This thesis examines chronologies of ice cores using their gas records. We first give two examples for constructing chronologies by the gas synchronization method.

The first example is the chronology reconstruction for the bottom sections of the GISP2 and GRIP ice cores from Summit, Greenland. This reconstruction is achieved by comparing two gas properties whose values vary uniformly with time (δ¹⁸O of O₂ and CH₄ concentration) with profiles of these properties in the stratigraphically intact Vostok ice core. This chronology is further constrained by NGRIP δ¹⁸O _ice values during the end of marine isotope stage 5e. The reconstructed chronology is partially validated by the triple isotope composition of O ₂. The reconstructed chronology indicates that the disturbed section primarily includes ice from the last interglacial (MIS 5e) and the penultimate glacial period (MIS 6). The oldest ice in the basal layer of GISP2 and GRIP has an age ≥ 237 ka.

The second example is the reconstruction of the age-depth relationship of the Vostok ice core for the section between 3,300 m and 3,347 m. Three gas properties in ice, CO₂, CH₄ and δ¹⁸ O_atm are compared with those in the EPICA Dome C core. We examined 14 depths from this interval. The reconstructed chronology shows that this section most likely contains ice from Termination V, but the stratigraphic order of ice is reversed. This 'flipped' layer starts somewhere between 3,316 m and 3,319 m, and ends somewhere between 3,340 m and 3,343 m. The oldest ice in this section is dated as ≥440 ka, confirming the existence of ice from MIS-12 (or possibly earlier) in the Vostok ice core.

We then explore O₂/N₂ ratios of occluded air in the GISP2 and Vostok ice cores. The GISP2 δO₂/N₂ record shows strong signals centered at the orbital frequencies, and δO ₂/N₂ is in anti-phase with local summer insolation. This observation is consistent with the earlier findings for the Vostok and Dome Fuji ice cores from East Antarctica [Bender, 2002; Kawamura, 2000]. It validates previous conclusions that fractionation during bubble close-off depends on ice grain properties set at the surface by solar insolation. In addition, the GISP2 δO₂/N₂ shows millennial duration signals that are in phase with the local temperature record of rapid climate change.

Using the O₂/N₂-summer insolation relationship, we derive an improved chronology of the Vostok ice core. The Vostok CH4 record plotted vs. the O₂/N₂ chronology is consistent with absolutely dated speleothem records, validating our new chronology. The ages for the last four glacial terminations in Vostok correspond to high obliquity (> 23.7° at terminations midpoints) and decreasing precession index (increasing boreal summer insolation). The Vostok temperature record, boreal summer insolation, and the rate of change of ice volume with respect to time (as represented by the SPECMAP record) are highly coherent in the precession and obliquity bands. These three properties vary almost synchronously, with a possibility that Vostok temperature lags behind the other two. Our new timescale supports the idea that boreal summer insolation played an important role in glacial-interglacial cycles.