alteration on age determinations

alteration on age determinations

Chemical Geology 287 (2011) 41–53 Contents lists available at ScienceDirect Chemical Geology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. ...

4MB Sizes 9 Downloads 7 Views

Recommend Documents

No documents
Chemical Geology 287 (2011) 41–53

Contents lists available at ScienceDirect

Chemical Geology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c h e m g e o

Unspiked K–Ar dating of Koolau lavas, Hawaii: Evaluation of the influence of weathering/alteration on age determinations Seiko Yamasaki a,⁎, Ryotaro Sawada a, Ayako Ozawa a, Takahiro Tagami a, Yumiko Watanabe a, Eiichi Takahashi b a b

Division of Earth and Planetary Sciences, Kyoto University, Kyoto 606–8502, Japan Earth and Planetary Sciences, Tokyo Institute of Technology, Tokyo 152–8551, Japan

a r t i c l e

i n f o

Article history: Received 1 October 2010 Received in revised form 29 April 2011 Accepted 2 May 2011 Available online 14 May 2011 Keywords: K–Ar dating Weathering Alteration Basalt Koolau volcano Hawaii

a b s t r a c t In order to evaluate the influence of weathering/alteration on K–Ar dating for Hawaiian tholeiites, unspiked K–Ar ages were determined for 21 samples from four lava flows with varying degrees of weathering or alteration collected from the Makapuu Head section of Koolau volcano, Hawaii. The samples were classified based on freshness of olivine phenocrysts and the groundmass olivine, and the presence of secondary minerals in vesicles. The results indicate that the ages for samples with fresh groundmass olivine are reliable, even though olivine phenocrysts may be slightly altered (thin reaction rims) or secondary minerals may have crystallized in the vesicles. The ages for the lowermost lava flow in the Makapuu Head section and the lava flow approximately 120 m above it are 2.58 ± 0.13 and 2.36 ± 0.09 Ma, respectively. The accumulation rate of this section is calculated to be 0.04–0.11 cm/year. We also report K–Ar ages for lava samples collected from the submarine flank of the Koolau volcano and the Nuuanu landslide blocks. The age for an early Makapuu-stage lava collected from the submarine flank of Koolau volcano is 2.5 Ma, similar to the age from the lower part of the subaerial Makapuu Head section. Another lava sample collected from the submarine flank of the Koolau volcano has an age of 3.3 Ma, older than any subaerial part of Koolau volcano. These results suggest that the onset of Koolau's shield-stage volcanism was no later than ~ 3.3 Ma, and the duration of the shield stage was at least 1.2 m.y. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Accurate and precise determination of volcanic rock ages is essential for understanding the history of Hawaiian volcanoes and the structure of the Hawaiian plume. Unspiked K–Ar dating has proven to be very successful for determining crystallization ages for many Hawaiian lavas (e.g. Guillou et al., 1997, 2000; Ozawa et al., 2005, 2006; Sherrod et al., 2007). To obtain reliable (nonspurious) K–Ar ages, the lava samples need to meet various requirements and lack of K and Ar loss during weathering or alteration is one of the most important considerations. Therefore, for reliable K–Ar dating, it is desirable to choose fresh rock samples that have not been affected by weathering/alteration; however, such samples are generally not available among the tholeiitic lava of shields older than about 1 Ma.

⁎ Corresponding author at: Tono Geoscientific Research Unit, Geological Isolation Research and Development Directorate, Japan Atomic Energy Agency, 959 31 Jorinji, Izuimi-cho, Toki, Gifu 509 5102, Japan. E-mail address: [email protected] (S. Yamasaki). 0009-2541/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.chemgeo.2011.05.003

In previous studies, the alteration of samples was evaluated by thin-section observations (e.g. Ozawa et al., 2005, 2006; Sano, 2006). However, the criteria for sample selection for K–Ar dating have not been examined closely. Therefore, the goals of this study are to evaluate the effects of alteration on tholeiitic lavas relevant to K–Ar dating and to establish criteria for sample selection. We measured unspiked K–Ar ages for 21 lava samples with varying degrees of weathering/alteration, collected from four flows in the Makapuu Head section of Koolau volcano, where about 200 m of tholeiitic basalt lava flows is exposed. The Makapuu Head section is an ideal for this study because its well-exposed strata have been studied in detail (Frey et al., 1994) and a reliable age was determined previously using fresh samples from one of the flows (Ozawa et al., 2005, their sample HVO211). We also determined K–Ar ages for lava samples collected from the submarine flank of the Koolau volcano and the Nuuanu landslide blocks. The duration for the shield stage of Hawaiian volcanoes has been estimated to be ~0.7 Ma (Clague and Dixon, 2000). However, the age of the submarine lava erupted in the early shield stage has not yet been clearly determined. The ages of the submarine lavas in the Nuuanu landslide blocks erupted during the main shield stage of

42

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53

Koolau volcano provide important information for understanding the duration of the shield stage volcanism of Hawaiian volcanoes. 2. Geological setting and analyzed samples 2.1. Koolau volcano Koolau volcano, on the island of Oahu, consists of shield-stage tholeiitic basalts (1.8–3.2 Ma; McDougall, 1964; Doell and Dalrymple, 1973; Laj et al., 2000; Haskins and Garcia, 2004; Ozawa et al., 2005; Herrero-Bervera et al., 2007) overlain by rejuvenated-stage alkalic basalts of the Honolulu Volcanics (0.8–0.03 Ma; Gramlich et al., 1971; Lanphere and Dalrymple, 1980; Ozawa et al., 2005). Much of the eastern half of Koolau volcano collapsed into the ocean (Fig. 1(a); Nuuanu landslide; Moore et al., 1994); consequently, the Koolau caldera is positioned on the east coast of the island. Two rift zones extend to the northwest and east–southeast from the caldera. The shield-stage lavas were divided into two parts on the basis of geochemistry: main shield stage and late-shield or Makapuu-stage (Shinozaki et al., 2002). Main shield-stage lava has been found in Nuuanu landslide blocks (Shinozaki et al., 2002) but not in the subaerial Koolau shield. Makapuu-stage lava, which is exposed in an approximately 800 m-thick section at the top of the shield (Takahashi and Nakajima, 2002), has a volume in the order of 500–1000 km 3 (Shinozaki et al., 2002), only a small part of the 34,000 km 3 that forms the Koolau volcano (Robinson and Eakins, 2006). Among the Hawaiian lavas, the Makapuu-stage lavas have the highest SiO2 and 87Sr/ 86Sr ratios and the lowest 143Nd/ 144Nd and 206Pb/ 204Pb ratios (e.g. Stille et al., 1986, Frey et al., 1994, Tanaka et al., 2002). In contrast, the main shield-stage lavas resemble Loa- or Kea-type compositions; they are characterized by lower 87Sr/ 86Sr and higher 143Nd/ 144Nd and 206 Pb/ 204Pb ratios than those of the Makapuu-stage lavas (Tanaka et al., 2002).

sediment to have been deposited between 2.5 and 3.3 Ma (Morgan et al., 2006). In order to investigate the structure of the sea floor of Koolau volcano and the Nuuanu landslide, seven dives and four dredges were conducted during the joint Japan-USA cruises, sponsored by the Japan Agency for Marine-Earth Science and Technology (JAMSTEC) in 1998 and 1999. Oahu North-a site is a shallower part (2000–3000 m) of the northeastern flank of Oahu (Fig. 1(a)), likely undisturbed by the Nuuanu landslide (Shinozaki et al., 2002). Subaerial or shallow submarine pillow lavas and hyaloclastite erupted during the Makapuu-stage occur on this site, whereas hyaloclastites are dominant in the deeper sites. This frank, which displays a rock-type transition similar to that of the Hilina slump on the south flank of Kilauea volcano (Lipman et al., 2002), is interpreted as a typical Hawaiian volcano flank that has undergone a volcano-tectonic growth cycle (Shinozaki et al., 2002). Sample S500-8 collected from the Oahu North-a site on dive S500 is from an in-place pillow lava, while sample S500-7 is from displaced lava that is probably a locally derived from submarine Koolau. Nuuanu blocks are landslide deposits derived from the main shield-stage of Koolau (Shinozaki et al., 2002). Dive S498 ascended the southeast wall of Tuscaloosa Seamount and non-vesicular olivine basalt (S498-4) was sampled from a rocky outcrop on the dive at the Nuuanu-2 site (Fig. 1(a)). The Nuuanu-3 site investigated on dive S499 (Fig. 1(a)), is along the southwestern slope of an unnamed hummock-shaped block and shallow submarine lava breccia (S499-6) was collected from this site. The petrological and geochemical signatures of the samples collected from the Oahu North-a site are similar to those of the subaerial Makapuu stage lavas. In contrast, the signatures of the Nuuanu landslide lavas are similar to those of Mauna Loa and Mauna Kea lavas on Hawaii (Shinozaki et al., 2002; Tanaka et al., 2002). 3. Analytical procedures

2.2. Makapuu head section 3.1. Classification for weathering/alteration of the lava samples The samples for evaluation of weathering/alteration for K–Ar dating studies were collected from the north side of the well-exposed and accessible Makapuu Head section investigated by Frey et al. (1994) (Figs. 1(a), (b), and 2). A schematic stratigraphic column of the section (M.O. Garcia, pers. comm.) and the sampling locations are presented in Figs. 1(b) and 2 and listed in Table 3(a). The ratio of pahoehoe to aa flows is roughly 40:60 for the entire 200-m-thick section. In the tropical environment of Hawaii, weathering can decrease K2O and SiO2, and K2O/P2O5 can be a good indicator of alteration (e.g. Frey et al., 1994). Frey et al. (1994) analyzed major element compositions for the Koolau shield lavas and reported that 20 of the 67 samples they analyzed had low K2O/P2O5 ratios (less than 1.0). They attributed the low ratios in the Koolau shield lavas to lowtemperature alteration (Fig. 3). For this study, 21 lava samples with varying degrees of alteration were collected from four flows (KOO-03, KOO-07, KOO-19, KOO-23; Fig. 2), which, for the most part, have relatively higher K2O/P2O5 ratios than those of the study by Frey et al. (1994). KOO-07 is a pahoehoe flow and the others are aa flows. Since the outcrops at Makapuu Head are along the coast, the lava flows there may have been affected by seawater as well as terrestrial water. 2.3. Submarine samples The Nuuanu landslide is one of the largest landslides on the Hawaiian Ridge. Debris extends more than 250 km out from the northeast coast of Oahu and contains giant blocks. The largest block, the Tuscaloosa Seamount is 20 by 30 km in size. The micro fossils in the mudstone collected from Tuscaloosa Seamount constrain the

The lava samples collected from the Makapuu Head section were studied with the naked eye, with a hand lens and in thin sections and have been classified into seven categories based on: (1) freshness of olivine phenocrysts, (2) presence of secondary minerals in vesicles and (3) freshness of groundmass olivine. The criteria for grouping the lavas into categories based on the degree of weathering/alteration are presented in matrix form in Table 1. The presence of secondary minerals in vesicles is dependent on the development of vesicles and the chemistry of any aqueous fluid that invaded the lavas during the weathering/alteration process. Therefore the presence of secondary mineralization in vesicles is valid only for samples with vesicles, and may be supplementary criterion for the evaluation of weathering/ alteration. In the six Makapuu Head samples classified in the B1 and C1 groups, there are no secondary minerals although vesicles are present. In terms of the presence of secondary minerals, one submarine sample was indeterminate because of the absence of vesicles. In previous studies, fairly fresh samples that would qualify as A1 and B1 groups in this classification were used for K–Ar dating (Ozawa et al., 2005; Sano, 2006). 3.2. Whole rock major-element analysis Whole rock major element compositions for the Makapuu Head samples were determined to estimate the degree of alteration (Table. 2). Sample preparation techniques and analytical procedures are based on those of Goto and Tatsumi (1994, 1996). To make fused glass beads, lithium tetraborate (Li2B4O7) was used as a flux. 0.40000 ± 0.00015 g of powdered sample and 4.00000 ± 0.00015 g of flux were weighed directly in a platinum crucible. Fusion and agitation

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53

43

(a) 160°W

158°W

156°W

Kauai

22°N

Oahu Molokai Maui Lanai Kahoolawe

Nuuanu-3 NuuanuS499

20°N km 0

Hawaii

Nuuanu-2 NuuanuS498

60

Nuuanu Slide

22°N

Oahu North-a S500 Oahu

Ko

Wailau Slide

ai

an

ge

an

ae

uR

ola

W

Makapuu Makapu Head Hea Molokai

km 0

50

21°N 158°W

157°W

(b) 539,000

0

540,000

0.5

1 km

Contour Interval 40 feet -23 KOO-03 -07 -19 2,357,000

2,356,000

Fig. 1. (a) Shaded relief map showing the locations of the dive sites and Makapuu Head. Dashed lines outline the debris fields for the giant Nuuanu and Wailau landslides from Koolau and East Molokai (Moore and Clague, 2002). Inset map shows the study area relative to the Hawaiian Islands and triangles are the locations of each volcano. (b) The detailed Makapuu Head sample location map. The topographic base map is from USGS Koko Head, 1999. Transverse Mercator projection. UTM grid in North American datum 1983, Zone 4.

was automatically carried out with a high frequency bead sampler. For measurements, we used Rigaku Simultix-3550 X-Ray fluorescence (XRF) spectrometry at Kyoto University. Accuracy for each element is better than 3.5%.

3.3. K-Ar dating About 80–100 g of rocks was crushed using a stainless steel pestle and then sieved to 250–500 μm. Crushed samples were cleaned with

44

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53 Altitude (m)

aa lava pahoehoe lava dike pahoehoe/scoria

131.5

KOO23 2.36±0.06 Ma (weighted mean age) (Samples MKPH-01, 02, 09, 10, 11, 12, 13, and HV02-11; Ozawa et al., 2005)

91.4

KOO-19 (MKPH-03, 04, 14, 15, 16, 17)

20.1

KOO-07 2.47±0.09 Ma (weighted mean age) (MKPH-18, 19, 20, 21) KOO-03 2.58±0.12 Ma (weighted mean age) (MKPH-05, 06, 07, 08)

4.9 0

Fig. 2. Schematic stratigraphic column of Makapuu Head section (M.O. Garcia, pers. comm.) and sampling sites.

deionized water and acetone in an ultrasonic bath. Phenocrysts were carefully removed using a Frantz isodynamic separator to minimize extraneous 40Ar (Ozawa et al., 2006) and groundmass samples were used for the analyses. An aliquot of each sample was ground further for potassium analysis. Potassium concentrations were measured using an Asahi Rika FP-33D flame emission photometer with an internal lithium standard and a peak integration method. Techniques were similar to those described by Matsumoto (1989). About 100 mg of powdered samples was dissolved in a mixture of HF and HClO4. Once dried, they were re-dissolved in HCl, diluted, and added lithium

2.5

2.0

young Kilauea Frey et al. (1994) this study

K2O/P2O5

fresh 1.5

1.0

altered

0.5

0.0 46

48

50

52

54

SiO2 Fig. 3. K2O/P2O5 vs SiO2 for the samples for Makapuu Head (this study), Koolau shield tholeiites (Frey et al., 1994), and young Kilauea tholeiites erupted over the past 25 years (Thornber et al., 2003).

standard solution as internal standard to sample solution. Dilution rate of the sample solution was determined by a weighing method instead of a volumetric method. Each sample was measured twice. Analytical error for potassium measurement is 1%, estimated from standard deviation of multiple analyses of standards JB-3 and JA-2, issued by the National Institute of Advanced Industrial Science and Technology, Japan. Because the samples dated in this study were expected to be relatively young and thus to contain a rather large proportion of atmospheric argon, argon isotopic measurements were performed using the unspiked method, also known as the peak height comparison method (Matsumoto et al., 1989; Sudo et al., 1996), and a mass fractionation correction procedure was applied to estimate the appropriate initial 40Ar/ 36Ar ratio (Matsumoto and Kobayashi, 1995) (See Appendix II for an explanation of this method and of the correction procedure). We used the VG Isotech VG3600 mass spectrometer operated in static mode and connected to extraction and purification lines. Samples wrapped in copper foil were melted in molybdenum curucible at 1500 °C to extract gases. Gases were purified with a Ti–Zr getter at ~700 °C, two SAES getter at 200 °C and room temperature. The extraction and analysis system, as well as the detailed techniques used for argon isotopic measurements, are described by Sudo et al. (1996). K–Ar ages were calculated using the isotopic abundances and decay constants for 40K recommended by the IUGS Subcommission on Geochronology (Steiger and Jäger, 1977; 40K/ K = 1.167 × 10 −4 , λ e = 0.581 × 10 −10 year −1 , λ β = 4.962 × 10 −10 year −1). All analyses were conducted at the Kyoto University geochronology laboratory.

4. Results and discussion 4.1. Evaluation of the influence of weathering/alteration on K–Ar dating The age data for samples from four lava flows with varying degrees of weathering or alteration are evaluated where possible by comparison with the age of equivalent fresh sample (Ozawa et al., 2005), and by means of K2O/P2O5 ratios, 36Ar volumes, and calculated atmospheric Ar contamination (See Appendix II). Fig. 3 shows K2O/ P2O5 vs SiO2 for the samples for Makapuu Head compared to the data from Koolau shield tholeiites (Frey et al., 1994) and from Kilauea tholeiites younger than 25 years of age (Thornber et al., 2003). Rhodes (1996) observed that the K2O/P2O5 ratios of altered lavas are typically less than 1.4 and are often below 1.0 based on Mauna Loa lavas recovered by the Hawaii Scientific Drilling Project (HSDP). Young fresh lavas from Kilauea have K2O/P2O5 ratios that range from about 1.25 to 1.9 (Thornber et al., 2003; Fig. 3). The concentration of P2O5 in the Kilauea samples varies little, ranging from 0.23 to 0.41 (723 data of 795 data are in the range from 0.23 to 0.29), whereas that of K2O ranges from 0.38 to 0.73. Since P2O5 can be difficult to analyze precisely, and is present in low concentrations, analytical artifacts could potentially affect the K2O/P2O5 ratios greatly. Despite this potential problem, the K2O/P2O5 values are limited to the range 1.25–1.9 in the freshest Kilauea tholeiites. The range of this ratio in the Koolau shield tholeiites is remarkably wide compared to that of the young Kilauea tholeiites, and the samples with lower K2O/P2O5 ratios have lower SiO2 (Fig. 3). Three out of 21 of Makapuu Head samples have significantly lower K2O/P2O5 ratios and SiO2 values, suggesting K loss from these samples. Atmospheric 40Ar contamination (%) can increase with loss of radiogenic Ar and replacement with atmospheric Ar during weathering. However, the portion of atmospheric 40Ar is also high in young samples, so this indicator may be valid only when comparing samples of similar age. On the other hand, 36Ar volume is considered to be proportional to an absolute volume of atmospheric contamination and the gross Ar concentration in the samples, and thus may be a good indicator of alteration. Finally, we note that higher

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53

45

Table 1 The alteration criteria for K–Ar age determination and classification of Makapuu Head samples. The photographs for each sample are shown in Appendix I. Criteria

Classification of Makapuu Head samples

Category

Freshness of olivine phenocrysts

Presence of secondary minerals

Freshness of groundmass olivines

A1 A2 B1 B2 C1 C2 D

Fresh Fresh Thin alteration rims Thin alteration rims Oxidation Oxidation Oxidation

No Formed in No Formed in No Formed in Formed in

Fresh Fresh Fresh Fresh Fresh Fresh Oxidation

vesicles vesicles vesicles vesicles

KOO-03 flow (aa lava)

KOO-07 flow (pahoehoe lava)

KOO-19 flow (aa lava)

KOO-23 flow (aa lava)

MKPH-18, 19 MKPH-05, 06, 08

MKPH-01, 02 MKPH-03

MKPH-07 MKPH-20, 21

atmospheric contamination increases the uncertainty of the calculated age. The unspiked K–Ar ages for lavas from the Makapuu Head section range from 1.42 to 2.95 Ma (Table 3(a), Fig. 4). The results for the B1-group lavas collected from flow KOO-23 are 2.39 ± 0.24 and 2.31 ± 0.13 Ma (all errors are 2σ, Table 4) for MKPH-01 and 02, respectively, and agree well with the previously reported age of 2.39 ± 0.12 Ma for this flow (Ozawa et al., 2005). The best age estimate for this flow is the weighted mean for these three ages, 2.36 ± 0.08 Ma. The D-group lavas collected from the same flow have high K2O/P2O5 ratios between 1.75 and 1.98 and show no evidence of K loss. The 36Ar volume and the percentage of radiogenic 40Ar for these D-group samples are similar to this flow, i.e., within ages agree well with the best age estimates for this flow, i.e., within 2σ errors (Fig. 4). The ages for all the samples collected from the KOO-19 flow have large uncertainties (Fig. 4). In this flow, the age for the B2-group

MKPH-04, 14, 15 MKPH-16, 17

MKPH-09, 10, 11, 12, 13

sample MKPH-03 is 2.83 ± 0.44 Ma. The K2O/P2O5 ratio of 1.30 for this sample is slightly low but in the range for the unaltered Hawaiian lavas. A slightly large 36Ar volume of 1.31 × 10 −9 cm 3 STP/g may cause large errors in the age corrected for mass fractionation. Sano (2006) rejected age data with 36Ar volumes N2.0 × 10 −9 cm 3 STP/g because corrections for mass fractionation produced uncertainties too large to permit discussion of the length of the hiatus between post-shield stage and rejuvenated stage magmatism. The ages for C2-group samples MKPH-04, MKPH-14 and MKPH-15 are 2.74 ± 0.36, 2.57 ± 0.76 and 2.95 ± 0.67 Ma, respectively. K2O/P2O5 ratios for these samples are comparable to that of the B2 sample from the same flow (MKPH-03). The sample MKPH-04 displays lower atmospheric 40Ar contamination (88.4%), and has a smaller 36Ar volume (1.18 × 10 −9 cm 3 STP/g) than MKPH-03. The samples MKPH-14 and MKPH-15 show extensive atmospheric 40Ar contamination (94.5% and 92.9%) and high 36Ar contents (2.10 and 2.06× 10−9 cm3 STP/g), suggesting that atmospheric

Table 2 Major element compositions for the lavas from the Makapuu Head section. Location

Sample name

Makapuu Head section KOO-23 MKPH-01 MKPH-02 MKPH-09 MKPH-10 MKPH-11 MKPH-12 MKPH-13

Alteration category

SiO2 (wt%)

TiO2 (wt%)

Al2O3 (wt%)

Fe2O3 (wt%)

MnO (wt%)

MgO (wt%)

CaO (wt%)

B1 B1 D D D D D

52.31 53.08 52.75 52.93 53.03 53.03

1.93 1.91 1.72 1.95 1.69 1.84 1.85

14.46 14.33 14.30 13.76 14.46 14.11 14.06

11.21 11.30 10.80 11.71 10.66 11.24 11.33

Na2O (wt%)

K2O (wt%)

P2O5 (wt%)

Total

K2O/ P2O5

0.15 0.14 0.15 0.15 0.14 0.14 0.15

6.95 6.88 7.29 7.21 7.36 6.97 7.21

9.48 9.07 9.37 8.55 9.16 9.01 9.09

2.75 2.63 2.65 2.55 2.59 2.67 2.61

0.35 0.50 0.40 0.49 0.44 0.49 0.44

0.25 0.26 0.23 0.25 0.23 0.25 0.24

99.85 100.10 99.74 99.37 99.66 99.76 100.01

1.38 1.92 1.75 1.94 1.97 1.98 1.82

KOO-19

MKPH-03 MKPH-04 MKPH-14 MKPH-15 MKPH-16 MKPH-17

B2 C2 C2 C2 D D

52.01 52.07 51.56 52.05 50.11 49.62

2.28 2.26 2.25 2.26 2.43 2.51

14.56 14.44 14.47 14.48 15.48 15.94

10.94 10.92 10.87 10.93 11.70 12.08

0.14 0.14 0.14 0.14 0.15 0.14

6.73 6.73 6.69 6.69 6.88 6.54

9.52 9.47 9.87 9.60 9.98 9.93

2.76 2.76 2.71 2.71 2.68 2.72

0.41 0.47 0.43 0.48 0.17 0.11

0.31 0.32 0.32 0.32 0.28 0.27

99.67 99.59 99.30 99.68 99.86 99.85

1.30 1.47 1.34 1.48 0.58 0.42

KOO-07

MKPH-18 MKPH-19 MKPH-20 MKPH-21

A2 A2 D D

51.24 51.23 51.04 48.67

2.12 2.12 2.37 2.60

13.98 14.06 14.12 14.88

11.49 11.59 11.90 12.72

0.15 0.15 0.15 0.16

7.32 7.19 6.75 6.60

9.89 9.89 10.07 10.61

2.68 2.59 2.71 2.66

0.52 0.48 0.44 0.15

0.31 0.31 0.32 0.34

99.68 99.59 99.88 99.39

1.66 1.55 1.39 0.44

KOO-03

MKPH-05 MKPH-06 MKPH-08 MKPH-07

B1 B1 B1 C1

51.98 51.27 52.20 51.21

2.11 2.14 2.11 2.17

14.38 14.60 14.42 14.62

11.20 11.36 11.16 11.34

0.15 0.15 0.15 0.15

6.93 7.01 6.66 7.14

9.57 9.73 9.55 9.69

2.97 2.85 3.01 2.73

0.34 0.33 0.47 0.36

0.29 0.30 0.30 0.30

99.91 99.75 100.03 99.70

1.16 1.11 1.58 1.18

A2

51.37

2.09

13.86

12.30

0.16

7.49

9.69

2.39

0.52

0.26

100.14

2.00

B1

51.52

2.11

13.97

12.76

0.16

7.14

9.76

2.43

0.47

0.27

100.57

1.73

B1 or B2 A2

48.50 47.67

2.09 1.74

11.75 11.09

12.72 12.58

0.17 0.17

13.15 16.10

9.71 9.02

1.81 1.59

0.19 0.19

0.21 0.17

100.29 100.31

0.93 1.12

Submarine Koonau Oahu 500-8 North-a Oahu 500-7 North-a Nuuanu 2 498-4 Nuuanu 3 499-6A

The data for submarine Koolau are from Shinozaki et al. (2002).

46

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53

Table 3(a) K–Ar dating results for lavas from the Makapuu Head section. Location

Sample name

alteration category

wt.(g)

K2O (wt.%)

40

Ar/36Ar

38

KOO-23 N21°18′56″ W157°39′25″ elev. 131.5 m

HV02-11 MKPH-01 MKPH-02 MKPH-09 MKPH-10 MKPH-11 MKPH-12 MKPH-13

B1 B1 B1 D D D D D

4.52 2.71 4.14 4.81 3.70 4.83 4.44 4.31

0.61 0.551 0.565 0.490 0.617 0.550 0.549 0.557

463.3 ± 2.6 347.1 ± 1.4 397.8 ± 1.6 389.3 ± 1.6 420.1 ± 1.7 403.0 ± 1.6 414.4 ± 1.6 421.7 ± 1.7

KOO-19 N21°18′57″ W157°39′34″ elev. 91.4 m

MKPH-03 MKPH-04 MKPH-14 MKPH-15 MKPH-16 MKPH-17

B2 C2 C2 C2 D D

1.48 0.73 0.74 0.74 0.74 0.73

0.500 0.512 0.433 0.479 0.149 0.137

KOO-07 N21°18′59″ W 157°39′36″ elev. 20.1 m

MKPH-18 MKPH-19 MKPH-20 MKPH-21

A2 A2 D D

3.72 3.75 2.10 3.70

KOO-03 N21°18′60″ W 157°39′37″ elev. 4.9 m

MKPH-05 MKPH-06 MKPH-08 MKPH-07

B1 B1 B1 C1

4.32 0.73 2.48 2.23

Ar/36Ar

40 Ar/36Ar initial*

40 Ar rad. (10−8cm3 STP/g)

36 Ar atm.(%)

36 Ar (10−9cm3 STP/g)

Uncorrected Age** (Ma)

Corrected Age (Ma)

0.1870 ± 0.0016 0.1869 ± 0.0015 0.1866 ± 0.0015 0.1868 ± 0.0015 0.1862 ± 0.0015 0.1873 ± 0.0015 0.1874 ± 0.0015 0.1874 ± 0.0015

295.9 ± 5.2 295.6 ± 5.0 294.7 ± 4.9 295.3 ± 4.9 293.3 ± 4.9 296.8 ± 4.9 297.0 ± 4.9 297.0 ± 4.9

4.72 ± 0.18 4.24 ± 0.42 4.22 ± 0.22 3.73 ± 0.21 4.34 ± 0.20 4.20 ± 0.22 3.92 ± 0.19 4.20 ± 0.19

63.9 85.2 74.1 75.8 69.8 73.6 71.7 70.4

0.28 0.82 0.41 0.40 0.34 0.40 0.33 0.34

2.40 ± 0.08 2.39 ± 0.09 2.29 ± 0.07 2.35 ± 0.07 2.14 ± 0.06 2.39 ± 0.07 2.24 ± 0.07 2.36 ± 0.07

2.39 ± 0.12 2.39 ± 0.24 2.31 ± 0.13 2.36 ± 0.14 2.18 ± 0.11 2.37 ± 0.13 2.21 ± 0.11 2.33 ± 0.12

325.0 ± 1.3 331.3 ± 1.3 311.8 ± 1.2 311.3 ± 1.2 309.2 ± 1.2 315.9 ± 1.5

0.1851 ± 0.0016 0.1860 ± 0.0015 0.1866 ± 0.0015 0.1848 ± 0.0015 0.1878 ± 0.0017 0.1852 ± 0.0030

290.0 ± 5.3 292.9 ± 5.0 294.7 ± 4.9 289.2 ± 4.9 298.1 ± 5.6 290.4 ± 9.8

4.58 ± 0.71 4.54 ± 0.60 3.60 ± 1.07 4.56 ± 1.04 0.76 ± 0.40 1.05 ± 0.40

89.2 88.4 94.5 92.9 96.4 91.9

1.31 1.18 2.10 2.06 0.69 0.41

2.40 ± 0.12 2.56 ± 0.11 2.45 ± 0.20 2.11 ± 0.17 1.97 ± 0.18 1.90 ± 0.14

2.83 ± 0.44 2.74 ± 0.36 2.57 ± 0.76 2.95 ± 0.67 1.59 ± 0.83 2.37 ± 0.92

0.454 0.541 0.411 0.178

403.5 ± 1.6 426.0 ± 1.7 340.0 ± 1.4 315.8 ± 1.2

0.1861 ± 0.0015 0.1870 ± 0.0015 0.1861 ± 0.0015 0.1877 ± 0.0015

293.2 ± 4.9 295.9 ± 4.9 293.2 ± 4.9 297.9 ± 4.9

3.70 ± 0.18 4.27 ± 0.19 3.87 ± 0.42 0.81 ± 0.23

72.7 69.4 86.2 94.3

0.34 0.33 0.83 0.46

2.47 ± 0.07 2.45 ± 0.07 2.77 ± 0.11 1.61 ± 0.10

2.52 ± 0.14 2.44 ± 0.12 2.92 ± 0.32 1.42 ± 0.40

0.380 0.294 0.447 0.312

382.1 ± 1.5 341.1 ± 1.4 343.6 ± 1.4 320.0 ± 1.3

0.1850 ± 0.0015 0.1858 ± 0.0016 0.1860 ± 0.0015 0.1850 ± 0.0015

289.7 ± 4.9 292.3 ± 5.2 292.7 ± 4.9 289.7 ± 4.9

3.24 ± 0.19 2.37 ± 0.26 3.52 ± 0.36 2.40 ± 0.40

75.8 85.7 85.2 90.6

0.35 0.43 0.69 0.79

2.47 ± 0.08 2.34 ± 0.08 2.30 ± 0.10 1.93 ± 0.08

2.64 ± 0.16 2.50 ± 0.28 2.44 ± 0.25 2.39 ± 0.40

HV02-11 data is from Ozawa et al. (2005), and other data for Makapuu Head lavas is from Sawada (2007). Datum for location is NAD83. Altitudes relative to sea level. Errors are given in 2σ. *initial 40Ar/36Ar calculated from 38Ar/36Ar assuming mass fractionation. **mass fractionation uncorrected.

40

Ar contamination increased during the weathering/alteration process. The ages from the B2 and C2 samples are similar to ages from less weathered samples, although the uncertainties are larger than those of A2 and B1 samples from other flows. The ages for samples MKPH-16 and MKPH-17, classified as D group, are 1.59 ± 0.83 and 2.37 ± 0.92 Ma, respectively. Extremely low K2O/P2O5 ratios for these samples (0.58 and 0.42) suggest significant K loss, and their high atmospheric 40Ar contamination (96.4% and 91.9%) and smaller 36Ar volumes (0.69 and 0.41 × 10−9 cm3 STP/g) suggest a decrease in the Ar volume of these samples, as well. The ages for A2 group samples MKPH-18 and MKPH-19 collected from the KOO-07 flow are 2.52 ± 0.14 and 2.44 ± 0.12 Ma respectively and agree within the range of their respective analytical errors (Fig. 4). The K2O/P2O5 ratios of 1.66 and 1.55, low atmospheric 40Ar contaminations (72.7% and 69.4%) and small 36Ar volumes (0.34 and 0.33 × 10 −9 cm 3 STP/g) indicate that there was no loss of K and Ar from these samples. Hence, the best age estimate for this flow is the weighted mean age for the A2 samples, of 2.47 ± 0.09 Ma. The ages for samples MKPH-20 and MKPH-21, classified as D group, are 2.92 ± 0.32 and 1.42 ± 0.40 Ma, respectively. MKPH-21 has a significantly lower K2O/P2O5 ratio (0.44), a higher atmospheric 40Ar contamination (94.3%), and a small 36Ar volume (0.46 × 10 −9 cm 3 STP/g), suggesting K and Ar loss from this sample.

The samples MKPH-05, MKPH-06 and MKPH-08 were collected from the lowermost flow (KOO-3) in the Makapuu Head section and classified in the B1 group. The ages are 2.64 ± 0.16, 2.50 ± 0.28 and 2.44 ± 0.25 Ma, respectively. The weighted mean for these three ages is 2.58 ± 0.13 Ma. Slightly lower K2O/P2O5 ratios for MKPH-05 and MKPH-06 of 1.16 and 1.11 suggest K loss from these samples. The age for sample MKPH-07, classified in the C1 group, is 2.39 ± 0.40 Ma and the error is larger than for the three B1 samples from the flow. The K2O/P2O5 ratio of 1.18 for the sample suggests that K loss, higher atmospheric 40Ar contamination (90.6%) and larger 36Ar volume (0.79 × 10 −9 cm 3 STP/g) compared to the B1 samples cause the larger uncertainty in the age corrected for mass fractionation. Fig. 5(a) to (c) shows the ages versus K2O/P2O5 ratios, 36Ar volumes and atmospheric 40Ar contaminations, respectively. The span between the best age estimates of the uppermost flow (KOO-23) and the lowermost flow (KOO-03) are also shown in these figures. The Makapuu samples with K2O/P2O5 ratios less than 1.0 have large uncertainties on the ages (Fig. 5(a)). Two of the samples show younger apparent ages than the best age estimate span. One explanation for this observation may be that they also have lost a significant amount of Ar through alteration. B2, C1, C2 and D samples with K2O/P2O5 ratios around 1.25 to 1.5 have larger errors compared to those with K2O/P2O5 ratios N1.5 (Fig. 5(a)). The sample MKPH-20

Table 3(b) Statistic results of the age data for Makapuu Head section. Location

Weighted mean age for all samples (Ma)

n

MSWD

Probability

Weighted mean age for preferred samples (Ma)

n

MSWD

Probability

KOO-23 KOO-19 KOO-07 KOO-03

2.30 ± 0.04 2.66 ± 0.23 2.46 ± 0.47 2.55 ± 0.11

8 6 4 4

1.8 1.7 12 0.94

0.088 0.12 0.00 0.42

2.36 ± 0.08 (3.4%) (for B1 samples)

3

0.45

0.64

2.47 ± 0.09 (3.6%) (for A2 samples) 2.57 ± 0.12 (4.6%) (for B1 samples)

2 3

0.75 1.05

0.39 0.35

Errors are given 2σ.

(2.0%) (8.5%) (19%) (4.5%)

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53

1.0

B2 2.83±0.44 Ma

KOO-07 (Pahoehoe)

KOO-03 (Aa)

w.m. for A2 w.m. for B1 2.47±0.09 Ma 2.58±0.13 Ma

Nuuanu 3

w.m. for B1 2.36±0.08 Ma

KOO-19 (Aa)

Submarine lavas

Nuuanu 2

KOO-23 (Aa)

0.5

l ower flow

N-Oahu-a

Makapuu section

upper flow

0.0

47

Age (Ma)

1.5 2.0 2.5 3.0 3.5 4.0 4.5

A2 B1 B2 C1 C2 D

5.0 Fig. 4. Unspiked K–Ar ages for the samples of each flow from the Makapuu Head section and submarine Koolau lavas. Uncertainties are 2σ. The gray bands are weighted mean age for preferred samples for each flow.

from the KOO-07 flow classified in the D group has a K2O/P2O5 ratio of 1.39, indicating that there was no K loss based on the criteria of Rhodes (1996). However, the age is older than the best age estimates beyond 2σ uncertainty (Fig. 5(a)). These observations suggest that careful consideration is needed if the K2O/P2O5 ratio of tholeiitic basalt samples is lower than 1.5. Four samples classified in the B2 and C2 groups, which have secondary minerals in the vesicles, have 36Ar volumes N1.0 × 10 −9 cm 3 STP, larger than those of other samples (Fig. 5(b)), although their ages are indistinguishable from the best age estimates. The large 36Ar volume implies addition of atmospheric Ar by secondary water

circulation. The most intensely altered samples from the KOO-07 and KOO-19 flows are classified in Group D, have relatively small 36Ar volumes (Fig. 5(b)), but also have low radiogenic 40Ar volumes (Table 2), suggesting secondary Ar loss in these samples. Samples in the C1, C2 and D Groups (except for the Group D samples from the KOO-23 flow) show atmospheric 40Ar contamination higher than 86% (Fig. 5(c)). Overall, atmospheric 40Ar contamination appears to correlate with degree of alteration because the Makapuu samples used in this study have nearly coincident ages. However, atmospheric 40 Ar contamination is also dependent on a sample's age, and careful consideration is needed when we use atmospheric 40Ar contamination

Table 4 K–Ar dating results for submarine Koolau lavas. Location

Sample Alteration wt. name category (g)

K2O (wt.%)

40

Ar/36Ar

38

Ar/36Ar

40 Ar/36Ar initial*

40 40 36 Ar rad. Ar Ar Uncorrected Corrected (10− 8cm3STP/g) atm.(%) (10− 9cm3STP/g) Age** (Ma) Age (Ma)

S500-7 Oahu North-a ″ N21°50′49″ W 157°46′16″ elev. − 2646 m

A2

0.72 0.495 2.28

370.5 ± 1.6 0.1871 ± 0.0018 296.2 ± 6.1 400.7 ± 2.2 0.1844 ± 0.0023 287.9 ± 7.5

5.20 ± 0.46 5.38 ± 0.40

80.0 71.8

0.70 0.48

3.28 ± 0.13 3.14 ± 0.12 w.m. 3.31 ± 0.20

3.25 ± 0.30 3.36 ± 0.26

S500-8 Oahu North-a ″ N21°50′36″ W 157°46′16″ elev. − 2743 m

B1

0.70 0.550 1.65

338.2 ± 1.4 0.1865 ± 0.0015 294.3 ± 4.9 342.7 ± 1.4 0.1867 ± 0.0016 294.7 ± 5.1

4.62 ± 0.55 4.37 ± 0.50

87.0 86.0

1.05 0.91

2.53 ± 0.11 2.42 ± 0.11 w.m. 2.53 ± 0.20

2.60 ± 0.31 2.46 ± 0.28

S498-4 Nuuanu 2 N22°15′24″ W 156°59′68″ elev. − 3718 m

B1 or B2

0.58 0.261

306.5 ± 1.2 0.1890 ± 0.0024 302.1 ± 8.0

1.41 ± 2.57

98.6

3.20

4.16 ± 0.47

1.67 ± 3.05

S499Nuuanu 3 6A N21°58′82″ W 157°02′14″ elev. − 4372 m

A2

0.62 0.284

387.3 ± 6.8 0.1902 ± 0.0120 305.5 ± 39.3 2.44 ± 1.19

78.9

0.30

2.94 ± 0.20

2.62 ± 1.28

2.04

392.9 ± 1.6

75.2

0.28

2.89 ± 0.11

295.5

2.70 ± 0.04

w.m. 2.89 ± 0.09 Major element data (K2O/P2O5) are from Shinozaki et al. (2002). Dutum for location is NAD83. Altitudes relative to sea level. Negative altitude is bathymetric depth. Errors are given in 2σ. w.m.: weighted mean ages. Weighting by inverse variance. *initial 40Ar/36Ar calculated from 38Ar/36Ar assuming mass fractionation. **mass fractionation uncorrected.

48

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53

(a)

A2 B1 B2 C1 C2 D (KOO-23) D (KOO-19) D (KOO-07)

0.5 1.0

Age (Ma)

1.5 2.0 2.5 3.0 3.5 4.0 0.0

0.5

1.0 1.25 1.5

2.0

2.5

K2O/P2O5

(b) 0.5 1.0

Age (Ma)

1.5 2.0 2.5

preservation may not be a good proxy for the extent of alteration in K-bearing groundmass minerals, since these are composed chiefly of plagioclase in the case of tholeiitic basalt. 4.2. Accumulation rate of lava in the makapuu head section Weighted means for the ages for A2-and B1-group samples for flows KOO-03, KOO-07 and KOO-23 (ascending order) are 2.58 ± 0.13, 2.47 ± 0.09 and 2.36 ± 0.09 Ma, respectively, and are consistent with the stratigraphy. This means that the unspiked K–Ar technique is equally applicable to fresh aa lava samples (KOO-03, -23) and pahoehoe lava samples (KOO-07). Since the stratigraphic thickness is ~ 120 m between KOO-03 and KOO-23, the accumulation rate calculated for this section is in the range 0.04 to 0.11 cm/ year. Accumulation rates for comparatively thick, well-sampled stratigraphic sections at other Hawaiian volcanoes are limited to Mauna Kea and Kilauea (HSDP; 2.7 km in thickness; Sharp and Renne, 2005), Waianae (115–220 m thick; Guillou et al., 2000) and West Maui (~ 300 m thick; Sherrod et al., 2007). The accumulation rates of Mauna Kea are staggering, as great as 0.86 cm/year in the lower part of the shield sequence (Sharp and Renne, 2005). They diminish upsection to 0.09 cm/year. The accumulation rates of Waianae volcano (Oahu) range from 0.038 to 0.20 cm/year (Guillou et al., 2000) and those of West Maui range from 0.025 to 0.08 cm/ year for their late shield stage (Sherrod et al., 2007). The accumulation rate of 0.04 to 0.11 cm/year for the Makapuu Head section of Koolau volcano confirms the diminishing accumulation rate at the end of the shield stage suggested by the results from other Hawaiian volcanoes.

3.0 4.3. Unspiked K–Ar ages for submarine Koolau lavas

3.5 4.0 0.0

0.5

1.0

1.5

2.0

Ar (10-9 cm3STP/g)

36

(c) 0.5 1.0

Age (Ma)

1.5 2.0 2.5 3.0 3.5 4.0 60

65

70

75

80

86

90

95

100

Atmospheric 40Ar contamination (% ) Fig. 5. Geochemical data that serve as indicators of weathering for the lavas from the Makapuu Head section, plotted with corresponding K–Ar ages of analyzed samples. (a) K2O/P2O5 ratio, (b) 36Ar gas volume, (c) Atmospheric 40Ar contamination. The gray bands show the span between the best age estimates of the uppermost flow (KOO-23) and the lowermost flow (KOO-03).

to evaluate alteration of samples. All ages of the samples from the KOO23 flow agree with each other within their analytical uncertainties, even though some of the samples are classified in group D. Olivine

The unspiked K–Ar ages for submarine Koolau lavas are presented in Table 3b. Duplicate Ar analyses were made for samples S500-7 and S500-8 (alteration criteria for A2 and B1 categories, respectively), and each pair of results agrees well within analytical error. Hence, weighted mean ages are used hereafter. The age for sample S500-8, in-place lava collected from the east flank of Koolau volcano, is 2.53 ± 0.20 Ma, roughly coincident with the lowest part of the subaerial Makapuu Head section (Fig. 5). The result implies that the Makapuu stage started no later than ~ 2.6 Ma. Published ages for Koolau shield stage range between 1.8 and 3.2 Ma, however, the younger Koolau samples have very low K concentrations indicating that these samples probably lost K during weathering (Haskins and Garcia, 2004; Ozawa et al., 2005). Ozawa et al. (2005) suggested the shield volcanism of Koolau probably ended at ~ 2.1 Ma. The age for sample S500-7 (not in place) of 3.31 ± 0.10 Ma is older than the oldest age for subaerial lavas (3.2 Ma) reported by Ozawa et al. (2005), suggesting that the onset of the shield stage of Koolau volcano was no later than ~ 3.3 Ma. This age is consistent with the age constraint from microfossils in the submarine sediment from Tuscaloosa Seamount (2.5–3.3 Ma; Morgan et al., 2006). Since the end of the shield stage is regarded to be ~ 2.1 Ma (Ozawa et al., 2005), the duration of the shield stage volcanism is at least 1.2 m.y. These results are somewhat longer than the estimate for the typical shield stage of Hawaiian volcanoes by Clague and Dixon (2000), who estimated a duration of 0.7 m.y. for their three phases of typical shield growth. Sample S498-4 was classified as B1 or B2 based on the degree of freshness of the olivine, but was not classified further because of the absence of vesicles. The mass fractionation uncorrected age for this sample is 4.16 ± 0.23 Ma. The corrected age (1.65 ± 3.04 Ma) has a large uncertainty, caused by high atmospheric 40Ar contamination (98.6%) and large 36Ar volume (3.20 × 10 −9 cm 3 STP/g). In addition, the low K2O/P2O5 ratio (0.93) of the sample suggests K loss. For these reasons the age is judged to be less reliable. Sample S499-6A collected from the Nuuanu landslide

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53

block likely correlates with the early main-shield stage. The alteration category of this sample is A2, which is basically acceptable for selecting datable samples. Duplicate analyses yield ages uncorrected for mass fractionation of 2.94 ± 0.20 and 2.88 ± 0.10 Ma, which agree well within the analytical error. The 38Ar/ 36Ar ratio of the first measurement of this sample has a large analytical uncertainty, and even on the second measurement using an increased amount of sample the 38Ar peak was not well determined, without any identified technical cause. Therefore, we prefer the uncorrected ages for this sample. Moreover, the K2O/ P2O5 ratio of 1.13 suggests significant K loss from this sample. Thus, we tentatively assign a weighted-mean uncorrected age of 2.89 ± 0.04 Ma for sample S499-6A.

comments that substantially improved this manuscript. This study was supported by a Grant-in-Aid (no. 16403009) as well as by a Grant-in-Aid for the 21st Century COE Program (Kyoto University, G3) from the Japanese Ministry of Education, Culture, Sports, Science and Technology. Appendix I. Photographs of the thin sections Photographs of the thin sections of the Makapuu Head samples for each flow. The alteration categories are also shown in the photographs. The scale bar is 0.5 mm. Appendix II. Unspiked K–Ar dating K–Ar age can be calculated from the equation;

5. Summary We classified several samples from four lava flows in the Makapuu Head section on the basis of freshness of olivine phenocrysts, freshness of groundmass olivine and presence of secondary minerals in vesicles. K–Ar dating for the tholeiitic samples with varying degrees of weathering/alteration led to the following findings:

49

t=

" #    40 1 Ar λe + λβ ln 40 +1 λe + λβ λe K

where t is the K–Ar age, λe and λβ are the decay constants of 40K to Ar (0.581 × 10 −10; Steiger and Jäger, 1977) and of 40K to 40Ca (4.962 × 10 −10; Steiger and Jäger, 1977), respectively. 40Ar* is the concentration of radiogenic 40Ar. 40K is the concentration of 40K calculated from the total K concentration with the present-day atomic abundance ratio of 40K/K (1.167 × 10 −4; Steiger and Jäger, 1977). In the unspiked K–Ar method (the peak comparison method; Matsumoto et al., 1989), the concentration of the total 40Ar is determined by the comparison of the peak intensity with that of a known amount of the Ar standard. The concentration of radiogenic 40 Ar is calculated by the equation; 40

(1) The uncertainty on the in an age tends to increase as the degree of alteration increases. (2) The ages for samples with fresh groundmass olivine are reliable, even though olivine phenocrysts are slightly altered (thin alteration rims) or contain secondary minerals crystallized in the vesicles (i.e. A2- and B1-group samples, see Table 1). (3) The ages of the samples classified as groups B2, C1 and C2 have larger errors, although they are not biased significantly from the reference age. (4) The ages for the samples with altered groundmass olivine (i.e. D-group samples) are disturbed beyond the errors, probably because of K and/or Ar loss. The geological implications of K–Ar dating for Makapuu Head sections are as follows: (1) The ages for the Makapuu Head section, which is approximately 120 m thick, are 2.58–2.36 Ma. The rate of accumulation of this section ranges from 0.04 to 0.11 cm/year, which implies a diminishing accumulation rate at the end of the shield stage of Koolau volcano. We also determined the K–Ar ages for the submarine Koolau lavas and concluded the following: (1) The likely age of the onset of the shield stage of Koolau volcano was no later than ~ 3.3 Ma. (2) The Makapuu-stage, the late-shield stage, in which lavas show enriched end-members in Hawaiian shield lavas, started no later than ~2.6 Ma. (3) The duration of the shield stage volcanism of Koolau volcano is at least 1.2 m.y. Acknowledgments We are grateful to Mike Garcia for valuable information on stratigraphy and sampling sites. Sample collections on the submarine Koolau volcano were thoroughly supported by JAMSTEC Kairei/Kaiko and Yokosuka/Shinkai 6500 operation teams. We appreciate the useful comments and English review by Mr. G. McCrank. We are indebted to the journal reviewers David Sherrod and Fred Jourdan, the editor Laurie Reisberg for their thorough reviews and constructive

40



Ar =

40

Artotal ð1−R0 = RÞ

where 40Artotal is the total 40Ar concentration, R0 is the initial Ar/ 36Ar, and R is the measured 40Ar/ 36Ar in a sample. Note that R0/R is the proportion of atmospheric contamination. In the conventional K–Ar dating, the initial 40Ar/ 36Ar is assumed to be equal to that of the present-day atmosphere (295.5). However, Ar isotopic ratios in many historical lavas have been shown to be mass fractionated from the present-day atmospheric ratio (e.g. Matsumoto and Kobayashi, 1995; Ozawa et al. 2006). In unspiked K–Ar dating, therefore, the initial 40 Ar/ 36Ar in each sample is estimated based on the assumption that the initial 40Ar/ 36Ar in all volcanic rocks lies on the mass fractionation line, as shown in the equations; 40

R0 = RA ð1 + 4δÞ

δ = ðr = rA −1Þ = 2 where RA is the 40Ar/ 36Ar in the present-day atmosphere, rA is Ar/ 36Ar in the present-day atmosphere, and r is the 38Ar/ 36Ar in a sample. Note that this technique cannot be used when conventional spiked K–Ar dating is employed, as a 38Ar isotopic tracer is added. However, the assumption that variations in the initial ratio result solely from mass fractionation is not always accurate, and some historical lavas have extremely high 40Ar/ 36Ar ratios, suggesting extraneous 40Ar contamination (e.g. Ozawa et al., 2006). Since extraneous 40Ar contamination cannot be corrected even with the unspiked method, it has to be removed prior to analysis. Ozawa et al. 38

50

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53

(2006) measured the K–Ar ages of the whole rock samples and the groundmass samples for some Hawaiian historical lavas, and they suggested that using fine-grained groundmass samples can significantly reduce extraneous 40Ar contamination. They proved that the argon isotopic compositions for the groundmass samples of Mauna Loa and Kilauea agreed well with the mass fractionation line through the atmospheric composition. The calculation of the uncertainty for age determinations by the unspiked method is summarized in Matsumoto et al. (1989), as following; 2

2

2

σt = σK + σAr

2

2

σAr = σx +

 AC 2  2 2 σR + σR0 2 ð1−AC Þ

2

σR0 = 2rσr = ð2r−rA Þ

AC = ð2r−rA ÞRA = ðrA RÞ where σt, σK, and σAr are the relative errors in the K–Ar age, the determination of 40K, and the determination of 40Ar. And σx, σR, σR0 and σr are the relative errors in the determination of the total 40Ar, the 40 Ar/ 36Ar ratio, the initial 40Ar/ 36Ar ratio and the 38Ar/ 36Ar ratio. Ac means fraction of atmospheric argon.

MKPH-01 (B1)

MKPH-02 (B1)

MKPH-09 (D)

MKPH-10 (D)

MKPH-11 (D)

MKPH-12 (D)

MKPH-13 (D)

MKPH-03 (D)

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53

MKPH-04 (C2)

MKPH-14 (C2)

MKPH-15 (C2)

MKPH-16 (D)

MKPH-17 (D)

MKPH-18 (A2)

MKPH-19 (A2)

MKPH-20 (D)

51

52

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53

MKPH-21 (D)

MKPH-06 (B1)

MKPH-05 (B1)

MKPH-08 (B1)

MKPH-07 (C1)

References Clague, D., Dixon, J., 2000. Extrinsic controls on the evolution of Hawaiian ocean island volcanoes. Geochemistry, Geophysics, Geosystems 1, 1010. doi:10.1029/ 1999GC000023. Doell, R., Dalrymple, G., 1973. Potassium–argon ages and paleomagnetism of the Waianae and Ko'olau Volcanic Series, O'ahu, Hawaii. Geol. Soc. Am. Bull 84, 1217–1242. Frey, F., Garcia, M., Roden, M., 1994. Geochemical characteristics of Koolau volcano: implications of intershield differences among Hawaiian volcanoes. Geochim. Cosmochim. Acta 58, 1441–1462. Goto, A., Tatsumi, Y., 1994. Quantitative analysis of rock samples by an X-ray fluorescence spectrometer (I). The Rigaku Journal 11, 40–59. Goto, A., Tatsumi, Y., 1996. Quantitative analysis of rock samples by an X-ray fluorescence spectrometer (II). The Rigaku Journal 13, 20–38. Gramlich, J.W., Lewis, V.A., Naughton, J.J., 1971. Potassium–argon dating of Holocene basalts of the Honolulu Volcanic series. Geol. Soc. Am. Bull. 82, 1399–1404. Guillou, H., Turpin, L., Garnier, F., Charbit, S., Thomas, D.M., 1997. Unspiked K–Ar dating of Pleistocene tholeiitic basalts from the deep core SOH-4, Kilauea, Hawaii. Chemical Geology 140, 81–88. Guillou, H., Sinton, J., Laj, C., Kissel, C., Szeremeta, N., 2000. New K–Ar ages of shield from Waianae Volcano, Oahu, Hawaiian archipelago. J. Volcanol. Geotherm. Res. 96, 229–242. Haskins, E.H., Garcia, M.O., 2004. Scientific drilling reveals geo-chemical heterogeneity within the Ko'olau shield, Hawai'i. Contrib. Mineral. Petrol. 17, 162–188. Herrero-Bervera, E., Browne, E.D., Valet, J.-P., Singer, B.S., Jicha, B.R., 2007. Cryptochron C2r.2r-1 recorded 2.51 Ma in the Koolau Volcano at Halawa, Oahu, Hawaii, USA: Paleomagnetic and 40Ar/39Ar evidence. Earth Planet. Sci. Lett. 254, 256–271. Laj, C., Szeremeta, N., Kissel, C., Guillou, H., 2000. Geomagnetic paleointensities at Hawaii between 3.9 and 2.1 Ma: preliminary results. Earth Planet. Sci. Lett. 179, 191–204. Lanphere, M.A., Dalrymple, G.B., 1980. Age and strontium isotopic composition of the Honolulu Volcanic Series, Oahu, Hawaii. Am. J. Sci. 280-A (Jackson Volume), 736–751.

Lipman, P.W., Sission, T.W., Ui, T., Naka, J., Morgan, J., Smith, J., 2002. Ancestrial submarine growth of Kilauea Volcano and instability of its south flank. In: Takahashi, E., Lipman, P.W., Garcia, M.O., Naka, J., Aramaki, S. (Eds.), Hawaiian Volcanoes: Deep Underwater Perspectives: American Geophysical Union Geophysical Monograph, 128, pp. 161–191. Matsumoto, A., 1989. Improvement for determination of potassium in K–Ar dating. Bull. Geol. Surv. Jpn. 40, 65–70 (in Japanese with English abstract). Matsumoto, A., Kobayashi, T., 1995. K–Ar age determination of late Quaternary volcanic rocks using the “mass fractionation correction procedure”: application to the younger Ontake volcano, central Japan. Chem. Geol. 125, 123–135. Matsumoto, A., Uto, K., Shibata, K., 1989. K–Ar dating by peak comparison method — new technique applicable to rocks younger than 0.5 Ma. Bull. Geol. Surv. Jpn. 40, 565–597. McDougall, I., 1964. Potassium–argon ages from lavas of the Hawaiian Islands. Geol. Soc. Am. Bull. 75, 107–128. Moore, J.G., Clague, D.A., 2002. Mapping the Nuuanu and Wailau landslides in Hawaii. In: Takahashi, E., Lipman, P.W., Garcia, M.O., Naka, J., Aramaki, S. (Eds.), Hawaiian Volcanoes: Deep Underwater Perspectives: American Geophysical Union Geophysical Monograph, 128, pp. 223–244. Moore, J.G., Normark, W., Holcomb, R., 1994. Giant Hawaiian landslides. Annu. Rev. Earth Planet. Sci. 22, 119–144. Morgan, J.K., Resig, J.M., Siesser, W.G., 2006. Microfossil assemblages on Tuscaloosa seamount and their constraints on the age of Nuuanu landslide, north of Oahu. HI. J. Volcanol. Geotherm. Res. 151, 269–278. Ozawa, A., Tagami, T., Garcia, M., 2005. Unspiked K–Ar dating of the Honolulu rejuvenated and Koolau shield volcanism on Oahu. Hawaii. Earth Planet. Sci. Lett. 232, 1–11. Ozawa, A., Tagami, T., Kamata, H., 2006. Argon isotopic composition of some Hawaiian historical lavas. Chem. Geol. 226, 66–72. Rhodes, J., 1996. Geochemical stratigraphy of lava flows samples by the Hawaii Scientific Drilling Project. J. Geophys. Res. 101, 11729–11746. Robinson, J.E., Eakins, B.W., 2006. Calculated volumes of individual shield volcanoes at the young end of the Hawaiian Ridge. J. Volcanol Geotherm Res. 151, 309–317. Sano, H., thesis. Unspiked K–Ar dating of rejuvenated and shield-building volcanism on Kauai, Hawaii. M.D., Kyoto Univ. (in Japanese with English abstract).

S. Yamasaki et al. / Chemical Geology 287 (2011) 41–53 Sawada, R., thesis. Evaluation of the influence of weathering on the K–Ar age; Unspiked K–Ar dating of submarine samples dredged from off Kauai, Hawaii. M.D., Kyoto Univ. (in Japanese with English abstract). Sharp, W.D., Renne, P.R., 2005. The 40Ar/39Ar dating of core recovered by the Hawaii Scientific Drilling Project (phase 2), Hilo, Hawaii. Geochem. Geophys. Geosys 6 (4). doi:10.1029/2004GC000846. Sherrod, D.R., Murai, T., Tagami, T., 2007. New K–Ar ages for calculating the end-of shield extrusion rates at West Maui volcano, Hawaiian island chain. Bull. Volcanol. 69, 627–642. Shinozaki, K., Ren, Z.-Y., Takahashi, E., 2002. Geochemical and petrological characteristics of Nuuanu and Wailau landslide blocks. In: Takahashi, E., Lipman, P.W., Garcia, M.O., Naka, J., Aramaki, S. (Eds.), Hawaiian Volcanoes: Deep Underwater Perspectives: AGU Geophysical Monograph, 128, pp. 297–310. Steiger, R.H., Jäger, E., 1977. Convention on the use of decay constants in geo-and cosmochronology. Earth Planet. Sci. Lett. 36, 359–362. Stille, P., Unruh, D.M., Tatsumoto, M., 1986. Pb, Sr, Nd and Hf isotopic constraints on the origin of Hawaiian basalts and evidence for a unique mantle source. Geochim. Cosmochim. Acta 50, 2303–2319.

53

Sudo, M., Tagami, T., Sato, K., Hasebe, N., Nishimura, S., 1996. Calibration of a new Ar analytical system for the K–Ar dating method and analytical results of K–Ar age known samples. Mem. Fac. Sci., Kyoto Univ. Ser. Geol. Mineral. 58, 21–40. Takahashi, E., Nakajima, K., 2002. Melting process in the Hawaiian mantle plume head: an experimental study. In: Takahashi, E., Lipman, P.W., Garcia, M.O., Naka, J., Aramaki, S. (Eds.), Hawaiian Volcanoes, Deep Underwater Perspectives: AGU Geophysical Monograph, 128, pp. 403–418. Tanaka, R., Nakamura, E., Takahashi, E., 2002. Geochemical evolution of Koolau volcano, Hawaii. In: Takahashi, E., Lipman, P.W., Garcia, M.O., Naka, J., Aramaki, S. (Eds.), Hawaiian Volcanoes, Deep Underwater Perspectives: AGU Geophysical Monograph, 128, pp. 311–332. Thornber, C.R., Hon, K., Heliker, C., Sherrod, D.A., 2003. A comparison of whole-rock and grass major-element geochemistry of Kilauea Volcano, Hawaii, near-vent eruptive products: January 1983 through September 2001. U.S.G.S. Open File Report 03–477. http://geopubs.wr.usgs.gov/open-file/of03-477/. U.S. Geological Survey (USGS), 1999. 7.5-Minute topographic Maps: Koko Head.