Disturbed age spectra are common for argon data. Discordant 40Ar/39Ar age spectra can occur because the argon step-heating technique, resulting from the differential thermal stability of mineral phases during in vacuo heating, has the potential to separate different argon reservoirs hosted in varying mineralogical domains. The contribution of each separate mineral phase to the age spectra, however, may be monitored by elemental variations (Ca, Cl, and K from neutron-derived 37Ar, 38Ar, and 39Ar, respectively). Meaningful interpretation of argon data, therefore, requires the support of petrographic and chemical analyses at a microscopic scale. High meaningless apparent ages in the low-temperature region such as those observed in Sample 185-1149A-4H-3, 96–97 cm, have also been observed in other argon step-heating analyses on volcanic rocks (e.g., Lo et al., 1994; Koppers et al., 2000). This was attributed to the lower release temperature of K-rich alteration mineral phases that may experience 39Ar recoil loss during irradiation. High meaningless apparent ages in the high-temperature steps such as we observe (e.g., Samples 185-1149A-12H-6, 79–80 cm, and/or 191-1179B-5H-1, 31–32 cm) were also observed by Lo et al. (1994) and attributed to the release of argon from phenocrysts that may host extraneous argon (both inherited or excess argon). However, the intermediate region of the age spectra in our samples, attributed in previous studies to outgassing of glass and groundmass plagioclase by Lo et al. (1994) and of interstitials (glassy or microcrystalline) by Koppers et al. (2000), yield good plateau ages that are similar to the isochron ages (see "OSU-191" and "OSU-185" in the "Supplemental Materials" contents list). For this reason, despite the complicated release pattern, we consider the obtained ages meaningful.
Samples 185-1149A-4H to 13H and Samples 191-1179C-15H-6, 106–107 cm, and 191-1179C-16H-4, 45–46 cm, contain fair to good amounts of radiogenic 40Ar, with some of the heating steps in each sample having ~50% or higher 40Ar. The rest of the samples contain significantly lower percentages of radiogenic 40Ar. The effect this low radiogenic 40Ar has on the accuracy of the sample analysis is reflected in the larger uncertainty for these samples (e.g., 24% uncertainty for Sample 191-1149A-2H-4, 85–86 cm). The obtained ages follow, in general, both the stratigraphy of the sample groups (i.e., deeper samples give older ages at both sites) and the magnetostratigraphic ages obtained shipboard. Therefore, we consider that the ash beds from Sites 1149 and 1179 are dated with success.
Based on our obtained ages, the uppermost sample at Site 1149 (Sample 185-1149A-1H-2, 55–56 cm) yields an age older than the ash beds below. Although this age is clearly incompatible with its position in the stratigraphic column, the plateau and isochron are quite acceptable by analytical criteria. Thus, it is unlikely this result is due to sample contamination. Instead, either the glass contains "excess" 40Ar that was not outgassed on eruption or the sample includes some older fragmental material that was incorporated in the eruption event. We prefer this last explanation for the older age since the isochron (see "OSU-191" and "OSU-185" in the "Supplemental Materials" contents list) shows no evidence of a younger age and a nonatmospheric 40Ar/36Ar intercept.
For most of the remaining samples, 40Ar/39Ar data yields, in general, internally consistent results based on correlation of the obtained 40Ar/39Ar ages with the shipboard magnetostratigraphic ages (Tables T1, T2). Few inconsistencies, however, are observed between the 40Ar/39Ar and the magnetostratigraphic ages. For example, older 40Ar/39Ar ages than those obtained by shipboard magnetostratigraphy in the same stratigraphic position are observed in Samples 185-1149A-10H-5, 80–81 cm; 191-1179B-4H-6, 18–19 cm; and 5H-1, 31–32 cm (Figs. F2, F4; Tables T1, T2). For Sample 191-1179C-5H-6, 68–69 cm, however, the 40Ar/39Ar age is consistent with the shipboard magnetostratigraphic age if we considered the uncertainty in the weighted plateau age (i.e., 2.03 + 0.16 = 2.19 Ma).
Based on the observations above, we summarize in Table T5 a new combined 40Ar/39Ar and magnetostratigraphic chronostratigraphy for the northwestern Pacific Sites 1149 and 1179. Age-depth plots for Sites 1149 and 1179 based in this new refined chronology are shown in Figure F6.