Holocene Delevelling of Devon Island, Arctic Canada: Implications for Ice Sheet Geometry and Crustal Response1
Total Page:16
File Type:pdf, Size:1020Kb
Color profile: Disabled Composite Default screen 885 Holocene delevelling of Devon Island, Arctic Canada: implications for ice sheet geometry and crustal response1 Arthur S. Dyke Abstract: The raised beachesDyke and deltas of Devon Island contain an abundance of dateable materials. A large set of radiocarbon dates (228), 154 of which are new, are used to construct relative sea level curves and isobase maps for the island. The best materials for this purpose are driftwood logs (61 dates) and bowhead whale bones (74 dates) from raised beaches and mollusc shells from marine-limit deltas (20 dates) or from altitudes close to marine limit (14 dates). During the last glacial maximum, the island is thought to have lain beneath the southeastern flank of the Innuitian Ice Sheet. The relative sea level history is congruent with that inferred ice configuration. The island spans half the ice sheet width. Relative sea level curves are of simple exponential form, except near the glacial limit where an early Holocene emergence proceeded to a middle Holocene lowstand below present sea level, which was followed by submergence attending the passage of the crustal forebulge. The response times of relative sea level curves and of crustal uplift decrease from the uplift centre toward the limit of loading, but the change appears strongest near the limit. The Innuitian uplift is separated from the Laurentide uplift to the south by a strong isobase embayment over Lancaster Sound. Hence, ice load irregularities with wavelengths of about 100 km were large enough to leave an isostatic thumbprint in this region of the continent. The apparent absence of a similar embayment over Jones Sound probably indicates a greater Late Wisconsinan ice load there, or a thicker crust than in Lancaster Sound. Résumé : Les plages soulevées et les deltas de l’île Devon renferment une abondante quantité de matériel apte à être datée. Un groupe de nombreuses datations au radiocarbone (228), incluant 154 nouvelles déterminations, a été utilisé pour tracer les courbes du niveau marin relatif et pour dresser les cartes des isobases de l’île. Les matériaux qui fournissent les meilleurs résultats pour cette étude sont les billes de bois échappé (61 âges) et les os de baleine franche (74 âges) dans les dépôts de plages soulevées et les coquilles de mollusques trouvées à la limite marine des deltas (20 âges) ou aux altitudes proches de la limite marine (14 âges). Nous croyons que durant le dernier maximum glaciaire l’île se trouvait sous le flanc sud-est de l’Inlandsis innuitien. L’histoire du niveau marin relatif est compatible avec la configuration déduite de l’étude de l’évolution de l’inlandsis. L’étendue de l’île représente la moitié de la largeur de la calotte glaciaire. Les courbes du niveau marin relatif prennent une simple forme exponentielle, sauf à la limite glaciaire où une émergence durant l’Holocène précoce a entretenu un stade de bas niveau durant l’Holocène moyen en dessous du niveau marin actuel qui fut suivi d’une submergence ennoyant le passage de la zone de bombement crustal. Les temps de réponse des courbes du niveau marin relatif et du soulèvement crustal décroissent en se déplaçant du centre de soulèvement vers la zone limite de charge des glaces, cependant le changement semble avoir été plus rapide à proximité de la limite. Le soulèvement innuitien est séparé du soulèvement laurentidien au sud par un enfoncement prononcé de l’isobase sur le détroit de Lancaster. Donc, les irrégularités de la charge glaciaire avec longueurs d’onde d’environ 100 km, étaient suffisamment importantes pour avoir laissé leur empreinte dans cette région du continent. L’absence apparente d’un enfoncement similaire de l’isobase sur le détroit de Jones indique que la charge glaciaire au Wisconsinien tardif était probablement plus élevée à cet endroit, ou bien la croûte était plus épaisse dans le détroit de Lancaster. [Traduit par la Rédaction] 904 (1970), for example, proposed that the Innuitian Ice Sheet covered the Queen Elizabeth Islands to account for the Much of the discussion of the extent of glaciation during broad, arch-like pattern of regional uplift. A similar, but the Last Glacial Maximum (LGM) in the Queen Elizabeth more qualitative, assessment had been advanced earlier Islands (Fig. 1) has involved postglacial rebound. Blake (Washburn 1947; Wickenden 1947). However, portrayals of the regional pattern of uplift have remained rather fluid, as Received November 29, 1997. Accepted April 1, 1998. reviewed by Dyke (1998). Furthermore, the interpretation of uplift became controversial when advances in the theory of A.S. Dyke. Terrain Sciences Division, Geological Survey of glacioisostasy allowed nonunique ice load distributions to Canada, 601 Booth Street, Ottawa, ON K1A 0E8, Canada explain the uplift pattern. The critical recognition was that (e-mail: [email protected]). ice sheets depress the crust about 200 km beyond their mar- 1Geological Survey of Canada Contribution 1997203; Polar gins because of the stiffness of the crust (Walcott 1970). As Continental Shelf Project Contribution 00298. a consequence, relative sea level (RSL) at an ice margin at Can. J. Earth Sci. 35: 885–904 (1998) © 1998 NRC Canada 1 I:\cjes\cjes35\cjes-08\E98-034tfh.vp Thursday, December 03, 1998 11:30:39 AM Color profile: Disabled Composite Default screen 886 Can. J. Earth Sci. Vol. 35, 1998 Fig. 1. Location of the study area (shaded) and principal place names. 120° 60° 84° 84° RSL CURVES 0 km 300 LR Lyall River Nansen Arctic Sound Ellesmere s BB Bere Bay Ocean nd Island a t sl i I a h r PR Port Refuge et t Greenland ab S liz s E e n r PAB Prince Alfred Bay uee a Q N TB Trito n Bay OP Owen Point PM Providence Mountain Devon LP Lovell Point Island CAP Colin Archer Peninsula Parry Channel Baffin RB Radstock Bay Bay TLI Thomas Lee Inlet Victoria FP Firkin Point Island Baffin CH Cape Hardy Island CB Croker Bay 66° 66° 120° 60° Norwegian Cornwall Bay Makinson Island Graham Inlet Island 050km Belcher Chan nel Ellesmere Hell Gate Island Peninsula LR North Kent Simmons Island GrinnellBB N Peninsual Barrow TB South Cape Harbour Colin Archer Fiord PAB Peninsula Goose Cape Sheills Fiord Storm Q Peninsula PR Eidsbotn CAP u Jones Sound e Dundas OP Fiord CH e n Baillie- s Island C Hamilton Viks Thomas Lee h Island FP Truelo ve Bathurst an W Fiord Inlet Baffin ne Sverdrup Lowland l e PM Inlet l l i n Bay g TLI Island t o n Devon Island Cornwallis C h a Island n LP CB n e RB Croker l Maxwell Cuming Viscount Radstock Bay Bay Inlet Melville Gascoyne Bay Barrow Strait Inlet r So und Sound Lancaste Parry Channel t le In y Bylot lt a Prince ir Island Somerset m Borden Prince of Wales Regent Brodeur d Island A Peninsula Island Peel Inlet Peninsula Sound © 1998 NRC Canada 2 I:\cjes\cjes35\cjes-08\E98-034tfh.vp Thursday, December 03, 1998 11:30:43 AM Color profile: Disabled Composite Default screen Dyke 887 equilibrium depression can be on the order of 100 m. Partly dated to the Early Wisconsinan (Klassen 1993; McCuaig for this reason, an alternative interpretation of the 1994). It possibly terminated in an ice shelf in Lancaster postglacial uplift of the Queen Elizabeth Islands was pro- Sound (Dyke and Prest 1987). Similarly, the extent of LGM posed (England 1976a), wherein the Laurentide and Green- ice in Jones Sound remains geologically undefined. How- land ice sheets each depressed swaths of the Queen ever, the RSL data presented below bear importantly on this Elizabeth Islands in their forefields and a series of noncon- question. tiguous intervening ice caps, named the Franklin Ice Com- Deglaciation was underway in the eastern coastal areas by plex, accounted for the rest of the observed crustal 10 000 radiocarbon years BP (10 ka BP). The sea had pene- deflection. Thereafter, the Queen Elizabeth Islands uplift trated to the head of Jones Sound by 9.3 ka BP and to Nor- pattern constituted inadequate proof of the general configu- wegian Bay by 9.2 ka BP. The Wellington Channel coast ration of the LGM ice load (Tushingham 1991; England et was deglaciated about 8.2 ka BP. After clearing of the chan- al. 1991). Indeed, Boulton (1979) even went so far as to in- nels, ice on Devon Island retreated to final remnants scat- sist that the inference of ice sheet configuration from RSL tered along the LGM ice divide. These vanished about 8 ka data, with specific reference to the Queen Elizabeth Islands, BP. The Devon Ice Cap did not disappear during the early was a “misuse of data.” Clearly, therefore, more direct gla- Holocene recession, but it probably shrank to become much cial geological evidence was required to demonstrate what smaller than its present size (Dyke 1998). this configuration might be. A further caution was raised when England (1987; see also England 1997) proposed that neotectonics may have influenced the Holocene uplift pat- tern. Similarly, Dyke et al. (1991) proposed that the strong Methods ridge pattern of early Holocene shoreline deformation over Prior to fieldwork, marine-limit deposits and features the structural Boothia Arch and vicinity indicated a large were mapped from airphotos and the areas of best-developed tectonic complication of the glacial rebound pattern. Be- raised beaches were identified. In the field, elevations were cause no alternative interpretation of that ridge or associated determined by Wallace and Tiernan surveying altimeters us- features has been advanced, we should remain cautious in ing shortest possible closure times between sample sites and assuming that Holocene uplift in glaciated Canada is every- sea level (Dyke et al.