Quaternary tephrochronology in Iceland dating principles & applications

Abstract

This thesis presents multi- method numerical ages and age estimates obtained from the previously undated Ókoli Tephra of the Vestfírôir Peninsula, NW Iceland and the Þórsmörk Ignimbrite (ÞIG), Eastern Volcanic Zone, S. Iceland. These provide unique chronological constraints for Quaternary landscape evolution, ice sheet extent and ice -sheet thickness. In addition, they allow ice -free area hypotheses in Iceland to be tested and terrestrial and marine records of palaeoenvironmental change in the North Atlantic region to be linked more securely. The 2.26±0.11 Ma fission-track age of the OÞoli Tephra is particularly important because it places a major new age restriction on glacially driven, macroscale landscape evolution processes of fjord network formation on the Vestfírdir Peninsula.

The pre- radiocarbon, Quaternary era is a vital period in which current glacial /interglacial cycles developed and caused widespread environmental change. Dating controls for this era are limited because of widespread glacial erosion, but in Iceland certain tephra deposits have survived glacial overriding and can provide accurate age constraints and precise spatial correlation for stratigraphic sequences. The two tephra deposits chosen for this study are of great palaeoenvironmental significance. The 013oli Tephra rests unconformably 580 -600 m.a.s.l. near the plateau surface of Skagafjall (NW Iceland). The 100 -160 m thick sequence of ice -dammed lake deposits beneath it were thought to have formed `20- 17,000' years ago. The ÞIG (Eastern Volcanic Zone, EVZ, S. Iceland) is the largest Quaternary -age ignimbrite deposit in Iceland and glass compositions have been geochemically linked with the highly silicic (Si0₂ >68 wt %) component of North Atlantic Ash Zone-2 (NAAZ-2), 48-58 ka.

Correlation, fission -track (FT) and thermoluminescence (TL) methods were used to produce ages from the glass shards that overwhelmingly dominate the highly silicic airfall and /or pyroclastic flow ash components of these deposits. Dating glass shards is advantageous because they form during the magma -quenching phase of an eruption and nearly always reflect the `true' eruption age. They are also the most far- travelled and geochemically diagnostic phase of Icelandic tephra. Geochemical correlation age estimates were obtained for the Ókoli Tephra and ÞIG by applying multivariate statistical methods to new and published glass -shard data from 104 highly silicic tephra layers /microtephra horizons of Tertiary- Quaternary age from various marine /terrestrial repositories in and around the northern North Atlantic. The isothermal plateau and diameter corrected fission -track (ITPFT & DCFT) methods used are well -established, grain specific, glass -phase dating techniques that produce accurate ages from tephra deposits older than ca. 70 ka. A new coarse -grain (90 -150 gm), glass -phase (2.3 -2.4 gcm⁻³/<2.4 gcm⁻³) TL dating method was developed to take advantage of the glass -rich, phenocryst and micro -inclusion poor nature of the selected deposits. To reconstruct palaeodose values, the established multiple aliquot additive dose TL (MAAD -TL) procedure and an adaptation of a single aliquot regeneration (SAR) luminescence dating procedure, (SAR -TL), that allows continuous assessment of, and correction for, changes in sample TL sensitivity, were used on the Okoli Tephra. The NG, with a ca. 50 ka correlation age estimate, was used as a pseudo -age control for the SAR -TL experiment.

Seven correlation age matches ranging from ca. 100 ka -13 Ma were found for the Ókoli Tephra. Composite correlation age matches for the NG are all in the 48.5 -58 ka age range with 48.5 ka the best correlative. Chemical correlation dating is therefore useful in constraining the age of distinctive tephra deposits in the late Pleistocene, but is not always effective over longer timescales. The weighted mean ITPFT/DCFT age of the Ókoli Tephra is 2.26±0.11 Ma and considered accurate. No spontaneous fission tracks were observed for the kirsmórk Ignimbrite, consistent with ca. 50 ka age and glass shard uranium content of 3-4 ppm. The natural SAR -TL signal is located in a second phase of approximately linear dose response growth of the TL signal beyond the first main phase of signal saturation. This meant age estimates greater than the generally quoted 0.8 Ma upper limit of luminescence dating could be calculated. Eight out of eleven TL dating attempts were successful with an age range of 0.6 -4.8 Ma. The weighted mean age from the three successful MAAD -TL runs with a 5 hour, 155°C pre -heat was 2.06 ±0.27 Ma. Longer, low temperature pre -heats (16 hrs, 135°C) reduced age estimates by approximately half. Chánges in TL sensitivity were observed during all SAR -TL experiments restricting age estimates to minimum -maximum age ranges. Nevertheless, the ÓÞoli Tephra SAR -TL age range of 1.74 ±0.37- 4.69±0.74 Ma (pre-heat: 2 minutes, 220°C) encompasses the fission-track age. The natural, glass -phase TL signal of the ÞIG ash is located in the first phase of linear dose response growth before first phase saturation. The minimum -maximum SAR -TL age range is 29 ±7 -40 ±10 ka. The ca. 40 ka maximum SAR -TL age estimate is significantly less than the minimum correlation age estimate, but considered most representative of the geological age because changes in TL sensitivity were greater than for the Ópoli Tephra. The TL methods used need refining, but these results are highly encouraging for developing and applying glass -phase dating procedures across a broader than expected chronological range.

The new numerical ages and context of the Ókoli Tephra suggest a proto-fjord network of similar configuration to the current fjord system had formed early in the Quaternary in NW Iceland and that Iceland-wide glacial activity in the first part of the early Pleistocene (2.5-2.0 Ma) was more extensive and /or more widely distributed than previously thought. The Óboli eruption is tentatively linked to the Húsafell volcanic complex on the western mainland and is thought to be one of the first deglaciation induced, massive eruption events of the Pleistocene. The Skagafjall deposits were preserved beneath cold -based non -erosive parts of overriding mid -late Pleistocene ice -sheets. Large TL age errors, caused by uncertain palaeowater content estimates, currently prevent the use of TL dating to establish if the I örsmórk Ignimbrite was deposited by single or multiple eruption events 60 -30 ka. Nevertheless, new ca. 40 ka SAR -TL and 48.5 ka correlation ages of the NG provide a critical 40 -50 ka age constraint on glacial and volcanic processes in the EVZ of southern Iceland. They suggest that volcanic activity in this region was responsible for at least one component of NAAZ -2. The ca. 40 ka SAR -TL dated deposit is an age underestimate or could be the second of at least two discrete 60 -30 ka eruptions in the EVZ of similar magnitude and geochemistry. These eruptions were massive events, possibly triggered by abrupt climatic upturns between 60 -30 ka during the rapidly fluctuating, but generally downward climatic trend towards the LGM.