TephrochronologyTephrochronology ( Gk ‘ashes’) David J. Lowe

1. What is tephrochronology? 2. How are dated? – Bayesian improvements 3. Identifying uncertainties – Six steps to miscorrelation

Photo: M. Gehrels 1.1. What What is is tephrochronologytephrochronology??

 A stratigraphic linking, dating & synchronizing tool – for palaeoenvironmental, geological or archaeological sequences or events

 Linking/correlating tephras relies on: – Statigraphic position (  relative age) – Characterisation using physical, mineralogical and geochemical properties = ‘fingerprinting’ (fingerprints must be diagnostic, ideally persistent) – Numerical age

 Hence unique power (advantages) of tephrochronology: – Deposits and palaeoarchival information positioned precisely on common scale using tephra isochron as stratigraphically fixed tie-point  even if age unknown or imprecise – Numerical age acquired for tephra, or relative age, can be transferred from one site to next because tephra eruption age geologically ‘instantaneous’ – Provides independent test of sequence ExampleExample 1: 1: FugloyarbankiFugloyarbanki tephratephra ()(Iceland)

Fugloyarbanki tephra Ice-core age: 26,740 ± 390 yr before 2000 AD Hence F tephra: (1) - Correlated ~3000 km (from Siwan Davies et al. 2008 JQS) - NGRIP  age & environ. proxies - Marine cores  environ. proxies - Ice-core age vs marine 14C ages  correct LGM ExampleExample 2: 2: Kawakawa Kawakawa tephratephra (New(New Zealand) Zealand)

Kk

Galway Tarn Photo: M. Vandergoes

Correlated ~1000 km Kohuora crater, Auckland KawakawaKawakawa tephratephra Links palaeoarchival records to common tie-point in Auckland (A)  Taranaki (B)  West Coast (C)

- Even though age imprecise 27,097 ± 957 cal yr BP (2)

Grass peaks Qtz peaks = coldest = coldest

Mild

Beetle study shows warm interstadial Mild

(After Rewi Newnham et al. 2007 JQS) 2.2. How How are are tephrastephras dated?dated?

 Historical  Radiometric – 14C – Fission track esp. isothermal plateau FTD on glass (also FTD on zircon) – 40Ar/39Ar – Luminescence (TL, OSL)  Incremental – – Varved sediments – Ice-core layering  Age equivalent – Palaeomagnetism – Correlation with marine oxygen isotope stages or biostratigraphy (e.g. palynostratigraphy) NZNZ--INTIMATEINTIMATE records records linked linked with with tephratephra isochronsisochrons

HOL

LGIT

eLGM

 Improve ages on tephras at Kaipo bog sequence using OxCal and Bpeat Upper section Kaharoa Kaipo bog

Taupo Section

Waimihia

Whakatane Tuhua Mamaku Rotoma

Opepe

 Poronui  Karapiti  Okupata  Konini

Grey mud  Poronui  Karapiti  Okupata  Konini

Peaty mud B Grey mud

Peaty mud A Waiohau

 Rotorua Mud  Rerewhak.

Lower section Chronology of Kaipo - 3 sets of 14C dates

20 radiometr. 14C 14 dates on tephras ages from Waikato (via tephrochronology) lab in red in black (Froggatt & Lowe 1990, - via Dr Alan Hogg Lowe et al. 1999)

20 AMS 14C ages from ETH lab (Switzeland) in blue - via Dr Irka Hajdas

Total of 51 independently dated points

(Lowe et al. 1999 NZJGG, Irka Hajdas et al. 2006 QR) Age model

Depth chronology grey scale likelihoods (darker = more likely ages)

Interpolated dates via Bpeat - maximum Kaipo bog sequence posterior densities of chronological fitted to INTCAL04 ordering-constrained using Bpeat & calibration ranges stratigraphic ordering (red = 100% outlier) (Dr Maarten Blaauw) TephraTephra ageage models models derived derived from from KaipoKaipo BpeatBpeat analysisanalysis

Volcanic centres Point age estimates from 9490 Red = Okataina single-best Bpeat iteration Yellow = Taupo 9488-9518 Bpeat 95.4% HPD (2 range) Green = Tong./Egmont 9480-9530 Purple = Tuhua (Mayor Is) OxCal 95.4% HPD (2 range)

9490 11,510 9488-9518 11,420-11,661 Okupata 9480-9530 11,430-11,810 11,180 8000 11,118-11,239 7979-8039 3390 Konini 11,560 11,110-11,270 7960-8050 3362-3422 11,480-11,752 3370-3450 11,500-11,940 9980 5490 9699-10,182 5474-5565 9920-10,230 5470-5590 13,670 13,502-13,744 6910 13,470-13,800 6832-7013 15,450 6850-7160 15,192-15,524 11,320 17,760 15,100-15,750 5220 11,269-11,450 17,200-18,050 5021-5233 11,220-11,600 17,426-17,999 4970-5270

(After Lowe et al. QSR 2008 with help of Dr Maarten Blaauw) 3.3. Identifying Identifying uncertainty uncertainty

 MiscorrelationMiscorrelation of of tephratephra layerlayer may may  –– IsochronIsochron drawndrawn at at incorrect incorrect position position –– MisassociationMisassociation ofof palaeoclimaticpalaeoclimatic oror archaeologicarchaeologic eventsevents = = incorrect incorrect linking linking//synchronizationsynchronization –– IncorrectIncorrect age age transferred transferred

 SixSix steps steps to to uncertainty uncertainty…… UncertaintyUncertainty in in tephrochronologytephrochronology

1.1. Faulty Faulty characterisation characterisation –– AnalyticalAnalytical data data compromised, compromised, e.g. . Unsuitable anal. method used . Block for EMPA (bulk sample vs. single grain) . Glass or crystals not polished . Appropriate standards not used . Incorrect probe conditions (e.g. glass needs de-focussed ~20-m beam)

. Grains weathered Electron microprobe, Victoria Univ. of Wellington UncertaintyUncertainty in in tephrochronologytephrochronology

2.2. Non Non--uniqueunique fingerprints fingerprints –– MajorMajor element element composition composition of of glassglass can can be be identical identical for for differentdifferent tephrastephras . e.g. Holocene Taupo tephras (Stokes et al. 1992 QI) –– NeedNeed additional additional data data . Trace element/rare earth data from single grain LA-ICPMS . Stratigraphic or chronological data UncertaintyUncertainty in in tephrochronologytephrochronology

Eruption North lobe 3.3. Multiple Multiple fingerprints fingerprints duration – Compositional heterogeneity South . Some eruptions tapped 23 magmas lobe simultaneously or sequentially  compositions vary in different fallout directions Biotite analyses - Kaharoa tephra  inadequate to characterise few samples (Phil Shane et al. 2008 QI)

from limited dispersal directions Photo: . May have miscorrelated tephras in past M. Gehrels . How many grains need we analyse? – Wide variability andesitic glass composition from crystal inclusions (microlites) . Can now correct using petrologic method Andesitic glass + microlites UncertaintyUncertainty in in tephrochronologytephrochronology

4.4. Reworking Reworking to to different different stratigraphicstratigraphic positionposition – Results in wrong positioning of isochron – Dissemination of grains through sediment . e.g. Shard concentration zone in peat: where is isochron ‘line’? – Tephra sinking in v. soft lake sediments

 How to recognise reworking? – Grain character – Stratigraphic/sedimentary context – Multiple populations shown by single-grain analyses – Reproducibility Tp Photos: M. Gehrels UncertaintyUncertainty in in tephrochronologytephrochronology

5.5. Dating Dating ‘‘errorserrors’’  WrongWrong age age transferred transferred – Misassociation – Contamination . e.g. Old carbon in-wash in lakes; young carbon migration; in-built age; hard-water, marine reservoir effects – Large counting errors (low precision) – Annealing problem glass fission-track dating  age underestimated, now corrected ITPFT ZFTD UncertaintyUncertainty in in tephrochronologytephrochronology

Kawakawa 6.6. Statistical Statistical misadventure misadventure Kk – Miscorrelation because database – Miscorrelation because database Okaia O not comprehensive or poor quality data (etc) . e.g. DFA generally OK (provides Rotoehu Re probability of miscorrelation, Mahalanobis distance stat.) but dependent on quality & comprehensiveness of training sets – Some tephras very similar compositionally, others separable, e.g. Kk vs Re vs O

DFA from Shane Cronin et al. 1997 NZJGG ReducingReducing uncertainty uncertainty

 NeedNeed – Multiple criteria = safer correlation – Rigorous analyses, standards must be documented, uncertainty reported – Ideally use comprehensive sequences esp. with interfingering tephras from different sources . Distal can be best, e.g. lake sediments, marine cores, ice cores

 EvenEven if if age age on on tephratephra isis imprecise, imprecise, correct correct characterisationcharacterisation means means isochronisochron stillstill transferrable transferrable – Effectively tephra tie-point fixed (stratigraphic position) on ‘elasticated’ chronology