Late Quaternary Displacement at the Hikihashi and

Transcription

1 Late Quaternary Displacement at the Hikihashi and Kitatake Faults, Miura Peninsula, Japan1) D. N. WILLIAMS2) Abstract As a result of field mapping and aerial photograph and map interpretation, maps and tables of late Quaternary displacements at the Hikihashi and Kitatake Faults are presented. Both faults, exhibiting clear topographic displacements, are Certainty I features according to the classification of MATSUDA et al. (1977). The Hikihashi Fault, upthrown to the south, is interpreted as dominantly right-lateral, nantly right-lateral, with a probable thrust component. Tabulated data show the Hikihashi Fault has an average horizontal displacement rate of at least 2m per 1,000 years and vertical displacement rate of about 0.5m per 1,000 3m per 1,000 years and vertical displacement rate of more than 0.4m per 1,000 years. Both faults, using Matsuda's 1977 classification, are thus Class A active faults, with net slip rates of greater than 1m per 1,000 years. Ratios of horizontal to vertical displace- of north. A comparison of recorded seismicity in Miura Peninsula with estimated recurrence intervals for transcurrent faults suggests there could be movement at an active fault in the peninsula within a few hundred years. Introduction In 1978 the writer participated in a joint project to map the active faults of Miura Peninsula some 45km south of Tokyo (Fig. 1). Separate publication of results having been requested by the mapping group, this paper presents maps of the Hikihashi Fault and of the central and eastern sections of the Kitatake Fault, together with tables listing displacement data and estimates of displacement rates. The most western part of the Kitatake Fault was not mapped in this study, because there was no time to examine its numerous branches in that area. Geological Setting and Surface Chronology The rocks of Miura Peninsula (GEOLOGICAL SURVEY of JAPAN, 1976) range in age from Miocene to Quaternary. The oldest rocks of the peninsula are massive mudstone and alternating tuffaceous sandstone and mudstone of the early to mid Miocene Hayama- Hota Group, cropping out in the hills between the Takeyama (Shitaura) and Kitatake Faults, and also in the northwest. These rocks are unconformably overlain by middle to late Miocene rocks of the Miura Groupprincipally alternating sandstone and mudstone, and pyroclastic sandstone. In the depression between the Takeyama and Minamishitaura Faults, in the centre of the peninsula, lie relatively unconsolidated sand and mud deposits of the middle Pleistocene Miyata Formation. Thin, mappable pyroclastic beds are found in the Miura Group rocks. Dips within the Miyata Formation are up 1) Accepted April 12, ) Earth Deformation Studies, New Zealand Geological Survey, P. O. Box 30368, Lower Hutt, New. Zealand.

2 290 The Quaternary Research Vol. 21 No. 4 Jan to 15 degrees, but the Miura and Hayama- Hota Groups are more deformed, with broad folds striking east-west and northwest-southeast. Some folds are overturned. The major faults delineated from geological work are those active in the late Quaternary, although the surface traces do not everywhere coincide exactly with the principal faults shown on the geological map (GEOLOGICAL SURVEY OF JAPAN, 1976). The major faults Fig. 1 Miura Peninsula and its location. Dashed lines indicate principal active faults. strike northwest-southeast, or east-west, and as pointed out by KANEKO (1969) divide the peninsula into a series of blocks. The blocks are further divided by minor faults striking north-east, especially in the north-central part of the peninsula. The minor faults, studied by KAKIMI (1976) have little or no expression in the landscape; none appears to be active. The rocks are in most places strongly weathered, with a mantle of late Quaternary tephra up to 15 metres thick, principally from the volcanoes Fuji and Hakone (lying 80km and 55km, respectively, to the west). Fluctuations of sea-level in the late Quaternary and tectonic uplift (especially in connection with the formation of the Neogene-Quaternary Kanto Basin), have resulted in flights of coastal terraces about the peninsula. A Holocene marine terrace, mantled by dune sand in several places, fringes the peninsula and has fluvial surfaces graded to it. For this study its age throughout the area is taken as about 6,000 years (cf. MATSUSHIMA, 1976). The marine Misaki, Obaradai, Hikihashi and Shimosueyoshi Terraces are also well distributed and from the dated tephra overlying them are estimated to have ages of 60,000, 80,000, 100,000, and 130,000 years B. P., respectively (MACHIDA, 1975). Previous Work Active faults in Miura Peninsula have been recognized by several workers (e.g. SUGIMURA, 1964), and faulting at the Shitaura Fault (now generally referred to as the Takeyama Fault) accompanying the 1923 Kanto Earthquake was recorded by YAMASAKI (1925). KANEKO (1969) summarized previous studies of active faulting in the area, presented a detailed map of the 1923 Shitaura Fault displacement and provided late Quaternary displacement data based upon field work and interpretation of aerial photographs. He listed the six major active faults (Fig. 1) and recognized that they were dominantly rightlateral and that all, with the exception of the Minami-Shitaura and Hikihashi Faults, have wide crush zones. KANEKO (1969) concluded that the right-lateral slip at the faults resulted from a regional compressive stress oriented north-south, or NNW-SSE. His studies indicated that the 1923 displacements at the Shitaura Fault were entirely dip-slip and secondary, the 1923 shock being centred on the Sagami Trough offshore to the southwest. In 1972 KANEKO described the Okusayama area (near the western section of the Kitatake Fault) in detail, and mapped up-and down-

3 buckled areas, suggested by him to result from fault curvature. Present Investigation Methods This study was commenced with stereoscopic interpretation of aerial photographs at scales of 1:40,000 and 1:20,000, to identify active faults and to map some prominent surfaces. Geological and geomorphological displacement data were noted in the field, and recorded on the accompanying tables (Tables 1-7), with localities plotted on 1: 10,000 and 1:20,000 scale field sheets, generalized here as Figs. 2 to 4. Urbanization has destroyed surface evidence of faulting in some areas, especially along parts of the Kitatake Fault; in such cases this study relies heavily upon previous work. In some rural areas, fault traces were difficult to locate with accuracy because of thick vegetation on hill slopes and artificial walls and embankments in valleys. In such places fault scarplets were difficult to recognize, but gross features such as grabens and offset valleys could still be seen clearly. Field studies were followed by further aerial photograph interpretation and analysis of maps of 1:2,500 and 1:3,000 scales, which had 2 m contour intervals. These data were supplemented by the special study on stream displacements carried out by ANDO (1972). Fault displacements were measured in the field by hand-level and by pacing, and in the case of large displacements, were measured directly from the 1:2,500 and 1:3,000 maps with interpolation from the 2m contours. Ages of fault displacements were estimated by noting the relationship of displacements to depositional surfaces of known maximum age (post-glacial tephra-mantled surfaces developed on the Obaradai and Misaki marine terraces, terraces dated as about 80,000 and 60,000 years B. P. by MACHIDA, 1975). Several valleys and intervening interfluves (spurs) have developed on these surfaces, and as reference lines, may be considered to postdate those surfaces, but the aggraded valley floors are graded to base levels close to dated Holocene sea-levels, and are therefore considered to form reference surfaces no older than about 6,000 years (e.g. MATSUSHIMA, 1976). Streams incised into such valleys are, as reference lines, less than 6,000 years old. Fig. 2 Hikihashi Fault. Upthrown side is on the south and southwest. Double trace indicates graben (net upthrow to southwest). Locality numbers refer to Tables 1 to 3.

4 292 The Quaternary Research Vol. 21 No. 4 Jan Vertical displacements of the Misaki Surface, and possibly of the Obaradai Surface, have occurred, but because of late tephra accumulation, heights of scarps preserved on the ground surface indicate merely the minimum amount of vertical displacement since terrace formation. The maps and tables form the essence of this report. The terminology in the tables follows that in general use, except that "offset stream" has been restricted to offsets of incised streams where lateral displacements can be measured to within about 10m and in most "offset valley" has been used to refer to an offset of an aggraded valley, where there are possible inaccuracies in displacement estimation of up to twice the valley width. Hikihashi Fault Displacement data are recorded in Tables 1, 2 and 3; relevant locality numbers are shown in Fig. 2. As already determined by KANEKO (1969), the active trace of the Hikihashi Fault extends from Hikihashi southeast towards the Kaneda area, where it gradually changes strike to lie east-west. However, the results of the present mapping differ in detail from those of Kaneko. Short (few hundred metres long) traces mapped south of the Hikihashi Fault proper (KANEKO, 1969 Fig. 5) could not be found. On the other hand, a depression associated with possible offset spurs south of Maruyama is considered to mark a probable active fault though not mapped in this study. A conspicuous feature of the fault is the development of grabens, especially where the fault strikes about east-south-east. As noted by KANEKO (1972) there is some downwarping of surfaces towards the grabens, but the amount is difficult to measure because of the uneven thickness of tephra on the surfaces, uneven erosion of the tephra, and a paucity of outcrops of terrace gravels or cut surfaces; in short, clear reference surfaces are few. As a result, the presence of transcurrent buckling such as that reported by KANEKO (1972) cannot be deduced. In contrast, it is considered that what warping can be observed has axes sub-parallel to the trace and is especially pronounced on the southwest side of the fault (e.g. locality 7), suggesting that such folding may be due to a thrust component of faulting, or drag. The grabens could be a response to transcurrent movement on right-stepping en echelon, fault segments. However, as right-stepping en echelon steps are usually major, and minor steps are usually left-stepping for right lateral faults, that mechanism is considered unlikely. A preferred cause for graben development along the western part of the trace (for example at observa- is there more nearly parallel with the principal axis of compression (see below). For most of its length, the Hikihashi Fault has controlled the Misaki shoreline which is cut into the Obaradai Terrace (Fig. 2). Thus the fault was repeatedly active at the culmination of the Misaki sea, uplift at the fault compensating for most marine erosion of the fault scarp. As a result, no pre-misaki reference lines or surfaces are preserved across the fault in such localities. Post-Misaki stream incision has been accompanied by continued dextral faulting. Aggradation during the Holocene has widened such valleys so that measurement of offsets is difficult, except where late downcutting has taken place. In several places Holocene alluvium has been displaced vertically. The Hiki. hashi Fault appears to terminate abruptly at both ends, although erosion may have destroyed late Quaternary fault traces at the extremities. To the west, the fault displacement may be dissipated across the zone between the Hikihashi and Minamishitaura Faults (Fig. 1). On balance, considering the relatively narrow crush zone developed on the fault (KANEKO, 1969), its relatively poor physiographic expression and its relative lack of displacement of local basement rocks (GEOLOGICAL SURVEY OF JAPAN, 1976), movement at the Hikihashi Fault was probably initiated in the late Quaternary, and the fault is considered to be a relatively immature feature.

5 Kitatake Fault The Kitatake Fault can be traced from the Uraga Channel across Miura Peninsula to Sagami Bay. It forms the boundary between the dissected upland of the Okusayama horst (KANEKO, 1969) to the north and the less dissected hills and high terrace lands on the northeast part of the Takeyama Range. As the fault is traced northwest from the Uraga Channel, displacement features are clear as far as Kitatake but the fault has progressively less expression in the Nagasaka area. This is likely to be due to displacement at the Kitatake Fault being distributed across several active faults in the Nagasaka- Akiya area (e.g. KANEKO, 1972). Further work would be necessary to establish the relationships of those faults. Displacement data are presented in Tables 4, 5, 6 and 7 and relevant locality numbers are shown in Figs. 3 and 4. As noted by KANEKO (1969, p. 203), the blunted scarp facets, fault gaps and grabens, or tectonic trenches, and commonly also by right-laterally offset ridges and streams". However, the "tectonic trenches" may be entirely due to erosion along the crush zone of the fault, as no definite evidence of displacement was found on the southwest side of the depressions. Several offset valleys noted by KANEKO (1969) and studied in detail by ANDO (1972) have since been destroyed, especially in the "Highland" housing area near Nobi. The data nevertheless are incorporated in this paper. No clear evidence of transcurrent buckling, as noted by KANEKO (1972) was found. However, his earlier observation (KANEKO, 1969) of downwarping of a wave-cut bench towards the fault, on the uplifted northeast side of the fault, is confirmed. Such tilting with an axis sub-parallel to the fault, is likely to be on the southwest limb of an anticline; its attitude and location suggest a thrust component of faulting. Minor thrusting is also suggested by the trace, over most of its mapped length, tending to occupy northeasterly positions in valleys and southwesterly positions on higher ground. Fig. 3 Eastern section of Kitatake Fault. Key to surfaces is on Figure 2. Locality numbers refer to Tables 4 to 6.

6 294 The Quaternary Research Vol. 21 No. 4 Jan Fig. 4 Central section of Kitatake Fault. Key to surfaces is found in Figure 2. Locality numbers refer to Tables 6 and 7. Summary and Discussion Both the Hikihashi and Kitatake Faults are recognized as Certainty I active faults, exhibiting clear topographic displacement with sense of displacement evident (MATSUDA et al., 1977). Based on the data tabulated (Tables 1-7), both faults are of Mass A, with average late Quaternary slip-rates of greater than 1 metre per 1,000 years (MATSUA, 1977). At the Hikihashi Fault, observations 3, 10, 12, 14 and 17 indicate dextral displacement averaging at least 2m per 1,000 years over the last 60,000 years and at about 3m per 1,000 years over the last 6,000 years, although there are few data on Holocene displacement. The rate of vertical displacement is about 0.5m per thousand years, upthrown to the south. There are too few data to determine changes in horizontal and vertical displacement due to varying strike. At the Kitatake Fault, observations 2 to 6 indicate that dextral displacement of about 1.2m per 1,000 years over the last 80,000 years, and at greater than 3m per 1,000 years during the past 6,000 years. A single observation (No. 3) indicates a vertical displacement rate of greater than 0.4m per 1,000 years, upthrown to the northeast, over the last 6,000 years. Both Kitatake and Hikihashi Faults have moved with consistent sense in the late Quaternary. The sense of future displacements is expected to be similar. Unfortunately, there are no data on the size of individual displacements that can be expected at the faults; the smallest observed lateral displacements of 20m (Kitatake Fault observation No. 12) and 18m (Hikihashi Fault, observation No. 5) are, based on recorded displacements elsewhere, interpreted as resultants of several discrete movements. Similarly, individual vertical displacements cannot be reliably estimated, especially as lateral movement has created apparent vertical displacements in some places. A few data, from localities where horizontal and vertical reference features can be matched, allow calculations of approximate shortening directions in the peninsula, following LENSEN (1958). (See Table 8.)

7 Table 1 Hikihashi Fault displacement data. Table 2 Hikihashi Fault displacement data (continued)

8 296 The Quaternary Research Vol. 21 No. 4 Jan Table 3 Hikihashi Fault displacement data (concluded) Table 4 Kitatake Fault displacement data Of the two sets of results, that of the Kitatake Fault is considered more reliable, principally because there the fault strike is more easily determined. This preferred val- with the northerly or NNW-SSE direction estimated by KANEKO (1969). Both the Kitatake and Hikihashi Faults show evidence of repeated late Quaternary and Holocene activity, but seismological data summarized by the RESEARCH GROUP FOR ACTIVE FAULTS (1980) detail no major earthquakes with an epicentre in Miura Peninsula in recorded history (-700 AD to 1978 AD), although it is possible that the magnitude 6.9 Uraga Strait Earthquake on 26 April 1922 was on the southeasterly extension of the Kitatake or Takeyama Faults (TOKYO ASTRONOMICAL OBSERVATORY, 1982). The recurrence intervals derived for major earthquakes and accompanying fault displacements-principally for reverse and transcurrent faults in Japan and elsewhere-are in the order of less than 1,000 years to 2,000 years (e.g. MATSUDA, 1977), and the recurrence interval for transcurrent faults would

9 Table 5 Kitatake Fault displacement data (continued) Table 6 Kitatake Fault displacement data (continued)

10 298 The Quaternary Research Vol. 21 No. 4 Jan Table 7 Kitatake Fault displacement data (concluded) Table 8 Derivation of approximate principal horizontal shortening direction in Miura Peninsula, a is the angle between the fault strike and the shortening direction (after LENSEN, 1958) be expected to be towards the lower end of this range. Thus movement at an active fault in Miura Peninsula, possibly at either the Kitatake or Hikihashi Faults, could be expected within the next few hundred years. Acknowledgements This work was carried out while on leave from N. Z. Geological Survey, D. S. I. R., New Zealand. I thank the Miura Peninsula active fault mapping group for allowing me to participate in this project. Numerous people assisted me in my studies in Japan, especially Prof. Y. OTA, Yokohama National University and Prof. T. YOSHIKAWA, University of Tokyo. I am indebted to Prof. T. MATSUDA, Earthquake Research Institute, University of Tokyo and to Mr. K. WATANABE, Geography Department, Yokohama National University, for constructive criticism in the field. The manuscript typed by IRENE GALUSZKA has benefited from criticisms by Mr K. R. BERRYMAN and Mr A. G. HULL. References (J) and (E) indicate papers in Japanese and English respectively. (J, E) indicates a two-language publication. (J+E) indicates a paper in Japanese with English abstract, and vice versa. ANDO, K. (1972) Amount of stream valley offset caused by strike-slip faulting in Miura and IZU Peninsulas and in the Yamazaki area, Hyogo Prefecture. (J). Chirigaku Hyoron, 45, p GEOLOGICAL SURVEY OF JAPAN. (1976) Geological Map of Tokyo Bay and adjacent areas 1: 100, 000. (J+E). KAKIMI, T. (1976) Relations between major and minor

11 fault systems in the central part of Miura Peninsula. (J+E). Niigata University, Faculty of Science, Geology and Mineralogy Research Report, 4, p KANEKO, S. (1969) Right-lateral faulting in Miura Peninsula South of Tokyo, Japan. (E+J). Journal Geological Society of Japan, 75, p KANEKO, S. (1972) Some Remarks on the Strike Slip Faulting in Miura Peninsula and Sagami Bay Area, South Kanto, Japan. (E+J). Journal Geological Society of Japan, 78, p LENSEN, G. J. (1958) Rationalized Fault Interpretation. (E) N. Z. Journal of Geology and Geophysics, 1, p MACHIDA, H. (1975) Pleistocene Sea Level of South Kanto, Japan, analysed by Tephrochronology. (E) Royal Society of. N. Z. Bulletin, 13, p MATSUDA, T. (1977) Estimation of Future Destuctive Earthquakes from Active Faults on Land in Japan. (E) Journal Physics of the Earth, 25, Suppl., S MATSUDA, T., OTA, Y., OKADA, A., SHIMIZU, F. and TOGO, M. (1977) Aerial Photo-interpretation of Active Fualts-the individual Difference and Examples. (J+E) Bulletin Earthquake Research Institute, 52, p MATSUSHIMA, Y. (1976) The Alluvial deposits in the southern part of the Miura Peninsula, Kanagawa Prefecture. Bulletin Kanagawa Prefectural Museum, no. 9, (J). RESEARCH GROUP FOR ACTIVE FAULTS (1980) Active Faults in Japan. University of Tokyo Press. p TOKYO ASTRONOMICAL OBSERVATORY (1982) Zishin, in Rikanenpyo, Maruzen, Tokyo. p (J). SUGIMURA, A. (1964) Notes on the Minamishitaura Fault and the Takeyama Fault, Japan. (J). Journal Geological Society of Japan, 70, p YAMASAKI, N. (1925) Physiographical Investigation of the Great Earthquake of southeast Japan. Rep. Imperial Earthquake Investigation Committee. 100 B, p (J).