IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.320321.213.2013

Results

NRM data compilation from DSDP, ODP, and IODP sites on the Pacific plate

The compiled data in this study are listed in Table T1. No data from IODP sites remain after application of the data selection criteria. Locations of sites from which data are reported are shown in Figure F2. Descriptions for each individual datum adopted in the present compilation are summarized below.

Site 169

During DSDP Leg 17, extrusive basalt, possibly with an ocean-ridge origin, was recovered from Site 169, just east of the Marshall Islands (Bass et al., 1973). Sager (2004) measured the NRM of the basalt from 10 discrete samples and estimated their ages as 148 Ma based on magnetic lineations. Because there is no information about any unit divisions (e.g., lithologic or magnetic units) for the measured samples, a single average value from 10 data points is tabulated in Table T1.

Sites 303, 304, and 307

Basaltic rocks classified as ocean-ridge-type tholeiites were taken from Sites 303, 304, and 307 during DSDP Leg 32 (Marshall, 1975). Larson and Lowrie (1975) reported NRM measurement data from five, eleven, and eight discrete samples from Holes 303A, 304, and 307, respectively. These holes are located on Mesozoic Anomalies M4 (Hole 303A), M9 (Hole 304), and M21 (Hole 307). The samples are described as slightly altered (Hole 303A), more extensively altered (Hole 304), and highly weathered (Hole 307). Average NRM values are tabulated for individual holes in Table T1.

Sites 420, 421, 422, 423, 427, 428, and 429

Paleomagnetism study of 1.5–4.6 Ma basalt samples taken from the western slope of the East Pacific Rise during DSDP Leg 54 was conducted by Petersen and Roggenthen (1980). Sites 420, 421, 422, 423, 428, and 429 comprise a transect perpendicular to the East Pacific Rise, whereas Site 427 is located ~100 km to the south (Rosendahl et al., 1980). Two to forty-two discrete samples were measured for individual holes at each site. Average NRM values are tabulated for individual holes in Table T1.

Sites 469, 470, and 472

Abyssal tholeiites were recovered off southern California and Baja California at Sites 469 (17 ± 0.5 Ma), 470 (15–16 Ma), and 472 (16 ± 0.5 Ma) during DSDP Leg 63 (Shipboard Scientific Party, 1981a, 1981b, 1981c). Geochemical evidence of the tholeiites is compatible with their formation at a spreading center (Shibata et al., 1981). Verma and Banerjee (1981) investigated magnetic properties of the tholeiites and gave NRM data from two to four discrete samples for Holes 469, 470, 470A, and 472. Average NRM values are tabulated for individual holes in Table T1.

Site 581

Sager (2004) reported NRM data from basements collected at Site 581, located north of Shatsky Rise, during DSDP Legs 86 and 88. Data were obtained from 8 to 38 discrete samples for Holes 581, 581A, 581B, and 581C. He estimated the corresponding crustal age as 124–127 Ma from seafloor magnetic lineations (Anomaly M1–M3). Geochemical study by Fountain et al. (1985) concluded that the basements are within the range of normal MORB (N-MORB). Average NRM values are tabulated for individual holes in Table T1.

Site 595

DSDP Leg 91 drilled three holes at Site 595 in the Southwest Pacific Basin, ~1000 km east of the Tonga Trench (Shipboard Scientific Parties, 1987). Three distinct types of basaltic rock were encountered in Holes 595A and 595B (Units 595A-1 and 2; Units 595B-1, 2, and 3), and all of the recovered basalts are light rare-earth-element depleted tholeiites with trace element ratios identical to N-MORB from elsewhere in the Pacific (Saunders, 1987). Montgomery and Johnson (1987) reported paleomagnetic and rock magnetic measurement results from basalts from Units 595A-1 and 2 (Hole 595A) and Units 595B-2 and 3 (Hole 595B) and stated that the basalts could be on magnetic Anomaly M29. Each unit average value is tabulated in Table T1.

Site 597

From the ocean crust of the eastern Pacific, tholeiites with the mineralogical and geochemical characteristics of N-MORB were taken from Site 597 during DSDP Leg 92 (Pearce et al., 1986). The crustal age of the site is estimated to be 27.5 Ma, and paleomagnetic measurements were conducted on discrete samples from three lithologic units (1–3) (Nishitani, 1986; Sager, 2004). Sager (2004) grouped the resultant paleomagnetic directions into eight groups, but the eighth group consists of results from both Units 2 and 3. In Table T1, average NRM values are tabulated for nine groups, since it is probably safe to further divide the eighth group into two subgroups (Units 2-2 and 3).

Site 801

A composite 131 m basaltic basement section was cored at Site 801 (ODP Leg 129) in the western Pacific Pigafetta Basin, a lower tholeiitic basalt sequence of which had typical N-MORB features (Floyd and Castillo, 1992). A whole-rock 40Ar/39Ar age of 166.8 ± 4.5 Ma was obtained for the lower basalt sequence, consistent with a prediction by simple linear extrapolation of the M-sequence magnetic lineation ages (Pringle, 1992). Paleomagnetic and rock magnetic properties of the basalt were examined on discrete samples taken intermittently from Flows 2 (top) to 32 (bottom) by Wallick and Steiner (1992). Average NRM values are tabulated in Table T1 for individual flows below Flow 13.

Site 843

ODP Leg 136 penetrated 70 m of basaltic crust ranging in composition from N-MORB to enriched MORB (E-MORB) at Site 843 in the central Pacific Ocean, and magnetism of the basalts were investigated on 10 discrete samples (Johnson and Pariso, 1993a). 40Ar/39Ar dating on a relatively unaltered specimen resulted in a crystallization age of 110 ± 2 Ma (Waggoner, 1993). An average NRM value is tabulated for Hole 843B in Table T1.

Site 864

Igneous rocks were drilled at Site 864 within the axial summit depression of the East Pacific Rise during ODP Leg 142 (Shipboard Scientific Party, 1993). Two lithologic units (1 and 2) were identified, and they are typical of N-MORB. NRM data were obtained from discrete samples taken from both units (Shipboard Scientific Party, 1993). Each unit average value is tabulated in Table T1.

Site 1149

At Site 1149, aphyric basalts were recovered from the ocean floor which should correlate to a basement age of ~132 Ma (Shipboard Scientific Party, 2000). The basalts are comparable to those in N-MORB or E-MORB, and remanence measurements were done with split half pieces of basalts during ODP Leg 185 (Shipboard Scientific Party, 2000). Remanence data are found in the ODP database (iodp.tamu.edu/database/) and tabulated in Table T1 for data measured from pieces >20 cm long (because the width of the response curve for the superconducting rock magnetometer was ~20 cm).

Site 1179

Sager and Horner-Johnson (2004) reported paleomagnetic measurement results from basement rocks (~129 Ma) collected from Site 1179 during ODP Leg 191. These basement rocks consist mostly of fresh aphyric basalts and are divided into 48 units, though paleomagnetism results were reported for 19 magnetic units. They are geochemically characterized as tholeiitic basalt with MORB affinity (Shipboard Scientific Party, 2001). Average NRM values are tabulated for individual units in Table T1.

Site 1224

Basalts ranging from differentiated to very differentiated N-MORB were cored from four holes in ~46.3 Ma crust at Site 1224 (ODP Leg 200; Shipboard Scientific Party, 2003a). Fourteen cooling units were identified, and paleomagnetic measurements were conducted on ~60 discrete samples, when hyaloclastite are excluded (Shipboard Scientific Party, 2003a). Average NRM values are tabulated for individual holes and units in Table T1.

Site 1243

During ODP Leg 203, basement material was recovered from the eastern equatorial Pacific at Site 1243 (Shipboard Scientific Party, 2003b). Eight basement units were defined, and their ages were estimated at 10–12 Ma. Moberly et al. (2006) concluded that the basement units had MORB geochemistry. Paleomagnetic measurement results were reported for the six units and later discussed in Zhao et al. (2006). Average NRM values are tabulated for individual units in Table T1.

Difference from the previous data set

It is thought that the newly compiled data set should be a superset of the “Pacific” data set by Johnson and Pariso (1993b): all Pacific data compiled in Johnson and Pariso (1993b) should be included in the present compilation. However, close inspection of original references turned out that many of the Pacific data compiled in Johnson and Pariso (1993b) do not meet the present criteria. Individual rationales are as follows.

  • At Site 163, there are no numerical NRM data in the report by Heinrichs (1973).

  • At Site 183, the recovered alkali-olivine basalts resembled Hawaiian alkali basalts, suggesting that the basalts are not mid-ocean-ridge origin (Stewart et al., 1973).

  • Sites 319, 320, 321, and 424 are at the Nazca plate (Yeats and Hart, 1976; Rosendahl et al., 1980).

  • Sites 425, 482, 483, 485, 487, 501, 504, 505, 506, 507, 508, and 510 are at the Cocos plate (Rosendahl et al., 1980; Shipboard Scientific Party, 1982, 1983; Langseth et al., 1983; Honnorez et al., 1983).

  • Site 462 is at the Nauru Basin and 40Ar/39Ar age of the recovered basalt shows much younger age of 110 Ma compared with the corresponding magnetic Anomaly M26 (Ozima et al., 1981). This age implies that Site 462 basalts are not mid-ocean-ridge origin.

  • At Site 471, all NRM data are from diabase samples (Verma and Banerjee, 1981).

  • At Site 601, no NRM data are reported in the initial report from DSDP Leg 92 (Leinen, Rea, et al., 1986).

Rock magnetic and paleomagnetic measurements on Expedition 320/321 basement rocks

Rock magnetic property: NRM and ARM intensity

Intensities of initial NRM and laboratory-imparted ARM are listed in Table T2 for individual cobbles or cores. NRMs range between 6.83 × 10–5 and 1.83 × 10–3 Am2/kg, and ARMs are between 2.70 × 10–4 and 1.14 × 10–3 Am2/kg. Assuming a density of 3 g/cm3, NRMs and ARMs respectively correspond to 0.205–5.49 A/m and 0.810–3.41 A/m. Interestingly, cobbles from Site U1332 show strong NRMs that are almost one order of magnitude higher than the others, except the core from Site U1337 (Table T2). In contrast, differences in ARM intensities are not large (about a factor of two) compared with those in NRM intensities (Table T2).

Rock magnetic property: hysteresis parameters and FORC measurements

Table T3 shows the resultant hysteresis parameters for selected individual cobbles or cores. The measured parameters are remanent coercivity (Brc), coercivity (Bc), remanent saturation magnetization (Mrs), and saturation magnetization (Ms). Regardless of the cobbles and cores, hysteresis ratios Brc/Bc and Mrs/Ms are very similar and close to the “threshold” value for single domain (SD) grains (Brc/Bc = 1.5 for titanomagnetites [Day et al., 1977]; Mrs/Ms = 0.5 for randomly oriented SD grains with uniaxial anisotropy). It seems that submarine basalts are very rapidly emplaced (quenched) and thus contain magnetic grains that are tiny. The resultant hysteresis ratios are consistent with this expectation. This is also consistent with FORC diagrams obtained from six selected specimens (Fig. F3): SD-like contribution is clearly recognized as elongated jetlike distribution along the horizontal axis (Hu = 0 [Hu is an interacting field]), whereas MD-like contribution appears almost absent.

A noticeable result in the hysteresis measurements is that Mrs values range between 2.704 and 8.703 × 10–2 Am2/kg. If we assume a density of 3 g/cm3, they correspond to 8.11–26.1 A/m. These values are at least one order of magnitude larger than the initial NRM intensities (Table T2). Mrs is equivalent to saturation isothermal remanent magnetization (SIRM). The weak NRMs compared to SIRMs imply that NRMs are not affected by large drilling-induced isothermal remanent magnetization (IRM).

Rock magnetic property: thermomagnetic property

Thermomagnetic curves measured on chip samples from all individual cobbles and cores can be classified into three types (A, B, and C). Representative curves are illustrated in Figure F4. Classified types are indicated in Table T2.

A common feature among the three types is that induced magnetizations rapidly decrease at low temperature (below approximately –150°C). This is probably due to Ti-rich titanomagnetite or titanohematite phases, though they are not considered to be contributed to any NRM. A second common feature is that cooling curves are always atop heating curves except intervals at very high temperatures. The cooling curves indicate an almost single phase of Tc (Curie temperature) at ~580°C, suggesting that Ti-poor titanomagnetite is produced during heating. This is probably caused by high-temperature oxidation of Ti-rich titanomagnetite or decomposition of (titano)maghemite. The greater magnetization of the cooling curves is more evident in the vacuum run. The reason for this is considered as follows: a thermal run in vacuum preserves the produced Ti-poor titanomagnetite, whereas that run in air perhaps further oxidizes it into lower magnetization hematite. A third common feature is that there is a minor phase of Tc > 400°C suggested from heating curves. This phase is possibly of primary origin and can hold primary remanence, although a part of this may be related to the Ti-poor titanomagnetite production during heating.

The Type A curve (Fig. F4A) above room temperature is characterized by an almost single phase of Tc at ~400°C in air heating. No bump is recognized above 400°C. In contrast, there is a subtle bump in vacuum heating, suggesting an inversion of (titano)maghemite. In the Type B curve (Fig. F4B) above room temperature, except for the main phase of Tc at ~350–400°C, a bump exists above 400°C even in air heating. This tendency is more exaggerated in vacuum heating. (Titano)maghemite content is obviously higher in Type B than in Type A. Behavior of the Type C curve (Fig. F4C) is analogous to that of Type A, but there is a local minimum at approximately –150°C and induced magnetization increases by ~140°C. Similar characteristics were observed in oceanic basalts recovered from the summit region of the Detroit Seamount, and these characteristics were considered to result from broad “N-type” thermomagnetic curves originating from low-temperature oxidation (Doubrovine and Tarduno, 2004). Although N-type possibly causes (partial) self-reversal in NRM (Dunlop and Özdemir, 1997), the Detroit Seamount samples showed single NRM components in thermal demagnetization results (Doubrovine and Tarduno, 2004).

Observations by FE-SEM

Figure F5 illustrates backscattered electron images of two specimens, one from interval 320-U1332B-18X-CC, 27–33 cm, and the other from interval 321-U1337C-33X-4, 0–47 cm. In the images, pore space appears as black regions. When the lowest magnification images are compared between the two specimens, the density of pore space is obviously lower in the Section 320-U1332B-18X-CC specimen, and Fe-Ti oxides appear to be well preserved (without [titano]maghemitization). On the other hand, in the Section 321-U1337C-33X-4 specimen, there is more pore space and Fe-Ti oxides are often associated with shrinkage cracks (e.g., the lower left image in Fig. F5). Such cracks are typical indications of (titano)maghemitization, and the pore space may be originated from long-term low-temperature oxidation.

Thermomagnetic curves of the two specimens correspond to Type A (Fig. F4A) for the Section 320-U1332B-18X-CC specimen and to Type B (Fig. F4B) for the Section 321-U1337C-33X-4 specimen (Table T2). Because the content of (titano)maghemite is considered to be higher in Type B than in Type A (see “Rock magnetic property: thermomagnetic property”), the present observations are consistent with the thermomagnetism experimental results.

Paleomagnetic inclination

For the cobble from Site U1333 (interval 320-U1333B-20X-CC, 18–24 cm), demagnetization results were obtained by AFD and ThD. AFD results show relatively simple behavior—that is, secondary components are removed by 18 or 24 mT, after which the remanences decay linearly toward the origin (Fig. F6A). Principal component analysis (PCA) by Kirschvink (1980) yielded inclinations of –19.2° and –11.5°. ThD results show complex behavior, but remanences carried by blocking temperatures (TB) higher than 325° or 340°C are considered to be primary (Fig. F6B). They give PCA inclinations of –16.2° and –20.4°, which are consistent with those obtained by AFD. Overall, average inclination of the cobble is calculated to be –16.9° (α95 = 5.5°, N = 4) by the inclination-only statistics of Arason and Levi (2010).

For Sample 321-U1337C-33X-4, 0–47 cm, we obtained 12 AFD and 6 ThD results. The ThD results show self-reversal behavior: remanence components with TB < 325°–375°C have antiparallel directions against those with TB > 325°–375°C (Fig. F6C). Components with TB > 325°–375°C are considered to be primary, and their PCA inclinations range between 11.3° and 37.8°. Thus, ThD results give positive primary inclinations. On the other hand, 7 out of 12 AFD results show relatively simple demagnetizing behavior: remanences are straightly demagnetized toward the origin of the orthogonal vector plot after removal of some secondary components (Fig. F6D). PCA inclinations calculated for these primary components range between –14.2° and –28.8°, which are negative and antiparallel with the primary components determined by ThD. Thus, AFD in these seven results is not considered to resolve the two components revealed by ThD. We reject the seven AFD results and adopt the remaining five AFD results (Fig. F6E), which yield PCA inclinations between 12.6° and 29.5° from high-coercivity (Hc) intervals. These inclinations are positive and consistent with the primary components determined in ThD. An average inclination of Core 321-U1337C-33X from the ThD and AFD results is given as 22.7° (α95 = 6.5°, N = 11) by the inclination-only statistics of Arason and Levi (2010).

Four AFD and three ThD results were obtained for the Sample 321-U1337D-49X-3, 18–39 cm. In contrast to the demagnetization results of Core 321-U1337C-33X, ThD and AFD results give concordant PCA inclinations from high TB (>300°–375°C) or Hc (>12–14 mT) intervals (Fig. F6F, F6G). These inclinations range between 12.6° and 22.6° and yield an average inclination for Hole U1337D of 16.7° (α95 = 3.4°, N = 7) by the inclination-only statistics of Arason and Levi (2010).

From the three average inclinations, the apparent paleolatitudes are calculated to be 8.6°S for Site U1333 and 11.8°S and 8.5°S for Site U1337. The three average inclinations are results of spot reading of the ancient geomagnetic field. Considering paleosecular variation, the estimated paleolatitudes are consistent with the expected paleoequatorial positions of Sites U1333 and U1337. These imply that NRMs of the Sites U1333 and U1337 samples are not affected by large drilling overprints.

Summary

NRMs range between 0.205 and 5.49 A/m, and ARMs are between 0.810 and 3.41 A/m. The cobbles from Site U1332 (49–50 Ma) yield strong NRM values that are almost one order of magnitude higher than other samples, except the core from Site U1337.

Hysteresis properties show ratios of Brc/Bc and Mrs/Ms that are very similar and close to the threshold value for SD grains. The ratios are consistent with FORC diagrams obtained from six selected specimens. Mrs values range between 8.11 and 26.1 A/m, and they are at least one order of magnitude larger than the initial NRM intensities. Thus, NRMs are not considered to be affected by large drilling-induced IRM.

Thermomagnetic curves can be classified into three types (A, B, and C). (Titano)maghemite content, namely degree of low-temperature oxidation, is higher in Types B and C than in Type A. Observations by FE-SEM reveal that density of pore space is obviously lower in the Type A specimen (Sample 320-U1332B-18X-CC, 27–33 cm), and Fe-Ti oxides appeared to be well preserved (without [titano]maghemitization). In contrast, in the Type B specimen (Sample 321-U1337C-33X-4, 0–47 cm), there is a lot of pore space and Fe-Ti oxides are often associated with shrinkage cracks which are typical indications of (titano)maghemitization.

Average paleomagnetic inclinations result in –16.9° (α95 = 5.5°, N = 4) for Hole U1333B, 22.7° (α95 = 6.5°, N = 11) for Hole U1337C, and 16.7° (α95 = 3.4°, N = 7) for Hole U1337D. Inclinations correspond to apparent paleolatitudes of 8.6°S for Site U1333 and 11.8°S and 8.5°S for Site U1337. Considering paleosecular variation, the estimated paleolatitudes are consistent with the expected paleoequatorial positions of Sites U1333 and U1337. These imply that NRMs for Site U1333 and U1337 samples are not affected by large drilling overprints.

Temporal variation in igneous basement rock NRM observed for the Pacific plate

Compiled NRM data from previous DSDP and ODP sites at the Pacific plate (Table T1) and newly obtained NRM data in the present study (Table T2) are shown versus age in Figure F7. For the latter data set, we assume a density of 3 g/cm3 to convert units of magnetization. The geomagnetic field is mainly dipolar and its strength has been shown to exhibit latitudinal dependence by a factor of two; Johnson and Pariso (1993b) treat similar data by normalizing NRM to a common latitude. We have not done such normalization in Figure F7. The reasons are as follows:

  • Paleomagnetism studies have revealed that geomagnetic field strength varies with time almost an order of magnitude even within a time period of ~100 k.y. (e.g., Valet et al., 2005; Yamazaki and Oda, 2005; Biggin et al., 2009). Errors of age estimates for compiled and present data are beyond 100 k.y.

  • The geomagnetic field consists not only of the dipole component but also of the nondipole component. Both the dipole and nondipole components are time varying. The time varying nondipole components can perturb the latitudinal dependence. NRMs of compiled and present data are probably more or less affected by the nondipole components at the timing of initial remanence acquisitions.

  • Because of the previous reasons, any normalizations of NRM can cause possible artifacts that may obscure the “real” feature.

NRM data at 46.3 Ma in Figure F7 are from Site 1224. They range between 0.0961 and 28.63 A/m, and the remanence values at the upper end seem high. Actually, 6 out of 14 individual data show NRMs >10 A/m, and some samples may be perturbed from drilling overprint. Shipboard Scientific Party (2003a) described the drilling overprint being removed by ~10 mT AFD for Site 1224 samples, though it was much less severe than vitric tuffs recovered at Site 1223. Two data points similar in age to the Site 1224 data points give a relatively high NRM value of ~3–5 A/m at ~50 Ma (Fig. F7). These data are from the Site U1332 samples and are not considered to be affected by large drilling overprint (see “Rock magnetic property: hysteresis parameters and FORC measurements”).