Abstract: Fjord valleys are carved during glaciation and then form local sediment sinks, which fill during retreat of the ice. Thus fjord valleys appear analogous to lower-latitude incised valleys, but they are remarkably different because fjords experience isostatic rebound during deglaciation, causing relative sea level to fall during infill. This paper explores stratigraphic architecture of fjord valley fills based on Late Quaternary deposition in Clyde Inlet, Baffin Island, Arctic Canada, as constrained by 11 cosmogenic dates and 9 accelerator mass spectrometry (AMS) 14 C datings. A major ice stream of the Laurentide Ice Sheet occupied Clyde Inlet at last glacial maximum and bulldozered through a U-shaped valley forming a lower sequence boundary. During the Early Holocene the system entered a deglacial stage; tidewater glaciers retreated rapidly (>100 km in 1000 yrs) through the fjord from 10.4 ka onward. Grounded ice lobes started retreating from the Clyde fjordhead by 9.4 ka. Then ice-contact fans (ICF) were deposited consisting of flat-topped fan deltas, covered with channels and boulder-strewn bars. Elevations of the surfaces vary between 62 and 77 m above sea level, which marks the relative sea level at the time of deposition and is considered to be the marine flooding surface. Marine muds have been draped directly onto the ICF complexes. Subsequently, coarse-grained glaciofluvial valley trains (GFVTs) prograde downstream caused by rapid base-level fall, despite possibly high sediment supply (i.e., forced regression). During the Late Holocene (3.5 ka) the last remaining lobes of the Laurentide Ice Sheet retreated from the middle parts of Clyde River basin to form the present Barnes Ice Cap. At this phase, the rate of base-level fall has decreased (~1.6 m/ka over the last 3.5 ka), still the river incises significantly, marking a reduced sediment supply. Narrow coarse sandy fluvial terraces were being deposited at the lowest level of the incised river valley. Clyde fjordhead may not have entered a postglacial stage by definition, nevertheless a strongly reduced sediment flux is apparent. Numerous upland lakes likely play a role in trapping sediment in the hinterland. In addition, we speculate that the glacial regime of the Barnes Ice Cap switched from a sediment producing regime to a nonerosive cold-based regime. In conclusion, stratigraphic patterns of valley fills in high-latitude areas display an evident signature of isostatic rebound and a strongly varying sediment supply. Rapid uplift causes ice proximal units to occur high in the infill and reverses classic fining upward valley fill sedimentary trends. The exact interplay of local sea-level change and sediment supply dictates the complexity of the valley fill, but coarsening upward trends with younger sandy fluvial deposits incising into the fill deposits ultimately have important implications for the interpretation of similar deglacial valley fill settings.

Abstract: Climate plays a significant role in determining the styles of depositional processes at different latitudes, which in turn influence the locations of hydrocarbon systems. Results of climate modeling may therefore provide important information for predicting the presence or absence of suitable hydrocarbon plays. To determine whether the models provide realistic results, the critical step is to validate the model results against proxy data where they are available. Paleoclimate proxy data are most often derived from more accessible low- to midlatitude regions and are biased towards warm climate states. However, general circulation models (GCMs) have traditionally been biased to colder temperatures, in particular at high-latitudes, struggling to maintain the high-latitude regions warm enough to sustain forests that were present during greenhouse periods, such as the mid-Cretaceous (~130–89 Ma), without exaggerated warming of the equatorial regions. To improve this approach, the HadCM3L coupled atmosphere–ocean GCM, a state-of-the-art model for the long simulations required to reach an equilibrium climate, was run for each stage of the Cretaceous using new paleogeographic base maps. Here, we compare the results for the Aptian (118.5 Ma) and Albian (105.8 Ma) with paleoclimate proxy data from the high northern latitudes in order to determine if the model produces viable results for this region. Paleoclimate analysis of fossil wood from conifer forests from Svalbard of Aptian–Albian age suggests that they grew in moist cool upland areas adjacent to warmer temperate lowland regions, probably with rivers and/or swamps present. Studies of conifers from the Canadian Arctic islands indicate that they grew under slightly cooler conditions than on Svalbard, similar to northern Canada today. The HadCM3L GCM results for Svalbard show that the dominant biome was evergreen taiga/montane forest with lowland temperate vegetation present during the Albian Stage, possibly with an element of deciduous taiga/montane forest in the Aptian (both cold boreal forests with short hot summers according to the Köppen–Geiger classification). The modeled mean annual temperature was ~−3.7° C at the sample sites, with summer temperatures rising to a mean of ~18° C during the Albian. Mean annual precipitation was ~571 mm. In the Canadian Arctic, the model results indicate that the biomes were more mixed than on Svalbard. The Aptian biome was dominantly deciduous taiga/montane forest with temperate vegetation in low-lying areas. The Albian landscape was dominated by evergreen taiga/montane forest, with some elements of deciduous taiga. Both stages were classified as cold boreal forest with short hot summers under the Köppen-Geiger classification scheme. Mean annual temperature was modeled to be ~−6.5° C at the sample sites, with summer temperatures reaching a mean of ~13° C, and mean annual precipitation was ~406 mm. These results suggest that the HadCM3L GCM, coupled with updated paleogeographic maps, can produce a good match to the climate proxy data in these difficult-to-model high-latitude areas.