CIRCUM-PACIFIC FAULTING IN THE PHILIPPINES TAIWAN REGION 1
CLARE;.JCE R. ALLE!\ Dfrisio11 oj Geological Scie11ces Caliiornia /11stit11te of Technology, Pasadena
ABSTRACT
Conflicting views of circum-Pacific Tectonics have focused on the Philippines Taiwan region, where there has been neither convincing documentation nor general agreement on the importance of transcurrent (strike-sHp) faulting or the possible ~ense of regional horizontal displacement. Structural and physiographic features of the 1200-km-long Philippine fault zone are fully as spectacular as those of the better-known San Andreas and Alpine faults, and current activity is indicated by n!any localities in which scarps cut Recent gravels. Predominance of horizontal over vertical displacements is indicated by linearity of the fault trace, failure of one side to be consistently higher than the other, disregard for gross physiography, and scis soring of individual scarps within the zone. Consistent stream offsets on Luzon, :.Iasbate, anti Leyte demonstrate unequirncally that the sense of Recent displace ment has been uniformly sinistral (le~t -h anded). The Philippine fault has no obvious geologic 1·elationship to active volcanoes, but the parallelism and proximity of the !ault to the Mindanao trench suggest a close causal relationship. The remarkable Longitudinal \"alley of eastern Taiwan represents another great transcurrent fault parallel to the western Pacific rim, and ground displacements during historic earth quakes indicate a sinistral sense of displacement here as well as in the adjacent Phil ippines. This st udy does not support the hypothesis of counterclockwise rotation of the Pacific basin, but more important is the further documentation of the predo minance o f transcurrent faulting in active circum-Pacific orogenic areas. These re sults reinfo1 ce field studies in Alaska, California, Chile, and 1\ew Zealand, as well as emphasizing the geological reasonability of the results of seismic fault-plane so lutions indicating the world-wide predominance of transcurrent movements.
THE PROBLEM widely divergent \'iews of circum Pacific orogeny focus here, especial The purpose of this study was to ly concerning the question of the investigate some of the major faults relative importance of horizontal uf the Philippines-Taiwan region fault displacements. Even grant with the hope of unequivocally de ing the importance of horizontal monstrating their extent, their de movements, major disagreement has gree of current activity, and - par existed concerning the sense of pos ticularly their sense of displacement. sible displacement: theories of Be- Unusual geophysical significance is 11ioff (1959) and St. Arm and attached to this region because ( 1959) implied the probable exist-
1 Reprinted from Journal of Geophysical Research, vol. 67, November 1962, no. 12, pp. 4794-4812.
122 123 CLARE:\CE R. A I.LEK ence of dextral or right-handed in the southern Philippines, the faults paralleling the western Paci southwesterly trend of the islands fic margin in this region, whereas in the southwestern Philippine con theories of \-ening ::V1einezz (195.J) trasting sharply \vith the southeast and Big (1960) supposed sinistral erly trend of the islands and moun or left-handed displacements in thif: tain chains along the southeastern s~:me r egion. In both hypotheses. rim of the ar chipelago. The island che assumed sense of displacement of Masbate,1 with its two di,·erging played a part in the interpretation 'arms', has often been cited as an of the general mechanism of cir example of the acute junction of cum-Pacific orogeny. Yan Bemme the two opposing trends. Willis was len ( 19.JO) and others ha Ye tended so impressed by the linearity of the to discount the importance of hori southeasterly trending features be zontal fault displacements in this tween Ragay Gulf and northern region altogether. Mindanao that he proposed the ex istence of the Philippine fault zone \Villis (1937) was apparently the along this line; his e\·idence was first inYestigator to recognize the admittedly based almost wholly on existence of a through-going fault gross physiographic forms. Other zone in the Philippines comparable investigators who followed Willis to the San Andreas fault in Cali have generally agreed upon the pr e fornia, although numerous workers sence of a fault zone, although there dating back to rnn Drasche (1878) have been disagreements concer n had pointed out the existence of ing its possible extension into north various tectonic lineaments in the ern Luzon and its exact location in archipelago, mainly on the basis of the central part of the islands. Evi the alleged alignment of islands. dence for its existence has continued Yolcanoes, and earthquake epicen to be based mainly on gross physio ters. Repetti (1935), in particular, graphy, inasmuch as there is no had proposed the "Master fault'' clear documentation of displace along the general trend of the Phil ments along its trace during historic ippine land zone, but as evidence earthquakes. Only on the island of other than that from northern Lu Mindanao do published geologic zon was based almost wholly on in maps clearly demonstrate and deli conclusive seismic data. The ter m neate the Yarious breaks that make 'Philippine fault zone' used by Wil up the Philippine fault zone (Ran lis (1937) seems to haYe become neft, et al., 1960). Probably the well-established in the literature, most accurate maps of the fault's notwithstanding the later use by location north of Mindanao, and Willis (1944) of 'Visayan rift' for the most cogent arguments for its the same featur e. Von Drasche existence or those by Irving (1951, ( 1878, p. 3) and others had partic 1953), based in part on an analysis ularly pointed out the bifurcation of the submarine morphology of the or virgation of the tectonic trends region.
1 All Philippine place names used in the text are indicated on Figures 1 and 2. CIRCU)l-PACIFIC FA ULTJ:>JG 12.t
\\'illis apparently considered the in Miocene time, as does the general Philippine fault to be dominantly trend Oi the folded mountain ranges transcurrent (i.e., 'strike-slip,' 'la north of Manila. Wisser (1952) cit teral". or 'wrench') because of its ed supporting evidence from local similarity to the San Andreas fault. fracture patterns in southeastern particularly in its length, linearity, Luzon and northeastern Mindanao. and general physiographic expres Such east-\vest compression is, of sion. But the sense of horizontal course, compatible with sinistral displacement has been the subject transcurrent displacement on the of more fabulous argument. Willis Philippine fault zone but by no argued that the fauit was sinistral, :11eans proves it. Kor has there and the most often-quoted evidence been complete agreement on the re is based on a single set of horizon gional stress pattern; the numerous tal asymetric slickensides that were north-south lineaments of the cen observed by Willis (1937, p. 39) in tral Philippines were visualized by the Aroroy mining district on Mas Alcaraz ( 1947) as tensional in ori bate, although this particular fault gin, which implies a regional stress is subsidiary to the main zone and system that would lead to dextral is apparently at an angle at least c.!isplacement along the Philippine 30° to it. Willis further argued that fault zone. Ranneft, et al. (1960) the regional plan of the islands sug indicated sinistral displacements on gested a sinistral sense of displace individual breaks of the zone in ment. Other authors agreed (Alvir, noi-theastern Mindanao but did not 19-ll; King and McKee, 1949). But ::;tate why. opinion has not been unanimous; Irving ( 1951, p. 70) thought that lt might be hoped that results of many of the offsets implied by such seismic faultplane solutions on the arguments were 'excessivem' and numerous earthquakes of the Phil Alvir ( 1926, p. -153) had used the ippines region would clear up the same kind of argument (based on ambiguity resulting from the limit gross physiographic features) to in ed geologic studies. In a summary dicate dextral displacement along of 35 fault-plane solutions from the fault in northern Luzon that is this region, Ritsema and Veldkamp herein considered to be an integral (1960) state that the predominant part of the Philippine fault zone. displacements have indeed been transcurrent and sinistral, and the Another line of evidence bearing implied direction of horizontal com on the sense of fault displacement is pression is about N60°E . None of that given by fracture patterns in the larger earthquakes of this group, Philippine mining areas, although however, occurred within the Phil none of the areas described in de ippine fault zone, and the largest tail has been within the fault zone shock (M = 8.2, Jan. 24, 1948) was proper. Leith (1938), Livingston clearly dextral. The epicenter of (1939), and Wisser (1939) pointed this great earthquake was near the out that the pervasive shear and west coast of Panay, where there tension joint of compression, at least is good evidence of a major north- 125 ('LARE.'.'ICE R. ALLE:\
east-trending fault that might rea relatiYely rec:ent displacements i~ sonably be considered conjugate to confined in many areas to a single the Philippine fault and thus dex clearly defined zone. and it is thi,: tral. The fault-plane solution, how most recent or acti,·e break that i:::: eYer, which is supported by the primarily shown in Figures 1 and ~. trend of the isoseismals (Irving and EYidence fo1· the fault':> location i:' Te,·es, 1948), indicated failure on a based mainly on physiographic fea fault striking more nearly parallel tures of Quaternary displacement to the Philippine fault than to the that are clearly \'isible on aerial local structure. Furthermore. Aki's photographs and in the field. These (1 9 6 0) source-function solutions features include fault and faultline based on Rayleigh "·aves continue scarps, fault troughs and valleys, to indicate dextral displacements side-hill ridges, and numerous fault around the entire circum-Pacific sags and sag ponds. In general, de margin. The one Philippine earth tailed features of active faulting are quake included in Aki's solutions almost as \Yell displayed along the indicates breaking on a fault sub Philippine fault as along the San parallel to the Philippine fault in Andreas fault in California. despite the northern Central Valley of Lu the extreme differences in climate. zon, and it is clearly dextral.
Thus, in brief, it was not clear at P r obaLly the best-displayed seg the beginning of this study that the ment of the Philippine fault zone i:-; in northwestern Leyte, between the Philippine fault zone ( 1) indeed ex isted, at least as a throughgoing fea towns of Yisares and Leyte. The ture of regional tectonic importance, zone here consists of a number of (2) was active, and ( 3) was neces parallel acti,·e breaks O\'er a width sarily a transcurrent and sinistral of about 500 meters, and the ex feature. Furthermore, it was hoped treme linearity of the topography that if the Philippine fault zone within the fault zone stands in was indeed a feature representing marked contrast to the topography regional horizontal shearing, then on both sides (Figures 3-5). Neither similar features might be obsened side is consistently higher than thf' in Taiwan or other adjacent circum other, and the central part of the Pacific areas. zone is, in general, a structurally and topographically d e p r e s s e cl trough that has caused the term PHILIPPI~E FAULT ZONE 'rift' to be applied to this fault in Location and ph ysiograpll ic ex analogy with the similar features JJl'ession. - Se\'eral in\'estigator" of nomenclature of the San Andreas haYe pointed out that the Philippine fault. Indeed, the general physio fault zone encompasses a wide re graphy is best described in the same gion interlacing and branching frac· words used by Noble (1954, p. 37) tures, so that its exact location i~ in portraying the San Andreas: 'It hard to define. This is probably has curiously direct course across true, but certainly the e\:idence of mountains and plains. with little re- ('IH<'C:\l-PA ClFl(' FACLTI~\:
120•
_J 1a·- cl: ~g2 M1ndonao ''tnclr Wts l ,Caro,,,,,,,.....-,/int ( z ,I f'olou lrt nch 0 100 200 FIG. 1 - - :\lap of northern Philippines. showing place names mentioned in the text and t races of aeti\·e bl'eaks of P hilippine fault zone. <'T.A REKCE R. ALLEN \ 120·200 300 .'Tl I ------< ___J F . .-:. :2 • 1 n Philipprnes.. . s IJO \\_.ng 1 place names ment10n· ed in the text :.Hapand nac:eso1 ~om of 1e1· aet1ve . b1·ea k's o f Philippine fault zone. Cl RC L\1-PACIFIC FAU LT!.'\(; 128 ;,; ::,, .. ;:: > 2 .. _,; c, .... - s >. -< "' ~"' :; ..:: .,- 0.: c., 0 -< ~ " '"' ...; d Ci. CLAREl\CE R. ALLEX Fir:. ·>. Aerial Yiew of features of recent displacement. Philippine fault zone, northern Le>·:e. FTC G - Aerial Yiew of recent scarps of Philippine fault zone east from near Bitulok, Luzon, toward Dingolan Bay (background). Photograph taken f1 om the help of ~I. R. Merrill. CIRCC:\l-PAIIFIC F ..\ n.TT:\C l:JI) garcl for gross physiographic fea the Agusan Hi,·er \"alley itself, the tures, yet it influences profoundly surficial rocks are :so young that it the local topographic and geologic i:;; not surprising that the fault zone features \Yithin it.' there lacks obYious physiographic express:on. Both north and south of the cen tral Philippines area. the fault se Current adi\'ity of the fault i~ parates into a number of branches. demonstrated in area~ in which the The fault that enters northern Lu fault clearly break:; Recent stream zo n at Dingalan Bay is herein con graYels. Prominent examples are: sidered to be the main Philippine (1) The fan of the Agno RiYer in fault on the basis of its similar northern Luzon is cut by a distinct strike, degree of complexity. and scarp that p:uses through the north sense of displacement as com pa reel ern part of the t Further eYidence of Quaternary nary Yolcanic rocks in the area. lr cli,;placement,; i::. giwn hy closed cle Ying ( 1951) has discussed the sub prel'sions along the fault, \\·hich cer marine morphology along the Phil tainly cannot sui·yive long in a re ippine fault. and his map show:-: gion \\·hich not only has an annual elongate closed depression along the rainfall exceeding 3 meters along extended trace of the fault between much of the fault trace but can al;;;o Alabat Island and Dingalan Bay, in claim some of the world's greatest Ragay Gulf. between Ticao a n cl short - period rainfrll,; associated Masbate, between Leyte and north with frequent t,vphoons. Probably eastern Mindanao. and south of the largest of t~e.c:;e depressions 1~ Mindanao. occupied b~- Lake Mainit in north eastern )1indanao. Lake Danao, 12 ~e11st of d1splace111c11t. - The km. northeast of Ormoc, Leyte. i,; known juxtaposition of grossly dif dearly fault controlled. as are seY ferent rock types along parts of the eral other dry closed depressions Philippine fault zone suggests t hat nearby. Lake Danao's location at the displacement has been apprecia the acute junction of two closely ble. and the existence of fault scarps diYergent branches of the fault (Fi in areas of lo\\' r elief indicates that gure 2. inset) appears to be a good at least some of the component of example of the mechanism of grab displacement has been vertical. But en formation by transcurrent mo\'e se,·eral lines of evidence suggest men t proposed by Len sen ( 1958). that the o\·er-all displacement dur •.\ sag pond bordered by fault scarp:-: ing the fault's history has been i:' illustrated by the small pond ~./ primaril~· horizontal. km. i"Outheast of Bitulok. Luzon, near the crest of the fault saddle ( 1) The c x tr em e linearity, or between Dingalan Bay and the Cen smoothness of curvature. of the tral \'alley. In addition, a multi fault trace oYer hundreds of kilo tude of ponds and swampy area::: meters is difficult to reconcile \vith along the fault in rice-paddy areas anything but horizontal displace of Lerte. Masbat2. and Luzon prob ment. Similar linearity is charac ably are natural closed depressions teristic of other great transcurrent but these are hard to Yerify because faults (Figure 9) but is unknown of possible cultural modifications. on faults with predominantly ver· On Leyte. the fault cuts across the tical displacement. To the author, flank of the volcanic cone of Mount the linearity of the Philippine fault Janagdan (Figure 2. inset) \\'hich. zone is in itself a diagnostic crite although not liRled as active. must rion of dominant horizontal dis· be of Quaternary age in Yiew· of placement. the pr esened crater lake at its summit. In cenfral Leyte, the fault (2) Along- most of the trace of trace is not a~ clear as it is farther the Philippine fault, neither side is north\\"est. Thi:; is uncloubteclly re consistently higher than t he other. lated to the abundance of Quater- The principal exception is in north- CIRCl.':\1-PACIFIC FAULT!!\(~ 132 C Eosl Bronci Loyog River, Leyte F ir.. 7 - '.\fa;-i of sinistral stream offsets along Philippine ,'ault. CLAREXCE R. ALLEX ern Luzon. \\·here the mountains are Granting the predominance of ho L'0n"istently hiirh northeast of the rizontal displacement. the Rense of fault. Rut in this al·ea. individual t!isµlaccment - clextral or sinistrnl J~ecent scarps at the ba"e of the - is a much more elu:'iYe question. rang-e exhibit Yar~·ing and opposing The author cannot accept the argu Yertical cli"placement,.;. a n cl th e ment or King and 1\1cI \\·ithin it. Jn iact, most of the larg ularly inasmuch a.-; the one area e,;t earthquakes of the region during from \\·hi<:h extensi \·e cracks \\'ere th i,.; period (1907-1959) occurred reported is the only area where the ebewhere in the Philippines. One actiYe break crosses dry land in thi,; !"hould note. on the other hand. that region. De.-;c:riptions of the great the diffuse spr ead of epicenters Surigao, :.\1ind~mao, earthquake of al,.;o typifies California and ~ ew July I. 1879 (Centeno y Garcia. Zealand, \Yhere active 'master fault::;' 1882). mention extensive fissuring are \\·ell-documented features. Suc:h c:hange.< or ground le\'el between major faults as the San Andrea:-: An:ioan and Lake :.\Iainit (Lgo Sa are eYidently characterized by great pongan), whic:h is the exact line of earthquake .., at only infrequent in the most recent appearing break on tervals, a lthough the tectonic strain modern aerial photographs. One of release ot' one such great shock may the gre:1test earthquakes in Philip lJe g 1·eate1· than the combined effect pine history is th:tt of June 21, 1893. of smaller s hocks o\·er a wide region in ea:;tern ) linclanao, which was during the preceding hundred year,;. cleal'iy as.;;ociated with wide-spread Thus the truly important tectonic subsidences in the swamps of the ,.;e i:-.mic eYents may not appear on upper Agusan River and \Vith pos an epicenter map that necessarily sil.Jle extensiw faulting near the cowrs only a ,·ery limited time present town of Monkayo (Mso. nmge. 1910). Allll the largest instrumen tally rewn.led earthquake of the Philippine;:; region CM = 8.3) oc :\s was stated in the introduction. u1rred on April 14. 1924, along the no historie earthquakes along the the projected trnce of the Philippine Philippine fault a re known to be un fault south of l\Iindanao, at 6-1 2°X. equiyocally associated with actual 1:26-1 2uE. At Mati, where the most ,.;u r ficial fault displacement. But se recent break of the fault goes out \'entl historic shocks were probably to sea. large-scale subsidence and related to surface faulting, and if possible faulting were reported at the historic record were more com that time (1\fowo, 1924 ; Melendres plete the Philippine fault zone might and ('mn:-.ti. 1%1. p. -l~). well appear more acti\'e than the better-known San Andreas and Al pine faults. For example, the great 7'1 clu11ic i'1iriru11111c11t. The earthquake of August 16, 1869. was pre::ence in the Philippines region of destructi Ye throughout much of the a i:rreat oceanic trench, abundant central Philippines and was appar active volcanism, numerous serpen ently centered near Masbate. Judg tinic intrusions, and earthquakes of ing from the reports of changes in shallow to deep focus indicate that ground level near Ticao and exten this region is truly part of the cir si\'e 'cracks' in southern Masbate cum-Pacific orogenic belt, although (Maso. 1895). this shock was prob the diverse tectonic trends of the ably associated with slippage along archipelago and the lack of align the fault's central segment, partic- mf'nt of rnlcanoes indicate that the re:non is hardly typical. The gen there is little regional ,-okanism at tle u11·yed trn('es of the Philippine all (Along the San Andreas and i'ault and the :\lindanao trench il southern Alpine faults). l uslra te the ar('Ua te structure, con ,· ex to\\'ard the Pacific, and the The slight kink in the Philippine parallelism of these two features is l'ault between Dingalan Bay and remarkable. The axes of the trench Alabat Island in Luzon is interest and the fault are scarcely 150 km ing be('.ause this is the only depar apart for long distances east of ture f r om the otherwise ,·ery :.Vlindanao, and one can hardly es smooth curYe of the fault tra<:e. l r cape the condusion that, if sinistral one Yisualizes the northeast coast of displa<:emenl is dominant on the Luzon to be fault controlled. as is Philippine I" au l l, the }lindanao suggested by the deep offshore trent·h must share the same stress trough (Jr,·ing, 1951) and by the :-ystem and likewise exhibit a large morphology of the coastline in the component of .-;inistral transcurrent Yicinity of the San Ildefonso Penin displa<:ement. This conclusion i.-; sula. then it is at the junction with ,:upported by the fault-plane solu the opposing Garlock fault. Indeed, tions (Rilsema an cl Veldkamp. e,·en the angles between the fault 1960), but fundamental questions systems are similar if one is Yiewed are left unanswered as to the origin as a rotated mirror image of the cit' this and othel' trenches. other. Thus, perhaps the best e\·i dence for clextral displacement on The Philippine fault zone has no the northeast-trending system I!-\ ob\·ious direct relation to Quaterna the kink in the Philippine fault al ry Yolcanism: about as many ,-olca though the absence of seismicity noes occur on one side of the fault sugge8ts that the northeast system a!-\ on the other, and no Quaternary like the Garlock-may he currently yoJcanoes within the fault zone, ex i na<:ti Ye. cept possibly on Leyte. This ap pears to be in ('Ontrast to other FAL'LTIXC'r IX 1'AJ\\'.AX great adiYe transt11rrent faults Of the circum-Pacific region (Figure Xorlhern Luzon is far closer to !) 1 , whel'c ( l) the fault zone lies Tai\\·an (Formosa) than it is to the p;trallel to and between the trench southern islands of the Philippines, and the line of active ,-olcanoes (the yet the tectonic relationship between Atacama fault of northern Chile Luzon and Taiwan is not at all (St. Amand and Allen, 1960) and dear. Wilson (1954, p. 184) con possibly in northern New Zealand). sidered the Yolcanic islands linking (2) the fault itself is marked by a the t\\'O areas to be representatiYe series of a c t i v e Yolcanic cones of a typical circum-Pacific arc be (probably the situation in parts of tv.;een the Ryukyu arc to the north Sumatra, southern C h i 1 e (St. and the main Philippine arc to the Amand, 1961, p. 30), and possibly south (Figure 1, inset) ; Taiwan it the Fossa :M agna of Japan). or (3) self ('.Ould thus be Yisualized as a CLAREKCE R. ALLEX cap range or secondary arc at the mountains on either side, and the acute junction of the Luzon and ,-ery odd drainage pattern in which Ryuku arcs. Biq (1960) and Ho only one of the major rivers drain ( 1961) considered Taiwan to be an ing the central range manages to integral part of a primary double breaks through the eastern wall of arc.: that is convex toward the Paci the Longitudinal Valley which lies fic, with a corresponding reversal athwart the normal drainage paths. in the order of the associated tecto One could hardly hope to find a nic features. Juan (1956, 1958) more diagrammatic example of a urged that Taiwan is a coastal structurally controlled valley, and range of the Asiatic mainland its fault origin has been recognized rather than an island arc of the for many years (Omori, 1908, p. type represented by e i the r the 16-!). In addition to the gross phy Ryukyus or the Philippines. The siography, the very different rock present study adds little to this par types of the Coastal Range and the ticular problem but does indicate Central Mountain Range demand a that major transcurrent faulting fault contact along the axis of the appears to typify Taiwan as well as Longitudinal Valley (HSu, 1956), other active circum-Pacific oroge although the contact itself is every nic areas. where concealed by alluvium. The much greater elevations west of the Central Taiwan is underlain by a valley than to the east suggest a longitudinal mountain range whose signifi<:ant vertical component of crest lies somewhat east of the is displacement, and m o st authors land's center line. The eastern coast have considered the faults under line of the island is consequently ex lying the Longitudinal Valley to be tremely precipitous in places, such predominantly thrusts (Hsu, 1956, as along the famous coastal cliffs p. 59; Juan, 1958, p. 286; Biq, 1960, 2 between Suao and Hualien. But the p. 209), although Biq argues for an central part of the east coast is additional component of sinistral bordered by a coastal range sepa displacement a 1 o n g the shallow rated from the main backbone of thrust plane:-;. island by a conspicuous longitudinal \·alley between Hualien and Tai-ung that is aligned with the coastal cliffs That the Longitudinal Valley is of the north and south (Figure 8). an active fault feature is indicated Indeed, the consistent trend of the by historic great earthquakes, by valley is so striking that it is com consistently high seismicity, and monly called simply the Longitudi by fault scarps that cut Quarter nal Valley, although many other nary alluvium of the valley floor. terms have been used in the geologic Scarps within the valley are not literature. This valley, 150 km long nearly as abundant as one might and 5 to 10 km wide, is remarkable expect from the high seismicity, for its extraordinary linearity, its because acti\'e alluvial fans of negligible relief as compared to the rivers draining the Central Moun- CIRCl.1:'.I-PACIFIC FAULTING 118 tain Range literally inundate most placement along the Longitudinal of the adjacent tectonic trough. An \'alley comes from the major earth nual rainfall in higher parts of the quakes of 1951, whose geological ef range exceeds 5 meters, and the fects have been described by Hsu steepness and height of the moun ( 1955, 1962). The largest shock of tain front compare favorably with this series (M = 7.3) occurred on the great eastern escarpment of the XoYember 25 and \Vas associated Sierra X evada in California. As a "·ith \Yell documented faulting cen result, despite the otherwise sub tred near Yuli. Displacement took tropical environment, the immense. place along the fault trace for about barren alluvial fans filling the Lon 40 km from near Shuishui to near gitudinal Valley appear to be far Fuli, within and parallel to the Lon more active than most typical desert gitudinal Valley. Strike slip was fans. Along the Longitudinal Val consistently sinistral, with a maxi ley, therefore, it is only at drainage mum displacement of 163 cm near d i v i d e s or otherwise protected Shuishui, and vertical displacement points that one can hope to find \Vas intermittent, with the east Quaternary fault scarps preserved. block relatively raised as much as Among the fe\v s u c h localities, 130 cm in some localities (Hsu, pointed out to the author in the 1955) . "J1ost of the fault trace has field by T. L. Hsu and C. S. Wang, been obliterated since 1951, but are ( 1) the prominent scarp trend three offset school buildings in Yuli ing north from Hualien to the sea, that were neatly transected by the associated with sag ponds along its fault remain standing in testimony base near the Hualien airport and to the sinistral displacement at that (2 ) a distinct linear scarp at least time. An earlier shock of the same 15 km in length that passes along ::;eries on October 22 had been cen the east shore of the fault-ponded tered farther north along the Lon Ta Po Chih near Chihshang. Scarps gitudinal Valley and \Vas associated of at least two ages are present with displacement for about 7 km along the edge of the valley east of along the base of the older scarp, Fuyuan, relatively protected here north of Hualien. Sinistral dis because of their proximity to the placement was 2 meters, and the drainage divide. All the above men east block was relatively raised 1 tioned scarps trend parallel to the meter. Thus, from the available Longitudinal Valley, but they are evidence, the current activity along so scattered and so dissected that the Longitudinal Valley appears to little information about the over-all be dominantly transcurrent, as is sense of displacement can be gain ed. In no areas do incised streams indeed suggested that perhaps de appear to cross the breaks in such manded by the extreme linearity of a way as to indicate or preclude the valley. The author sees no rea lateral displacements. son to suppose that the controlling faults are thrusts, but instead visu The most powerful line of evi alizes them as essentially vertical Jence concerning the sense of dis- fractures similar to those of other c I.\ l~L.'.\1 F: R.. \LLLX 120· ~\ 2~· 0 Con1ours at 200.~.1000. 2000 m ~lap of Taiwa•1 1 Fu1nH,,;a). showing: faultin!.. clul'ing hi::turil· earthquak.,,;. . fapar1l'~t· 11ames in pare1Jtheses. CJ r:c l >I l'ACIFH' FA l.L TI~G 140 great tran:'currcnl faults where Lhe ed. cut into the ea.:;t coasl of Tai\\·an attitudes of the fault planes can be almo,;t opposite the midpoint of the more satisfactorily determined. H::-:u Longitudinal \'alley. Ukewise, the ( 1906) points out that the nearl~' Ycry smooth geanticlinal (nonrnl longitudinal faults of the Coastal canic) axis of the Ryukyu arc (Hes::;, Hange are nearly all east-dipping 19-18) \1·ould intersect Tai\rnn at thrusts, b u t the association of the north end of the valley near thrust faults \\"ith ma.ior Yertical Hualien, ~·et there is little e\·idence transcur rent faults has been •veil of an impeding major change in documented in many areas of the trend of the Longitudinal Valley \rnrlcl. The ::;ense of cli::;placement fault :::~·stem. It is difficult to es along the Longitudinal Yalley at the cape the conclusion that this fault present t ime is clearly sinistral. sy:::tem and possible parallel off \\"hich ob\'iously corresponds to the shore equiYalent~ repre::>ent an ac regional displacement along the tual truncation. and not merely Philippine fault zone farther south. sharp bending, of tectonic elements of the Ryukyu arc. l\Iore \\·idely kno\\"11 t h a n th ~~ 1951 shocks are l\rn earlier earth CO.NCLl!SlONS quakes in Taiwan that both demon strated significant components of The results of this study tend tu transcurrent. displacement of dex rein rorC'e earlier obsenations made tral sense. Geologic effects of these in northern Chile (St. Amand and shoch, in 1960 and 1935, haYe been Alle11, 1960) which indicated that :-ummarizecl from the Japane::e lite great transcurrent faults parallel to rature by Richter (1958), and the thE· oceanic margins are typiral associated breaks are sho\\'n in Fig features of actiYe cir cum-Pacific ure 8. As opposed to the 1951 earth arcs and. are not limited to uniq11e quakes, both of the earlier shocks or unusual ar eas. The best docu are on the west side of the island. mented transcurrent faults ha\'e and the part:-; of the breaks associ generally been consider ed to be the ated \Yith dominant strike slip trend ~an Andreas fault of California and tne Alpine ·fault of southern :\ew ~ G0-70 E . Thus they might •vell Zealand. yet both of these areas are !Jc Yisualized as fractures conjugate ;1clmittedly atypical of the cir cum to the ·master fault' represented b,v Pacifo: rim in that they lack inter the Longitudinal \'alley. mediate and cl e e p earthquakes, abundant Yolcanism. and oceanic The linearity and continuity of trenches. But the Philippines and. the Longitudinal \'alley fault zone particularly, northern Chile are so emphasize the apparent truncation trnly ein·um-Pacific in their struc uf tectonic elements of the Ryukyu ture that little doubt remains that ar c by Taiwan. The Ryukyu trench major trnn:=:current faults must and its southwest\\·ard extension now be included among the typical (Figure 1. inset) •rnuld. if prolong- tee ton il· !'ea tmes of the circum-Pa- l-11 CLARE -cE R ALLEN 0 ~o 1000 Km' H • Active volcano Trt"ch r .. .. r ' Al ..J _, I ) I l: (' v J ,.' I I ( I I ) I f F1G. !I - - '.\laps at sa me si:ale of Philippine, San A ndreas, Atacama, and Alpine faults. llRCT:.\f-l'ACIPf(' FAl"LTl;\C <.:ific ·ring of fire.' This point of concealed beneath the sea and the Yiew was emphasized by the great author suspec t s that similar Alaskan earthquake of 1958 (Toch throughgoing fractures of transcur er, 1960), which was associated rent character probably underlie with almost pure strike slip in a many of the less well exposed is region where most geologists proba lana-arc areas. Likewise, the 150- bly would not ha\'e predicted this km length of fault exposed in the type of displacement. Furthermore, Longitudinal Valley of Taiwan pre the much-debated results of numer sumably represents only a short ous fault-plane solutions which in segment of its total length; every dicate that three-quarters of the thing that is known about this fault world's earthquakes are caused by indicates that it, too, is a master dominantly transcurrent displace fault of regional tectonic impor ments (Hodgson, 1957; Schiedgger. tance. 1959) begin to look more and more geologically reasonable. I'.egarding the question of possi ble counterclockwise rotation of the The Philippine fault zone is phy Pacific basin (Benioff, 1959; St. siographically and structurally fully Amand, 1959), it seems clear that as spectacular as the better-known the sinistral displacement that pre San Andreas and Alpine faults, and vails in the Philippines-Taiwan re <.:ertainly it deser\'es a place among gion is not in harmony with the dex these as one of the great regional tral displacements that have been tectonic features of the earth (Fig tlemonstrnted in Alaska, California, ure 9). Its current acti\'ity is dem Chile, and New Zealand. This point unstrated by abundant eYidence of of \·iew has pre\•iously been express Hec:ent and probably historic dis ed by Biq ( 1958), and the principal placements, and consistent offset contribution of the present study is streams indicate that Quaternary not so much to refute the rotation displacements have been predomi hypothesis as to put forth concrete nantly transcurrent and sinistral. geologic evidence for the sense of In the opinion of the author, the displacement on major faults of the linearity - or Yery smooth curva Philippines-Taiwan region. Indeed, ture - of the fault trace in itself one might argue that the true edge demands a history of horizontal dis of the rotating 'disk' is along the placement has been at all large. The Bonin-l\fariana-Palau arc system, total length of the Philippine fault following the Andesite Line (Hess, zone within the archipelago is more 1948), rather than along the Ryuk than 1200 km, and its remarkable yu-Taiwan-Philippine system, and parallelism and proximity to the thus sinistral displacement in the :Mindanao trench suggest a close latter area does not necessarily des causal relationship between the two troy the rotation hypothesis. On the features. If the Philippine Islands other hand, there are a \'ariety of were slightly more submerged, the reasons for arguing against the ro trace of the fault would be largely tation hypothesis, the most cogent 1-l ~: CL\ TIEXl F. R. ALLEX being the failure of the hypothesi" 11am, 19-10). Similar complexities to account for the conYexity of the probably exist else\\"here (in Japan. arc::; or the acute cusps between ad for instance) . where no clear-cut jacent arcs. as has been pointed out picture or systematic transcurren t by Benioff (1961) in a recent re cli;:;placements has as yet e\'OlYecl. c,·aluation of the rigid rotation hy thesis. The pattern of individual currents underlying the circum Pacific rim appears to be far more This study was carried out with complicated than a single rotational tL111ds from the first Grove Karl Gil cell. For example. demonstrated bert A ward in Seismic Geology from sinistral displacement in the Philip the bequest of the late Harry 0. pines and Taiwan area giYes strong Wooc.1. The fiE>ld work was possible support to the proposal of Vening only with the generous cooperation ::\Ieinesz (193-1) that the Indonesian of the Geoligical Survey of Taiwan, archipelago is be i n g protruded Taiwan Xational Cniversity, Chi southeastward toward Australia nese Petroleum Corporation, Uni \\"ith accompanying sinistral shear Yersity of Philippines. I am espe ing along the Philippine flank and cially indebted to field colleagues C. dextral shearing along the southern S. \\'ang, T. L. Hsu, C. Y. Meng, flank. This theory is supported by E. A. Gamus, J. Pilac, and E. Y. obsenations of transcurrent fault Tames is. ing in Sumatra (Reid, 1913; Dur- REFERENCES Aki, K .. Furthe1 study of the mechanism --- . Mo,·ements on major transcur of circum-Pacific: earthquakes from rent faults. chapter -l in Continental Rayleigh waves. J. Geophys. Research. Dript. edited by S. K. Runcorn, Acade li.i. -l lfi:i--l 172. HJGO. mic Press, London. l 9G2. Akarnz. A. P., The major structural lines Riq Chingchang, Is the Pacific basin ro of the Philippmes. Philippine Geologist. tating- counter-clockwise? (In Chinese I t 2). 1:)-17. 19-J'i. English summary). Trans. Chinese As soe. Advance. Sc:i., 2(1), 10-11 and lG Alvi1·, A. D .. A theory on the major tecto l!l. J!l.)8. nic: structure of Luzon, Philippine Is lands. Proe. Pan Pacific Sci. Congr.. - - - . Circum-Pacific tectonics in Tai :lnl Cong1 ., 1, -l51-4!i-l. 192G. wan, Rept. Intern. Geo!. Congr., '.?l5t Congl' .. 18, 203-21-1, 19GO. ---. The Philippine fault zone and its relation Philippine structure, Philip Centeno y Garcia, J ., Informe sobre los pine J. Sci., 'i-1, -111-421, 19.U. temblores de tierra ocurridos en el mes de Julio de 187-, en el distrito de Su Benioff, H., Circurn-Pacific tectonic:s, l'igao, isla de :Mindanao, Bo!. Comision Pubis. Dominion Observatory Ottawa. de :.\lapa Geologico de Espana, 9, 215- 20 (2)' 395--102, 1959. 221, 188'.?. CIRCC::\1-PACIFIC FA CLTIKG 1-14 Durham,,). \\-.. Oeloe Aer fault zone. Su ---- . Continental rifting and igneous matra. Bull. Am. A ssoc. Petrol. Geolo actiYities in the Xeogene marginal geo g ists. ~ I, 35'.J-:JtiZ, lfi-10. synclines of Taiwan, Proc. Geo!. Soc. China. 110. l. 27-3.i. Jft58. Gutenberg, B .. and C. F . R ichter, Seismi city of t he Earth, 2nd ed .. Princeton hing, l'. B .. and E . i\l. :\lckee. T errain Cniversity Press. Princ:eto, >:. J., 195-1. d iag-rams of the Philippine Island:; . Hull. Gcol. Soc. Am.. GO, 1820-1!)3fi. Hess. 11.. :\lajor structural features of JI. HH~'- "-estcrn north l'acific, an interpretation of 11. 0. ;)-185, Bathymet ric Chart, Ko Leith. A.. The geology of t he Baguio rea to Kew Guinea, B ul l. Geo!. Soc. Am., Gold district. Philippine (Common ;:i'. 1• -l I 1--1-t(i, 1 ~'-18. ,,·ea lth ) Dept. Agr. and Commerce, Tech Bull. !l, ]!)38. l!o. C.~ .. (;eo!ogic relationships a nd com parison between Taiwan and t he Philip Len:,en. G. .J .. A method of grnben and pines, Proc. Geo!. Soc. China. no. -1, 3- horsl format ion, J. Geo/., uU, 579-;:iSI, 31, 19Gl. 1\li'i8. llodg-so11. .J. l l.. K ature of faulting in Li\·ingston. C. \\·., :\Iechani's of vein for large earthquakes, Bull. Geo!. Soc. Am .. mation in the northern half of the Ba 1;z<_ Gll-643, Hli>I. guio di:;trict, Eng. 1\1 ining J., 1-10 (!!), ;Jb- -1~ , and J..10(10), 4!J-5l, 10:~9. ll su. T. L.. The earthquakes of Taiwan (in Chinese). Quart. J. Bank of Tai :\l a s:i, :.\1. S., La seismologia en Filipinas. wan, 7(2), 1-18-lG-!. 1U;:i5 . T1pog ralia de Hamirez y Cia., ;\lanila ( Obsen·ato rio de i\Ianila J, 18:15. ----, Geolog y of t he Coastal Range, eastern Taiwan. Bull. Geo!. Survey Ta - ---, The seismic centers of Samar, iwan. 8, 3'.l-G:3, 1!•5G. Leyte and eastern ;\Iindanao. Philippine \\'eather Bur., ~l on t hly Bull. A ug. 1:11 0, - ---, Recent faulling in the Longitu 27!.J-291. l!JlO. dinal \'alley of eastern Taiwan. Geo!. Soc. China i\1em. 1, in press, HH52. - ---, The ;\lati earthquake. April 1-1. 1\.12-l . Philippine Weather B ur., Seismol. ln·ing. E. i\l., Submarine morphology of Bull. l~t2-J , 1-2, l\!2-l. the P hil ippine archipelago and its geolo g ic s ignificance. P hilippine Sci., 80 J . .i\Ielendre:;, :\I. :.\I.. Jr., and F. Comsti, Hc (1), 55-88, 1951. rnnnissance geology of southeastern Da Yao. Philippine Geologist 5 (2). 38-46, ----. Geologic map sheets. Philip 1951. pines Proc. Pacific Sci. Congr .. Pacific Sci. Assoc.. 8th Congr .. 1953. :\l ii1oza, \Y. A .. A. T. Ocampo, Jr., and .S . Bellos illo. Jr. . S ignificant Philippine lr\·ing, E. i\l., and J. S. THes, T he lloilo earthquakes, 19-1 9-1959, Philippine Wea earthquake of January 25, Ht-1 8. Panay ther Bur. Sci. Paper 101. 19GO. Island. Philippine Islands. Philippine GC:o logist. 2(2). (i-1 7, l!J-1 8. :\oble, L. F .. The San Adreas fault zone from Soledad P ass to Cajon P ass. Cali Juan. \". C .. Physiography and geology of t o rnia, chapte1-, -1. pp. 37--18, in Geolog~· Taiwan. Proc. Pacific Sci. CongT., Pa of Southem California. edited by R. lf. cific Sc:i . Assoc., 8th Congr.. 2, 2131-;JJ ~. .Jahn :;, California Div. ;\lines Bull. J'iO. l '.J5t5. 1!151. 145 CLA.REI\CE R. ALLEN Omori. F., On the Bokusekikaku and Bas Bull. Geo!. Soc. Am., 71. 1!)65, 1960. sh isho (Formosa) earthquake of .Jan uar~· 11 . 1908. Bull. Imp. Earthquake Tocher. D.. The Alaska earthquake <>f Investigation Comm. (Japan), 2, 1.)li- July 10, Hl5 : :.\lovement on the Fair 165, mos. weather fault and field investigation of : Heid, ;\l. F ., Su Repetti. W . C., Tectonic lines of the Phil \"onDrasche. R , Fragmente zu einer geo ippine Islands, Philippine Weather Bur.. ologie der Insel Luzon (Philippines), Seismal. Bull. 19:3-t, 57-71, 1935. C:erold's :"ohn, \'ienna, 1878. Richter, C. F., Elementary Seismology. \\". \V ill is. B.. Geologic obsen·ations in the H. F reeman and Co .. San Francisco, Ca Philippine archipelago, Natl. Research lif.. Hl.58. Council P hilippines, Bull. 13. 1937. Ritsema, A. R., and J. \'eldkamp, Fault --- P hilippine earthquakes and plane mechanisms of sout heast Asian structure, Bull. Seismol. Soc. Arn., 3.t, earthquakes, Komikl. Ned. l\1eterol. Inst. i>fl-81, Hl.t-1. :'lfededel. en Verh .. no. 7!1. HIGO. Wilson, .J. T.. The dHelopment and Scheidergger, A. E., Statistical analysis structure of the crust, chapter -1 in The of recent fault plane solutions of earth Barth as a Planet. edited by G. P. Kui quakes, Bull. Seismol. Soc. Am., 4~1. per. Univers i t~· of Chicag-o Press, Chi 3:3 7 _;3.i ' . 19.?tl. cag-o, III.. 1'.l!il. St. A mand. P., Circum-P acific orngeny. \\'isser. E .. Deformation in t he Baguio Pubis. Dominion Ob:servatory Ottawa, g-old district, P hilippine I slands (abs 20 (2)' ·103-411, 1959. tract) : Bull. GC' ol. Soc. Arn .. 50. l!)..1:-l HJ.t.t. 193!1. ----. Observaciones y interpretacion de 1960, Cniv. Chile. Communicaciones ----. Preliminary notes on the rela Escuela Geologia, 1 (2), 1961. tion of Philippine mining- districts to the structural history of the archipela- and C. R. Allen , Strike-slip go. Philippine Geologist. 6 ( 4), 77-84, faultini:r in northern Chile (abstract, l llii2.