu is situated In the Ismetpa婁a Region10km east of near the westem end of the1944Bolu－Gerede ：Kapak11，the creep movement by the North Earthquake Fault．The Bolu Basin，one of the Anatolian Fault Zone is well known （AM－ intramountain basins along the North BRAssEYs，1970）．The average slip rate of Anatolian Fault Zone，measures25km in right lateral sense is estimated at about l cm／ length，8km in width，and is700m to800m year based on the instrumental data since high above sea leve1．The North Anatolian 1969（AYTuN，1980）．On the northem side of Fault Zone extends along the westem margin the fault zone Eocene turbidite rests on the of the basin．At Loc．1（Fig．5），an isolated late Cretaceous ophiolitic melange．At a few small hill trending E－W，which is related to kilometers north of the fault zone，this the active faulting of the North Anatolian Eocene sedimentary rocks are covered with FaultZonerises790mabovesealevelinthe

一104一 乃φ加鱈勧郷％擁翻3カ7加」勧朋o励θ魏ル吻（翫’06’の 團1 麗・ 皿3 騒・ 騒・ 圏6

L－oc．2 ．￥eηlce・ 圏7 唱¶llm一 □8

Fig。5 Simplified geologic map of the Bolu Basin and its adjoining areas．1 Gneiss，mica schist and amphibolite of mknown age，2： Devonian rocks，3：Jurassic to UpPer Cretaceous sedimentary rocks，4：Upper Cretaceous flysh，5：Eocene flysh，6：Quatemary tuff，agglomerate and andesite lava，7： Holocene fan deposits，8：Holocene deposits except fan deposits．

basin．The northem slope of this hill dips40 few meters．This suggests the accumulated －50。N and the southern slope dips10－20。S． displacement of vertical components of the This hill is composed of andesite lava and active faulting related to the North Anatolian andesitic tuff breccia of presumably Pliocene Fault Zone including that of the earthquake to Pleistocene age，which are mconformably fault in1944． covered with Quatemary fluvial conglomer－ 垣V－2 A恥ha tr我ck朧e我s腿re㎜e盟ts臨 t翫e ate．The earthquake fault in1944extends K理蹴k旦且一亙s朧e腹職§a－Ye醜萱ce Reg董o醜 alongthenorthem foot inthe westempart of In the：Kapakll Region a small ridge is the hi11，but it extends along the trough－1ike displaced right laterally by the active faulting landform between two peaks of the hill in the of the North Anatolian Fault Zone．The eastempart．Av皿agertoldthatthevertical measuring line crossed this fault trace near displacement of the earthquake fault is about the creepmeter settled by MTA，and the 50cm though the horizontal one is mknown． measurement was carried out inAugust，1983． Near Loc．2（Fig．5）the fault displaced the The track density is maximtm at measuring Quatemary travertine right laterally．At Loc． point No．3where the fault passes through， 2near Yenice，a shallow trough caused by the though the value is less than100as shown in earthquake fault in1944extends about E－W． Figure4A． The groundwater springs out and is stagnant At Ismetpa婁a the alpha track measurement in some places of the trough．On the southem was carried out on two measuring lines along side of the trough，10ng and slender mounts the masonry walls of the railway station composed of Cretaceous limestone trending E displaced by the creep movement of the fault． －W are aligned，and their height is about a The values of the track density were rather

一105一 Bz4」ε癒％qプ焼εG60409寛αJ Sz57∂εy oゾノ4夕α％，Viol，4ヱ，No．3

Y．ζ）． τ．D． A 400 400 B

300

200 200

10

ひ 0 層茗1’2 1 10 藩 20 78 9で0 層986 123456789 醗87

Fig。6 Results of the alpha track measurements in Yenice（Loc．2in Figure5）．A： cup method for the l944earthquake fault（KATo，1986），B：pipe method for the1944earthquake fault （Locality is just on the fault trace），

10wer than expected，and a clear peak of the magnitude is estimated at about6．42based on track density was not observed． the strong motion record（YILMAz，1984）． In the Yenice Region（Loc．2in Figure5） Eastern Turkey including the epicenter area two kinds of alpha track measurement were of this earthquake is defined as the triple carried out．The result by the cup method in junction of plates by the North Anatolian and July，1985is shown by the average of neigh・ East Anatolian Faults．This area is under the bouring three values in Figure6A．The track N－S trending compression regime due to the density shows relatively higher than those of northward movement of the Arabian Plate． the Kapakh and Ismetpa鼻a Regions．Though The main characteristics of the deformation the maximumvalue ofthetrack densityis346 resulted from the N－S trending compression at the measuring point No．15where the fαult are as follows． trace mns，original measured data show （1）E－W striking reverse or thrust faults，（2） broad range of high values near the fault NNE－SSW or NE－SW striking left lateral trace and therefore，unique peak of the track strike－slip faults，（3）NW－SE striking right density cannotbedistinguished．Themeasure・ lateral strike－slip faults，and（4）N－S trending ment by the pipe method was carried out extension fractures or normal faults． from July，1986to August，1987．The pipe was In the Horasan－Narman Region discontinu－ set just on the fault trace at Loc．2in Figure ous surface break zone of approximately12 5．Any change indicative of the earthquake km in length and4km in width was detected was not observed．It is noted that two peaks after the earthquake． of alpha track density are observed as shown The assembly of surface breaks occurred in Figure6B． mostly within the ophiolitic melange produced a wide shear zone，not a contimous V。麗o膿s踊一N麗朧我蝿Reg畳繍 single fault．Most ofsurface breaks are domi－ nant in the left lateral motion with the verti－ v－l Ge（》璽ogi《協皇 蹴岨 se魯凱o旦ogie魏 set国 cal component of east side down．At just 色亘聡gs north of Cimenli the left lateral displacement A destructive earthquake named the Hor・ of25cm was measured．Less developed right asan－Narman （or Erzurum－Ka．rs） Earth・ lateral segments of the fault produced a con・ quake occurred on 30 0ctober，1983 in the jugate pair with the left lateral ones．N－S Erzurum Province，northeastem Turkey．The striking tension cracks develop to the north

一106一 ／1々）hα ’名とzo為 ηz6αsz6名召彫6n云sノわ7カz6」応 1n％oπh67n ル猟海ツ （κα孟0 6∫6z♂．）

イ〃・一跨1噸禦，．

o駐r翻醐朧 42。一

100 一．D．

90

80

7（〉

畠 60

T．D． 50 50． A 臨 40

／ 30 ！

20 ￥ ￥ F 10 F 遍 0 0 1 2 3456 Fig．7 Map showing active faults and localities ofmeasuring points123456 by the cup method in the Horasan－Narman Region（modified from BARKAαα」．，1983），and results of the alpha track measurements．Dashed line indicates the expected tendency though the alpha track density value of that measuring point is lacked．

of Klzlarkale． of1，500minlengthoccurredanddisplaceda The pattem of these various surface breaks small road．The left lateral displacement is indicates the N－S trending compression cor・ 100cm and the vertical displacement of east responding to the compression axis estimated side down is60cm．On the fault plane20－300 from the focal mechanism of this earthquake NE pitch of slickenside lineation was obser－ （BARKA6！41，．，1983）． ved。The measuring line for the cup method V－2 A恥地a t聡ek朧e薦駅e㎜e魏総睡蝕e crossed this break．The film at point No．4 Ho聡s蹴一N躍㎜魏Reg亘艦 just on the fault plane was lost and that of The measurement was carried out from point No．5was dug out to exposed by some・ July to August，1984at following two loca1－ one． ities． At Loc．B（Fig．7）the fault shows the en AtLoc．A（Fig．7）theclearsurfacebreak echelon pattem composed of tension cracks

一107一 βz4」6あ館 （ゾ 孟h6 060乙0910αl Sz67z76：y qプノ41》α多zン Vbl．41，〈ηo．3

AR〈《UTLU PENINSULA we5↑empα㎡ cen曹rdpqr曹 eqs穿ern pαr曾 LAKE ADAPAZARl午 …這i≡≡iMARMAR．《．SEA…養、 憺A面鷲「『一一＿晶 、／ マ、、 づノN●A’E

ロ ぎ 、 ／一。ぐ∠藷’ 亀 厩 ク Porηukovσ 墨撃藷 イ 認 Ploin ヂ鈴’ 一薩舜藤謬響 ，’ 夢鐸ヂ

塵 翻躍翻罐曄囲睡麟 夢ダづ夢夢 』昌km 慶塞翻醗甕翻匿醒甕圏 露BURSA

夢 Fig．8 Simplified structural map of the Armutlu Peninsula Region（modified from ERENDIL6オσ1．，1989）．Thick dashed lines show the North Anatolian Fault and its branches．A脚hite arrow shows the measuring locality of alpha track measurements at Mekece。

and pressure ridges．The total left lateral 1967． displacement is about100cm．The measuring In the Armutlu Peninsula，Miocene：Lime－ line for the cup method crossed one of thes今 stone and Pliocene to Pleistocene formations cracks．The apparent peak of the alpha track consisting of conglomerate and sandstone of density is situated on the fault segment． continental origin overlie unconformably pre－ Tertiary rocks．Holocene deposits such as fan V亘．Mekece Reg量o聡 deposits，riverbed deposits and1＆ke deposits cover extensively this area and veil the active v夏一l Geo且09董暇L且 窺鵬d se豊s凱o盈09亘c我l set困 faulting of the branch（KATo，1989）． t睡gs V亘一2 A且麟我tra¢k朧e我s偲e朧e薦s睡曲e The North Anatolian Fault Zone seems to Me藍e¢eReg套艦 branch away into a few strands indicated by a Measurement by the cup method was car－ series of depressions around Marmara Sea in ried out from July to August，1985between northwestem Turkey．The direct extension of Iznik and Gey▽e as shown in Figure8．Cups the North Anatolian Fault Zone is geomor－ were buried along a narrow unpaved path phologically defined by ＆ 1inear depression striking about：N－S across the most active passing through Izmit Bay and Sapanca Lake branch of the North Anatolian Fault Zone at the northem margin of the A㎜utlu Penin・ mentioned above．This measuring spot is sula．But the most active branch is considered situated a few hundreds meters east of a to be a pronounced linear depression passing pond．Measurement by the pipe method was through Gemlik Bay，Iznik Lake and the carried out from July in1986to August in Pamukova Plain at the southem margin of 1987．Films were replaced and retrieved about the Armutlu Peninsula．This branch shows every month to detect alpha track density． right lateral strike－slip sense with someverti－ The result by the cup method is shown by the cal component。The area around this branch average of neighbouring three values in Fig－ is known to have been a seismic gap since ure9（A）．It is noted that two peaks of alpha

一108一 鋤h召’名罐吻郷膨彫燃力7加傭初％oπh醐ル吻（κ4云06’の

T．D． 1500）o 8

7．D． 1000 10000

500 50

1 10 20 留 30 78 9署0 1署 12 1986 Fig．9 Results of the alpha track measurements at Mekece．A：cup method for one branch of the North Anatolian Fault（KATo，1986），B：pipe method for one branch of the North Anatolian Fault（Locality is near the measuring point No．2by the cup method）．

track density，at which two small fault seg－ Eocene basin．The Ayancik Reverse Fault ments of the active branch of the North appeared during the Eocene．The Balifakl Anatolian Fault Zone，are considere（1to pass Reverse Fault making the southem margin of through，are found，and that all the values of Neogene－Quatemary basin，came to exis－ track density are extremely high，especially tence at the end of Eocene to Oligocene time． at point No．2（Fig．9A）．The result by the This fault is covered by（1ate）Miocene sedi－ pipe method is shown in Figure9（B）．Any ments in the eastem part of the Sinop Penin・ change related to the earthquake was not sula．Therefore，the activity of this fault observed．It is noted that values of alpha ceased during the early to middle Miocene or track density are generally higher t血an t上ose earlier．This northward migration of reverse at other points，and that the maximum value faults probably continued throughout the 、 ammounts to about1，500（T．D）． Miocene and created the Pontic Escarpment all along the southem Black Sea Coast．The ▽朋。 S亘恥o亘》Reg亘o恥 Kurtkuyusu Fault is an alleged fault or only a lineament depicted from aerial photographs． v販一l Geo且09且c翻 a脳丑 se豊s醗o且09量e翻 se色四 The Sinop II Fault is a reverse fault striking ti醜gs WNW－ESE，dipping to the north，which Approximately E－W trending reverse appearedinlateMiocene．Aftertheerosionof faults develop in the Sinop Region and show the hanging wall of this fault in Pliocene to an clear pattem of the northward migration early Pleistocene time，marine terracedeposit （Fig．10）．The Erikli Reverse Fault ac－ and sand eolian deposits covered the base－ companied with an overtumed anticline， ment rocks in late Quatemary．The Balifak1， which extends over100km in the Sinop Kurtkuyusu and Sinop II Faults distributed Region，borders the southem margin of within or adjacent to the agglomerate and

一109一 β毎JJ召差勿z（～プ孟h召 060Jo9づαzl Sz67∂6ッ（ゾノψごz銘， V＝oJ。41，．八b．3

Kur曾kuyusuFαul骨 駐LACK SEA 竈8翫 Singpl繍 馳 OP 42。N u5曾α㍗盟 尾剛昌 2罐隙。瀞醤昏▽誌8縮飾好， Fα謝

冶軍 ξrik牌裾篭Gerze 幽 35・旺

TD．

層脇 舳 鶴⑪ 晦 幽

5⑰

T．肛）． ⑪ 層 5㊨7鯵⑭層⑰層鋼咽2 咽3

2c 5⑪⑪ T．D．

咽 5 7

層 2 3 4 5 4⑪⑪ 3

T．D． T．D． 30⑰ 3⑭⑪ 3◎⑪

囁c

2⑪⑪ 2⑪⑪ 2⑰⑪ 鞭

竈b

恥⑱ 働⑪ 働⑪ 騙 融

⑪ ⑪ ⑪ 眉3騒7謬9瑠⑫ 租一5’7 帽234567翻鰯眉⑪ Fig．10 Simplified stmctural map of the Sinop Region，showing localities（1a，1b，2a， 2b，2c，3，4a，4b，5a，5b，6a，6b，and6c）of alpha track measurements and results by the cup method．

basaltic dikes of late Cretaceous age．All active at preSent， faults mentioned above have no geological In the Sinop Region there is no record of and physiographic evidence that they are historical or instrumental large earthquakes

一110一 助hα磁・為％硲膨耀n云sカ7伽1醜競07∫h醐丁％吻（猷・伽」・） although small and moderate earthquakes of the hanging wall of the reverse fault is occurred occasionally in the Pontide Escarp・ much fractured．Therefore it can be recog・ ment． nized that the track density increased gener－ 鞭一2A旦曲at聡e匙朧e訊麗eme臨置館翫e ally from the fault line which passes through s董恥q》Regio盟 themargin oftheh皿to thetop ofthehanging Measurements by the cup method were wal1，that is，the buldge．The second result in done at the localities la，1b，2a，2b，2c，3，4a， 1984is similar to the first one． 4b，5a，5b，6a，6b and6c insummerto aut㎜n， At the Bahik segment，one of the1939 1983 （Figure10）． earthquake fault segments，the result of the The results of the Erikli（1a and lb），the first measurement in1983shows two peaks of Ayancik（2a，2b and2c），one profile on the alpha track density，that is at points No．6 Balifak1（4a），Sinop I（5a）and Sinop II Fault where the main fault trace passes through （5b）are rather lower than expected．The and at point No．9where a subsidiary fault is Ustabumu Fault which is the westem contim・ situated．In another measuring lust on the ity of the Balifakl Fault showed the highest main fault trace in the dried－upstream where track density during the present sur▽ey．On is a few meters east，the track density is the southem side of this fault altemating nearly the same as that at No．6．The result of sandstone and mudstone are distributed and the second measurement in1984is also simi－ on the northem side limestone is distributed． 1ar to the first one as well as at the Mihar Both of them are of late Eocene age．The segment．These results strongly support the Balifakl Fault at locality4b also shows high good reproducibility of the alpha track mea－ track（1ensity． surement． At the H111r segment，the track density 懸．D五鵬ssio囎 shows the maximum value at point No．10 ，where the fault may pass through．The track At the Mihar segment，one of the1939 density at the point No．9is relatively low earthquake fault segments，the reason why because the film surface was completely cov－ the track density of the measuring points No． ered with drops of water when the film was 9and10near the fault trace was fairly low in retrieved．Although the high value of track the first measurement in1983，may be ex－ density at point No．1could not be well inter－ plained that cups were placed in the small preted， a subsidiary fault probably passes artificial troughs for irrigation．Because the through at that point． diffusion of radon to the ground surface is At Ismetpa婁a where the North Anatolian often disturbed if the surface soil were culti－ Fault Zone has a creep movement，the result vated within a few years before the measure－ did not show meaningful values．This is ment．In general，it seems that the track because measuring conditions were wrong． density increases from the north to the south That is，the water gushed and settled in the tmevenly，especially from point No．11to No． depression of the ground surface near 18．These measuring points were located on a masonry wall when films were retrieved， gentle slope resulting from active faulting． therefore puddles might intermpt the difus－ The southem side of the fault moved upward sion of the soil gas including radon，and the several meters for a few earthquake events． ground surface was hardended artificia11y． The Mihar seglnent seems to be a high angle Also at Kapakll showing a creep movement reverse fault，though its main displacement is of the North Anatolian Fault Zone，the result right latera1．It is well known that the buldge did not show meaningful values．MTA settled

一111一 B％lJ6∫’％ （ゾ 云h6 （3セoJo9づαzl S鋸7∂6：y げノψごz7z， 『Vlol．4ヱ，ノ〉io．3 a creepmeter at this point，but no creep of the Ustabumu Fault（3in Figure10） mevement of the fault has been recorded． resulted from that agglomerates in the depth At Yenice the peak of track density by the may well be the source of high radon as well cup method is rather broad but the maximum as rock type and that the water supply value in346at point No．15and is higher than through the fault may be excessive．Note this those at other places mentioned above．It is measuring place is very close to the sea coast because the fractured zone of this segment facing Black Sea．Also the high track density might be broad here．Limestone around here of the Balifakl Fault（4b in Figure10）was is crushed along the fault segment by brittle taken place near a soda spring．Although the deformation．The underground water through Kurtkuyusu Fault was recognized by the this crushed zone can supPly soil gas and expression of a lineament and a relative enough to make possible the higher concen－ bench，no geological evidence has been found． tration of radon．The result by the pipe The basement rock is agglomerates which method show the clear relationbetweentrack occurs only within a few tens of meters below density and temperature，as the track density the ground．However this profile gives high is lower in winter and higher in summer． track density，it should be noted that there is In the Horasan－Narman region，the a small pond near the measuring place． measuring at Loc．Awasmfortunatelydisturb・ ed．Therefore we can not say anything about 夏x．Co脆c且櫨樋 it』At Loc．B，the apparant peak of track density is situated just on the fault segnlent． The alpha track measurement by the cup The peak value is not so high as expected method is capable of detecting not only the although these measurements were carried active f段ult including the earthquake fault but out within a year after the earthquake occur・ the geological fault even in the region where rence． rocks containing a little radon are distributed Inthe Mekece Region，itisnotedbythecup in Turkey．The results of the alpha track method that two peaks of track density， measurements have good reproducibility if where fault segments are considered to pass the weather condition，especiallythe tempera－ through are found，and that all values are so ture is nearly the same． high，especially at the measuring point No．2。 In the seismic gap area of one branch ofthe Even though the underground water caused North Anatolian Fault Zone，that is Mekece by the neighboring Pond supPlied excessive region，the maximum alpha track density soil gas，this value seems to be anomalously obtained by the cup method is abnormally high radon concentration．The result by the high although the basement rocks contain pipe method in this region also shows very little radon．Furthermore，the results by the high concentration of radon as a whole． pipe method also show the high values of the Furthermore，it shows that the maximum alpha track density even in winter．The track density was attained in winter．This branch fault of this region is physiographica1－ region has no record of historical earthquake， 1y active and there is no historical record of but can be considered to show one ofthe most earthquakes in this area．These facts indicate active features of the North Anatolian Fault that this branch fault of the North Anatolian Zone geomorphologically．Therefore，these Fault is most active now and in near future． results may indicate the earthquake occur－ Therefore it is necessary to carry on the rence in the near future． alpha track measurement by the pipe method In Sinop Region，the high amount of radon in some places in this region．But the mder一

一112一 肋h4枷漉彫幽名㈱6廊力γ加」醜朋o励醐丁％吻（κ伽6加ム） gromd water supply must be taken into con－ （Turkey）；aspects of the westem North sideration before the evaluation of fault activ－ Anatolian Fault Zone．Report of intema－ ity， eapecially earthquake occurrence． tional Research and Development Coop－ Because not only active faults but also non－ eration，ITIT Projects No．8513，Ed． Geo1．Surv．Japan，p．59－65． active faults sometimes show high track den・ KATo，H．（1984） Earthquake faults along the sities if measuring Points are near the sea，the eastem part of the North Anatolian pond or spring． Fault（Turkey）。ノδ％名G6097．．vo1．95，P。

So as to compare the absolute values of the 77－93．（in Japanese）． track density between different faults，it is （1986） Notes on earthquakes and required that the basement and cover rocks active faults in the North Anatolia should be similar to provide a comparable （Turkey）．Chゑsh狢％〈石召卿s，no．387，p．38－ amount of radon and its diffusion along faults 53．（in Japanese）． ，BARKA，A．，KAToH，K．and Ku多gu，1． through the surface．Then it would be possible （1985） Some results of the alpha track to assess whether high track density indicates measurements on faults in northern Tur－ fault activity or not． key。Report of International Research The results by the pipe method show the and Development Cooperation，ITIT seasonal change depended on the tempera－ Project No．8212，Ed．Geo1．Surv。Japan， ture．It is better to detect the fault by the cup p．53－69． method of the alpha track measurement in （1989） Some remarks on geologic and summer generally because the diffusion ofthe tectonic features of the westem part of the North Anatolian Fault，northwestem gas containing radon through the fault plane Turkey． Report of international is sprightly．But for evaluating the fault activ－ Research and Development Cooperation， ity，it is necessary to keep measuring radon ITIT Projects No．8513，Ed．Geo1．Surv． by the pipe method in a few years． Japan，p．1－16． ，KへToH，K，KuSgu，1．and TsuKuDA，E． Re奮ere配es （1989） Alpha track measurements in the westem part of the North Anatolian AMBRAssEYs，N。N．（1970） Some characteristic Fault．Report of international Research features of the Anatolian fault Zone． and Development Cooperation，ITIT T66ホoηoρhッs∫os，vo1．9，p．143－165． Projects No．8513，Ed．Geol．Surv．Japan，

AYTuN，A．（1980） Creep measurements in the P．30－34． Ismetpa婁a region ofthe North Anatolian KETIN， 1． （1948） Ober die tektonisch－ Fault Zone．動IsIKuRA，A．M．and mechanischen Folgenmgen aus den gros・ VoGEL，A。（eds．），Multidisciplinary sen anatolischen Erdbeben des letzen apProach to earthquake prediction， Dezenniams．G601．ノ～観4soh側，voL36，p． Friedr．Vieweg and Sohn， Brauns－ 77－83． chweig／Wiesbaden，p．279－292． KING，Chi－Yu（1980） Episodic radon changes in BARKA，A．，SARoGLu，F．and GuNER，Y．（1983） subsurface soil gas along active faults 1983Horasan－Naman depremi ve bu and possible relation to earthquakes． depremin Dogu Anadolu neotektonigin－ ノ；oz6z．980ρhダs．ノ～6s．，vo1．85，3065－3078． deki yeri． y診勿yz6zノごz7げ ∂6 乃zsごz％，vo1．8， ToKAY，M．（1964）Zongludak sheet，1：500，000 p．16－21．（in Turkish）． scale geologic map of Turkey，Mineral DEwEY，」．F．（1976）Sei3micity of Northem Research and Exploration Institute of Anatolia。B％Jl．S6ゑs郷ol。z4郷．，voL66，p． Turkey． 843－868． ToKs6z，M。N．，SHAKAL，A．F．andMlcHAEL，A．」。 ERENDIL，M。，Ku婁gu，1、and KATo，H．（1989） （1979） Space－time migration of earth－ Tectonics of the Annutlu Peninsula quakes along the North Anatolian Fault

一113一 B％lJ（3∫∫％（～プ 孟h（3060Jo9∫αzJ Sz67∂6ッ げノ4ψα多z， Vb乙41，ノ〉io．3

Zone and seismic gaps．鞠60φh．．vol． participants at the International Insti． 117，p．1258－1270． tute of Seismology and Earthquake YILMuz，R．（1984） Engineering seismological Engineering，Intemational Inst．SeismoL analysis of strong ground motion records Ea耽hquake耳ng。Building Res．I Inst．， at the l983Erzumm－Kars，east Turkey， vo1．20，P．137－153． Earthquake．lndividualstudiesbythe

トルコ北部におけるアルファ・トラック測定

加藤磧一・加藤 完・アイクット・バルカ・イスマイル・クシュ

要 旨

トルコ北部を北に緩く弧を描いてほぼ東西に約1200kmにわたって発達する北アナトリア断層は右 横ずれのトランスフォーム断層で，1939年本断層東部のエルジンジャンで発生した地震以降，本断層沿 いにM7クラスの地震が相次いで発生しその震源が西方移動していることが知られている． 筆者らは，それらの代表的な地震断層である1939年エルジンジャン地震断層，1944年ゲレ・デ・ボルー 地震断層，クリープ運動を示すイスメット・パシャ付近の活断層，近い将来の地震発生が懸念される同 断層西部のメケジェ付近の地質学的調査と断層トレースの追跡及びその活動性調査のためのアルファ・ トラック法による地中のラドンの相対濃度測定を実施してきた．また，比較のため，北アナトリア断層 と関連しないが，最も近年トルコ東部で発生した1983年ホラサンーナルマン地震断層及び黒海沿岸部の シノップ地域に発達する地質断層である東西性の逆断層群においても同様な調査を行った． アルファ・トラック法は断層トレースに簡便なカップ法と継続測定に有効なパイプ法の二通りを使い 分けた．前者は，上底部にアルファ線に鋭敏な硝酸セルロースフィルムを添付したプラスチックカップ を想定される断層線に直交する測線上に数mおきに地下浅く埋設し，1週間から10日程度経過した後取 り出したフィルムを化学処理して単位面積・単位時問あたりのトラック数を測定しその地点のラドンの 相対濃度とするものである．．後者は，径6cm，長さ60cm程度のプラスチックパイプを断層トレース 上に埋め込み，上端を塞いだゴム栓にとめた糸でフィルムを吊るして，前者と同様に測定し，長期にわ たって繰り返すものである． その結果，調査地域のようなラドン含有濃度の低い岩石の分布地域においても，断層の精密トレース に，カップ法が有効であり，しかも測定結果の再現性も良いこと，ラドンの相対濃度値は，測定点付近 の地下水などの賦存状況に影響され，非活動性の地質断層でも高い値を示す事，パイプ法によって地震 空白域であるメケジェ付近では地下水などの賦存状況や地温変化などを考慮しても異常に高いラドンの 相対濃度値が連続してえられ，このことは，同地域の活断層の活動度が極めて高いことを示唆している ことが明らかとなった。

（受付：1989年6月12日；受理：1989年10月26日）

一114一