Paleomagnetic directions of the Kawabata Formation
Reflecting a complicated subduction and collision history on the eastern Eurasian margin, central Hokkaido has been a site of various types of basin formation. Thick piles of the Cretaceous and Paleogene sediments (Figure 1; ) buried a regional forearc basin subducted by the Izanagi/Kula and Pacific Plates. Paleomagnetic studies of the Cretaceous Yezo Supergroup [2,3] showed that the present forearc is divided into some basins developed in different areas. Sedimentary system and forearc basin architecture in the Paleogene was studied in detail by Takano and Waseda  and Takano et al. .
Under the influence of arc-arc collision on the Pacific convergent margin, vigorous mountain building and formation of foreland basins became active since the late Cenozoic. The Ishikari-Teshio belt (see Figure 2) is underlain with thick middle Miocene clastic strata. These are the Kawabata and its correlative formations, derived from the longitudinal mountainous ranges that were uplifted and eroded during that time . It is generally regarded as a typical foreland setting, and the burial history of turbidites and associated coarse clastics of the Kawabata Formation has previously been studied from a sedimentological viewpoint (e.g., ). The process through which the Miocene basin developed in central Hokkaido is not only governed by compressive stress in the collision zone, but also by coeval tectonic events like back-arc spreading in the Japan Sea (e.g., ) and dextral transcurrent faulting along the Eurasian margin (e.g., ).
In this paper, we present preliminary results of rock magnetic analyses of the Cretaceous Yezo Supergroup, the Eocene Ishikari Group and the Miocene Kawabata Formation in order to detect tectonic movements around the basin and to describe the microfabric of sedimentary rocks related to the tectonic regime and sedimentation processes in the mobile zone. This study is an attempt to apply magnetic properties to tectono-sedimentology.
The Yezo Supergroup deposited on the Cretaceous forearc and consists of monotonous mudstone intercalated by coarse clastics and ash layers. After a stagnant subsidence stage at the beginning of the Cenozoic, fluvial sediments of the Ishikari Group and its correlative units began to bury depressions on the forearc. As a result of strong deformation and continued sedimentation on the active margin, surface distribution of the Eocene Ishikari Group is rather restricted. However, numerous exploration drilling clarified that voluminous Paleogene units are concealed under the alluvial plain (Figure 1). Paleogene depositional sequence and facies classification were described by Takano et al. . They are shown in Figure 3 using abbreviations.
The study area of the Kawabata Formation is located in the southern part of the middle Miocene basins of the Ishikari-Teshio belt. Folded sedimentary units are distributed with a NNW-SSE trend, and are cut by numerous faults (Figure 2). The area is divided into the following formations in ascending order : the Takinoue Formation, the Kawabata Formation, the Karumai Formation, and the Nina Formation (Figure 4). They represent the sequence by which an elongate N-S foreland basin was filled. The middle Miocene Kawabata Formation comprises mainly turbidites and associated coarse clastic rocks derived from the eastern hinterland .
2.2. Sedimentary facies of the Miocene unit
This study conducted sedimentary facies analysis for the Kawabata Formation along the Rubeshibe River (Figure 2). The analysis revealed that the turbidites of the Kawabata Formation mainly consisted of sheet-flow turbidite facies association and channel-levee facies association (Figure 5). The sheet-flow turbidite facies association comprises aggradational stacking of rhythmic alternating beds of turbidite sandstone and mudstone with rare upward thickening or thinning successions, and is interpreted to be sheet-like turbidites with minor occurrences of depositional lobes, which occupied major part of the trough-like foreland basin fill [7,10]. The channel-levee facies association is composed of thick amalgamated sandstone facies with slump blocks and thinly bedded alternating beds of sandstone and mudstone. These two facies appearing coupled is indicative of an elongated channel-levee system made of the main channel with levees on both sides. These two facies associations are believed to have been deposited in an elongated trough-like foredeep in the foreland basin . The turbidites of the Kawabata Formation commonly contain sedimentary structures indicating paleocurrent directions; e.g., sole marks (mostly flute marks) at the bottom of individual turbidite bed, and current ripples in Bouma Tc division .
3. Rock magnetism
We obtained samples for rock magnetic analyses exclusively from fine-grained parts of the target sedimentary units, since fine sedimentary rocks generally preserve stable detrital remanent magnetization (DRM). Few visible markers of the sedimentation process accompany such sediments, so we attempted to measure their microscopic magnetic fabric, which may be related to paleocurrent directions (e.g., ).
3.1. Basic measurements
The Cretaceous and Eocene samples were taken from outcrops along the streambed in central Hokkaido (Figure 3) using an engine or electric drill at 21 sites. Samples of the Kawabata Formation were collected with a battery-powered electric drill at 21 sites along the Rubeshibe River (Figure 2). The bedding attitudes were measured on outcrops to allow us to compensate for tectonic tilting later. Between seven and sixteen independently oriented cores 25 mm in diameter were obtained at each site using a magnetic compass. Cylindrical specimens 22 mm in length were cut from each core and the natural remanent magnetization (NRM) of each specimen was measured using a cryogenic magnetometer (model 760-R SRM, 2-G Enterprises). Low-field magnetic susceptibility was measured on a Bartington MS2 susceptibility meter, and the anisotropy of magnetic susceptibility (AMS) was measured using an AGICO KappaBridge KLY-3 S magnetic susceptibility meter. After the basic measurements, pilot specimens with average NRM intensities, directions and susceptibility levels were selected from each site for subsequent demagnetization tests.
3.2. Demagnetization tests
In order to isolate stable components of the remanent magnetization, progressive alternating field demagnetization (PAFD) and progressive thermal demagnetization (PThD) tests were carried out on two pilot specimens per site that had average NRM directions. The PAFD test loading ranged from 0 to 80 mT using a three-axis tumbling system with specimens contained in a μ-metal envelope. The PThD test was performed using an electric furnace, with a residual magnetic field less than 10 nT, beginning at 100 ºC and continuing until the specimen was either fully demagnetized and a characteristic remanent magnetization (ChRM) component was isolated, or until the thermal treatment provoked erratic behavior of the magnetic direction. Specimens’ low-field bulk magnetic susceptibilities were measured using a susceptibility meter after each PThD step in order to monitor chemical changes in ferromagnetic minerals.
Figure 6 presents typical PThD and PAFD results for the Yezo Supergroup and Ishikari Group. It is obvious that the ChRM direction was not isolated because of unstable behavior in thermal treatment (Figure 6a), overlapping spectra of primary and secondary magnetization (Figure 6b) and partial remagnetization within a site (Figure 6c,d). Therefore further analyses for magnetic granulometry were not applied on the Cretaceous and Eocene samples. On the other hand, PThD treatment was effective for isolating stable ChRM in the sedimentary rocks of the Kawabata Formation. Figure 7 shows typical results of the progressive demagnetization tests.
3.3. Hysteresis properties
Hysteresis parameters were determined for the Kawabata samples with an alternating gradient magnetometer (Princeton Measurements Corporation, MicroMag 2900). Ten sample chips up to 1 mm in size were randomly selected from site RB16, where stable ChRM has been successfully isolated. Figure 8 displays typical hysteresis of the Kawabata mudstones. The raw diagram seems to suggest the absence of ferromagnetic material. After correcting the linear gradient of paramagnetism, a weak ferromagnetic behavior signature can be recognized. Saturation magnetization (Js), saturation remanence (Jrs) and coercive force (Hc) values were determined for all samples from their hysteresis loops. Their relatively low Hc (~ 100 mT) implies that magnetite is the dominant remanence carrier. After acquiring coercivity of remanence (Hcr) values through backfield demagnetization experiments, we constructed a correlation plot of Jrs/Js versus Hcr/Hc  as shown in Figure 9. All the data are plotted in the pseudo-single domain (PSD) region of magnetite.
4.1. Rotational motions
We found stable magnetic components at three sites of the Kawabata Formation. Their directions were determined with a three-dimensional least squares analysis technique . Figure 10 and Table 1 present site-mean ChRM directions obtained from the Kawabata Formation. They exhibit antipodal directions, and precision parameter (κ) improves after tilt correction. Although the number of data points is minimal for tectonic discussion, we can interpret the site-mean directions as a record of the Earth’s dipole magnetic field, acquired before the strata tilted. The declination of the formation mean exhibits a significant westerly deflection, which suggests counterclockwise rotation of the study area.
A previous study  suggested a clockwise tectonic rotation around central Hokkaido based on a paleomagnetic study of the Kawabata Formation. Takeuchi et al.  proposed a coherent rotational model with ‘domino-style’ rigid crustal blocks. However, Tamaki et al.  criticized the block rotation scheme as being overly simplistic based on differential rotations inferred from Oligocene paleomagnetic data. They restored crustal deformation in central Hokkaido using dislocation modeling, and found complicated vertical-axis rotations around terminations of the faults that contributed to the formation of N-S elongate sedimentary basins. Figure 11 demonstrates differential rotation in central Hokkaido since the middle Miocene.
4.2. Sedimentation process inferred from AMS fabric
We found that the AMS fabric (orientation of principal axes) were precisely determined at all the sampled localities. Tables 2 and 3 show the AMS parameters for the Cretaceous/Eocene units and the Miocene unit, respectively. Figure 12 delineates typical AMS fabric obtained from the Ishikari (left) and Yezo (right) samples. After tilt-correction, the maximum (K1) and intermediate (K2) axes of AMS are bound to the horizontal plane with a subtle imbrication suggestive of hydrodynamic forcing.
|Site||N||K1 Str.||K2 Str.||K3 Str.||L||F||P||PJ||T||q||Unit / Sequence|
|(D, I)||(D, I)||(D, I)||(K1/K2)||(K2/K3)||(K1/K3)|
|BP01||11||1.010||1.005||0.984||1.005||1.021||1.026||1.028||0.619||0.213||Bb / Isk-2HST|
|(207, 4)||(297, 1)||(35, 86)|
|BP02||11||1.009||1.008||0.982||1.001||1.026||1.027||1.031||0.917||0.043||Bb / Isk-2HST|
|(226, 3)||(136, 4)||(0, 85)|
|BP03||11||1.010||1.005||0.985||1.005||1.020||1.025||1.027||0.593||0.229||Ik / Isk-4TST|
|(233, 9)||(141, 7)||(16, 78)|
|BP04||13||1.010||1.002||0.988||1.008||1.014||1.022||1.022||0.259||0.458||Ik / Isk-4TST|
|(225, 6)||(134, 3)||(15, 83)|
|HP01||18||1.010||1.004||0.987||1.006||1.017||1.023||1.024||0.479||0.303||Ak / Isk-3HST|
|(271, 4)||(180, 8)||(29, 81)|
|HP02||11||1.007||1.004||0.989||1.003||1.015||1.018||1.019||0.662||0.186||Ak / Isk-3HST|
|(38, 6)||(307, 4)||(187, 83)|
|NP01||15||1.009||1.006||0.985||1.003||1.022||1.025||1.027||0.762||0.128||Bb / Isk-2HST|
|(264, 2)||(174, 7)||(11, 83)|
|NP02||9||1.013||1.006||0.981||1.006||1.026||1.032||1.034||0.597||0.227||Bb / Isk-2HST|
|(252, 10)||(162, 2)||(59, 80)|
|NP03||10||1.006||1.004||0.989||1.002||1.016||1.018||1.019||0.779||0.118||Bb / Isk-2HST|
|(253, 1)||(163, 7)||(349, 83)|
|NP04||7||1.009||1.006||0.985||1.002||1.022||1.024||1.027||0.791||0.111||Bb / Isk-2HST|
|(80, 10)||(170, 4)||(282, 79)|
|NP05||9||1.007||1.003||0.990||1.004||1.013||1.017||1.018||0.536||0.264||Bb / Isk-2HST|
|(39, 28)||(141, 21)||(261, 54)|
|(71, 5)||(340, 11)||(187, 78)|
|(66, 2)||(156, 4)||(315, 86)|
|(51, 4)||(321, 11)||(163, 78)|
|(336, 11)||(70, 19)||(217, 68)|
|(8, 9)||(99, 11)||(240, 76)|
Figure 13 delineates typical AMS fabrics of the Kawabata Formation. Site RB08 typifies an elongate (prolate) fabric reflecting aligned detrital grains. Site RB14 has highly oblate fabric, as shown by a positive T parameter near unity. This fabric is essentially confined to the bedding plane under gravitational force. As the hysteresis study showed a negligible amount of ferromagnetic material in the Kawabata samples (Figure 8), we consider the AMS fabric as being governed simply by the shape anisotropy of paramagnetic minerals, i.e. alignments of elongate or platy grains such as amphibole or mica.
|Site||N||K1||K2||K3||L (K1/K2)||F (K2/K3)||P (K1/K3)||PJ||T||q|
|(D, I)||(D, I)||(D, I)|
|(16, 6)||(107, 7)||(242, 81)|
|(3, 20)||(98, 11)||(215, 66)|
|(326, 11)||(58, 11)||(190, 74)|
|(159, 2)||(249, 6)||(52, 83)|
|(4, 19)||(101, 19)||(232, 62)|
|(336, 5)||(66, 7)||(208, 82)|
|(135, 24)||(41, 9)||(291, 64)|
|(152, 2)||(61, 36)||(245, 54)|
|(121, 1)||(31, 7)||(218, 83)|
|(343, 2)||(252, 12)||(84, 78)|
|(313, 42)||(105, 45)||(210, 15)|
|(3, 6)||(93, 1)||(192, 84)|
|(119, 26)||(214, 11)||(325, 61)|
|(78, 4)||(348, 5)||(206, 84)|
|(188, 10)||(96, 10)||(324, 76)|
|(281, 4)||(11, 5)||(152, 83)|
|(292, 6)||(201, 6)||(66, 81)|
|(120, 1)||(30, 2)||(234, 87)|
|(26, 15)||(277, 50)||(127, 36)|
|(300, 19)||(205, 15)||(78, 66)|
|(215, 75)||(90, 9)||(358, 12)|
Sedimentological context of the AMS fabric is demonstrated in Figures 14 and 15. Paleocurrent directions inferred from the Eocene AMS data tend to align in N-S azimuth (Figure 14), and accord with development process of the forearc basin . Takano and Waseda  demonstrated that the Eocene paleo-Ishikari basin experienced differential subsidence during deposition. Such deformation may be related to longstanding strike-slip faulting around central Hokkaido , and tectono- / sedimentological context of the AMS fabric will be better evaluated in the light of quantitative study of basin-forming processes described in this book. For reliable interpretation of AMS data, it is necessary to assess properties of ferromagnetic minerals, such as composition, grain size and contribution to bulk magnetic susceptibility, as shown in this paper.
Our field survey revealed indicators of paleocurrent directions in the Kawabata Formation along the Rubeshibe River as depicted in Figure 15. After correction for the counterclockwise rotation identified in our paleomagnetic study, most of the markers indicate a westward current direction with minor southward flow contributions. This is consistent with a tectono-sedimentary model of rapid burial of the Miocene N-S foreland basin by clastics derived from the eastern collision front presented in such research as Kawakami et al. . Notably, the imbrication of the oblate AMS fabric matches visible sedimentary structures. Although the transport direction of muddy detrital material spilled out of a levee is not necessarily parallel to the turbidity current within a channel, AMS data can serve to indicate paleocurrents after the contributors to the magnetic fabric have been identified. Also note that K1 of prolate samples (with negative T parameters) tend to align perpendicular to the paleocurrent direction, implying that elongate grains roll on the sediment surface.
Figure 16 delineates groups of microscopic fabrics identified in the Kawabata Formation as a function of the AMS shape parameter (T). The intensity of alignment forcing inferred from AMS data is closely related to sedimentary facies (shown on the right in the figure) determined by field observation. For example, weak hydrodynamic forcing corresponds to fine rhythmically alternating facies in channel-levee systems. Thus, the sedimentological context of muddy sediments’ AMS fabric can be interpreted in the light of sandy sediments’ facies analysis.
Azimuths of AMS maxima in natural sediments vary significantly, reflecting the size or shape of magnetic grains and changes in current velocities (e.g., ). Figure 16 presents the relationship between paleocurrent proxies estimated from the imbrication of the AMS minimum axis (K3) and the K1 trend. Tarling and Hrouda  stated that the angle between K3 and K1 changes as a function of current velocity and the slope of the sedimentary surface. Our result suggests that the orientation between those AMS sedimentary indicators can vary, regardless of the level of hydraulic forcing, based on the shape parameter (T), which implies development of a preferred orientation. Although the AMS fabric is a diagnostic tool for patterns of sediment transportation, laboratory-based experiments that analyze natural sediments under conditions where a few of the prevailing factors are controlled, are essential to allow firm sedimentological interpretation of formation processes.
4.3. Re-deposition experiment and the origin of AMS
In order to consider the origin of the AMS in the Kawabata samples, we organized a re-deposition experiment. A silty sandstone (SP1C-1) and a mudstone (SP2F-1) samples were crushed and sieved into coarse, medium and fine fractions. The fine fraction (< 63 μm) was then separated into magnetic and non-magnetic fractions with an isodynamic separator. The ‘magnetic’ fraction actually contained no ferromagnetic opaque minerals such as magnetite, but had abundant biotite and common hornblende. It also contained garnet, probably derived from metamorphic rocks exposed around the hinterlands during the rapid deposition of the Miocene turbidite.
A suspension of the fine fraction was poured into a vertically settled plastic tube 1 m in length and 2.5 cm in diameter, filled with water. This deposit of artificial sediment was dehydrated at room temperature. After being soaked in an adhesive resin, the samples were trimmed into standard-sized specimens for rock-magnetic measurements. The AMS was measured with an AGICO KappaBridge KLY-3 S magnetic susceptibility meter. The AMS parameters for the artificial samples are summarized in Table 4.
Figure 17 presents the magnitudes of magnetic fabrics in natural sedimentary rocks and the re-deposited sediments of the Kawabata Formation. Obviously, the magnetic separation results in remarkable decrease of both the bulk susceptibility and the degree of anisotropy (PJ). It is also noteworthy that the shape parameter (T) of the artificial sediments is almost null, suggesting a neutral magnetic fabric. The directions of the principal AMS axes (see Table 4) are not bound to the horizontal plane or to geomagnetic north. Thus, the detrital particles, free from paramagnetic minerals having shape anisotropy, like platy biotite, are deposited without any gravitational or geomagnetic forcing, creating an isotropic sediment.
|Sample||N||K1||K2||K3||L (K1/K2)||F (K2/K3)||P (K1/K3)||PJ||T||q|
|(D, I)||(D, I)||(D, I)|
|(167, 75)||(265, 2)||(356, 15)|
|(250, 27)||(343, 6)||(84, 62)|
Rock-magnetic investigation of sedimentary rocks provides insights into the basin’s formation and sedimentation processes on an active margin. Cretaceous (Yezo Supergroup) ~ Eocene (Ishikari Group) strata and middle Miocene (Kawabata Formation) turbidites in central Hokkaido represent forearc and foreland settings, respectively. Progressive demagnetization successfully isolated characteristic remanent magnetization (ChRM) of the Kawabata Formation. Mean declination of the formation’s ChRM exhibited significant westerly deflection, suggesting counterclockwise rotation of the study area since the middle Miocene. This differs from previous reports that indicated clockwise rotation. We attribute the difference to complicated deformation around the terminations of faults that form the N-S elongate Kawabata sedimentary basin. Anisotropy of magnetic susceptibility (AMS) principal axes were clearly determined for both the Cretaceous/Paleogene samples and Neogene samples, and regarded as a proxy of sediment influx directions. Paleocurrent directions inferred from the Eocene AMS data tend to align in N-S azimuth (Figure 14), and accord with the results of sedimentological paleoenvironment reconstruction, which suggest a northward downstream trend in fluvial to tidal estuarine systems . As for the Cretaceous, further acquisition of AMS data is necessary to assess the effect of intensive syn-depositional deformation of the forearc . After correcting for the tectonic rotation, most of the paleocurrent markers in the Kawabata Formation indicated a westward current direction with minor southward flow contributions, consistent with a sedimentary model that envisions burial of the Miocene N-S foreland basin by clastics derived from the eastern collision front. The intensity of alignment forcing of sedimentary particles inferred from the shape parameter (T) of the AMS data was closely related to sedimentary facies observed in the field. In investigating the origin of the AMS fabrics of turbidite deposits of the Kawabata Formation, we conducted a re-deposition experiment of fine detrital particles with no magnetic fraction including paramagnetic minerals with relatively high magnetic susceptibility, which demonstrated the significance of the alignment of paramagnetic minerals having shape anisotropy.
The authors are grateful to N. Ishikawa for the use of the rock-magnetic laboratory at Kyoto University and for thoughtful suggestions in the course of the magnetic analyses. We thank S. Oshimbe and S. Nishizaki for their help with field work. Thanks are also due to N. Yamashita and Y. Danhara for their support in mineral separation. Constructive review comments by G. Kawakami greatly helped to improve early version of the manuscript.