Maps showing the thickness of sediments above the bedrock (depth to bedrock, or DTB) are important for many geoscience studies and are necessary for many hydrogeological, engineering, mining, and forestry applications. However, it can be difficult to accurately estimate DTB in areas with varied topography, like lowland and mountainous terrain, because traditional methods of predicting bedrock elevation often underestimate or overestimate the elevation in rugged or incised terrain. Here, we describe a machine learning spatial prediction approach that uses information from traditional digital elevation model derived estimates of terrain morphometry and satellite imagery, augmented with spatial feature engineering techniques to predict DTB across Alberta, Canada. First, compiled measurements of DTB from borehole lithologs were used to train a natural language model to predict bedrock depth across all available lithologs, significantly increasing the dataset size. The combined data were then used for DTB modelling employing several algorithms (XGBoost, Random forests, and Cubist) and spatial feature engineering techniques, using a combination of geographic coordinates, proximity measures, neighbouring points, and spatially lagged DTB estimates. Finally, the results were contrasted with DTB predictions based on modelled relationships with the auxiliary variables, as well as conventional spatial interpolations using inverse-distance weighting and ordinary kriging methods. The results show that the use of spatially lagged variables to incorporate information from the spatial structure of the training data significantly improves predictive performance compared to using auxiliary predictors and/or geographic coordinates alone. Furthermore, unlike some of the other tested methods such as using neighbouring point locations directly as features, spatially lagged variables did not generate spurious spatial artifacts in the predicted raster maps. The proposed method is demonstrated to produce reliable results in several distinct physiographic sub-regions with contrasting terrain types, as well as at the provincial scale, indicating its broad suitability for DTB mapping in general.

Basal gravel and sand mantling the bedrock floors of buried valleys throughout the Canadian Interior Plains, and conformably overlying proglacial lacustrine sediment, comprise the Empress Group. While previously conceptualized as stratigraphically equivalent deposits of preglacial rivers prior to the first and most extensive continental and montane glaciations, subsequent stratigraphic studies indicated that buried valley basal gravel must have been deposited between, or during, progressively more extensive continental glaciations and could not be stratigraphically equivalent throughout the buried valley network. However, in the general absence of formation-rank stratigraphic description of basal gravel units that might better inform the geologic history of the deposits, most workers simply consider Empress Group sediments time-transgressive. In this paper, we examine basal gravel at provincial and regional scales to understand its genesis and geologic history. At the provincial scale, we map basal gravel in three dimensions using a novel machine learning approach. At the regional scale, we formally define basal gravel formations at either end of the largest buried valley system in Alberta, which informs its glacial history and physiographic development and shows the importance of formation-rank stratigraphic description. Our results indicate that the buried valley network across Alberta is palimpsest in genesis and basal gravel units within it are chronostratigraphically intercalated between tills. We advocate that the Empress Group definition be extended across Alberta with modifications to improve its clarity and utility, and formally define the Old Fort, Unchaga, Ipiatik, and Winefred formations as part of the Empress Group.

The North American Ice Sheet Complex (NAISC; consisting of the Laurentide, Cordilleran and Innuitian ice sheets) was the largest ice mass to repeatedly grow and decay in the Northern Hemisphere during the Quaternary. Understanding its pattern of retreat following the Last Glacial Maximum is critical for studying many facets of the Late Quaternary, including ice sheet behaviour, the evolution of Holocene landscapes, sea level, atmospheric circulation, and the peopling of the Americas. Currently, the most up-to-date and authoritative margin chronology for the entire ice sheet complex is featured in two publications (Geological Survey of Canada Open File 1574 [Dyke et al., 2003]; 'Quaternary Glaciations - Extent and Chronology, Part II' [Dyke, 2004]). These often-cited datasets track ice margin recession in 36 time slices spanning 18 ka to 1 ka (all ages in uncalibrated radiocarbon years) using a combination of geomorphology, stratigraphy and radiocarbon dating. However, by virtue of being over 15 years old, the ice margin chronology requires updating to reflect new work and important revisions. This paper updates the aforementioned 36 ice margin maps to reflect new data from regional studies. We also update the original radiocarbon dataset from the 2003/2004 papers with 1541 new ages to reflect work up to and including 2018. A major revision is made to the 18 ka ice margin, where Banks and Eglinton islands (once considered to be glacial refugia) are now shown to be fully glaciated. Our updated 18 ka ice sheet increased in areal extent from 17.81 to 18.37 million km(2), which is an increase of 3.1% in spatial coverage of the NAISC at that time. Elsewhere, we also summarize, region-by-region, significant changes to the deglaciation sequence. This paper integrates new information provided by regional experts and radiocarbon data into the deglaciation sequence while maintaining consistency with the original ice margin positions of Dyke et al. (2003) and Dyke (2004) where new information is lacking; this is a pragmatic solution to satisfy the needs of a Quaternary research community that requires up-to-date knowledge of the pattern of ice margin recession of what was once the world's largest ice mass. The 36 updated isochrones are available in PDF and shapefile format, together with a spreadsheet of the expanded radiocarbon dataset (n = 5195 ages) and estimates of uncertainty for each interval. (C) 2020 Elsevier Ltd. All rights reserved.

The extent of proglacial lakes following the initial separation of the southwest Laurentide Ice Sheet from the Cordilleran Ice Sheet and its eastward retreat from the Canadian Rocky Mountains has been reconstructed across regions of Alberta at a range of scales. However, to date, no studies have integrated all available geological information to produce a province-wide deglacial reconstruction that considers the evolution of proglacial lakes as components of the ice-marginal system. In this paper, we utilize a geologically constrained shoreline projection method with a high resolution digital elevation model to reconstruct the evolution of the ice-marginal system along the southwest LIS during the last deglaciation. This method provides new details on the configuration, volume, drainage history and routing of similar to 240 proglacial lakes as they migrated across Alberta and establishes a succession of paleogeographic reconstructions that can place geological evidence of regional and local ice-flow reorganizations into a spatiotemporal context. These reconstructions demonstrate that although the evolution of proglacial lakes was largely driven by the topography of the emerging landscape and the configuration of the ice margin, positive feedbacks in the ice-marginal system, particularly where margins transitioned from terrestrial to subaqueous settings played a major role in deglacial ice dynamics. Narrow, ribbon-shaped lakes that paralleled the ice margin induced relatively minor changes in style and rate of deglaciation, whereas the evolution of progressively larger, longer-lived lakes extending obliquely to the ice margin promoted surging and subsequent rapid retreat. Crown Copyright (C) 2019 Published by Elsevier Ltd. All rights reserved.

Any process that causes a sudden brittle failure of material has the potential to cause earthquake-like seismic events. Cryoseisms represent an underreported class of seismic event due to their (often) small magnitudes. In this paper, we document the phenomenon of some of the largest magnitude lake-associated icequakes (ML2.0) yet reported. These events occurred nearly simultaneously (within ∼2 h) on geographically separate lakes in Alberta, Canada, starting 1 January 2018. We conjecture that these events were caused by the sudden brittle failure of lake ice due to thermal expansion; the effects of the thermal expansion were compounded by the lack of insulating snow cover, high lake water levels, and a rapid onset of atmospheric warming. These factors also contributed to ice-jacking — a repeating process in which thermal contraction produces tensile cracks (leads) in lake ice that are then filled with water that is frozen during the cooling cycle. Thus, any subsequent thermal expansion must be accommodated by new deformation or brittle failure. This ice-jacking process caused creeping ground deformation after the initial brittle failure and again two weeks later following a second warming period. In many cases, the resulting ground deformation was significant enough to cause property damage.

Glacial lake outburst floods occurred frequently during the last deglaciation of the Laurentide Ice Sheet. Within the Interior Plains, these floods carved large spillway systems; however, due to a lack of abundant sediment, deposits within prairie spillways are rarely preserved. Here, we present geomorphic and sedimentary evidence and hydraulic modelling of the eastern Beaver River Spillway, formed by the catastrophic drainage of the ice‐dammed glacial Lake Algar, in north central Alberta. During this flood, coarse‐grained sediment eroded from local till formed large pendant bars. Within the first ~50 km of the spillway (Reach 1), pendant bars contain downstream orientated foresets overlain by horizontally bedded coarser gravels. The remaining pendant bars (Reach 2), present downflow of a moraine barrier, differ, comprising massive, matrix‐supported, inversely graded gravels capped by a boulder layer. We use a HEC ‐Geo RAS / HEC ‐ RAS system in conjunction with palaeostage indicators to estimate the steady‐state water surface elevation. Modelling results show that peak discharge within Reach 1 of the eastern Beaver River Spillway was approximately 14 000–21 000 m 3 s −1 . For Reach 2, 30 km downstream, the peak discharge was estimated at 23 000–40 000 m 3 s −1 (n bulked 18 000–26 000 m 3 s −1 ). The downstream discharge increase, consistent with the sedimentary change in pendant bar deposits, is attributed to sediment bulking of the flood flow. This provides the opportunity to observe a range of flow conditions, and associated sedimentology, from a single flood event. The reconstructed flow conditions, coupled with lake volume estimates from the ponding above the moraine barrier suggest a minimum flow duration of 3–5 days.

Ice sheets that advance upvalley, against the regional gradient, commonly block drainage and result in ice‐dammed proglacial lakes along their margins during advance and retreat phases. Ice‐dammed glacial lakes described in regional depositional models, in which ice blocks a major lake outlet, are often confined to basins in which the glacial lake palaeogeographical position generally remains semi‐stable (e.g. Great Lakes basins). However, in places where ice retreats downvalley, blocking regional drainage, the palaeogeographical position and lake level of glacial lakes evolve temporally in response to the position of the ice margin (referred to here as ‘multi‐stage’ lakes). In order to understand the sedimentary record of multi‐stage lakes, sediments were examined in 14 cored boreholes in the Peace and Wabasca valleys in north‐central Alberta, Canada. Three facies associations ( FAI – III ) were identified from core, and record Middle Wisconsinan ice‐distal to ice‐proximal glaciolacustrine ( FAI ) sediments deposited during ice advance, Late Wisconsinan subglacial and ice‐marginal sediments ( FAII ) deposited during ice‐occupation, and glaciolacustrine sediments ( FAIII ) that record ice retreat from the study area. Modelling of the lateral extent of FA s using water wells and gamma‐ray logs, combined with interpreted outlets and mapped moraines based on Li DAR imagery, facilitated palaeogeographical reconstruction of lakes and the identification of four major retreat‐phase lake stages. These lake reconstructions, together with the vertical succession of FA s, are used to develop a depositional model for ice‐dammed lakes during a cycle of glacial advance and retreat. This depositional model may be applied in other areas where meltwater was impounded by glacial ice advancing up the regional gradient, in order to understand the complex interaction between depositional processes, ice‐marginal position, and supply of meltwater and sediment in the lake basin. In particular, this model could be applied to decipher the genetic origin of diamicts previously interpreted to record strictly subglacial deposition or multiple re‐advances.

The impact of the Laurentide Ice Sheet ( LIS ) deglaciation on Northern Hemisphere early Holocene climate can be evaluated only once a detailed chronology of ice history and sea‐level change is established. Foxe Peninsula is ideally situated on the northern boundary of Hudson Strait, and preserves a chronostratigraphy that provides important glaciological insights regarding changes in ice‐sheet position and relative sea level before and after the 8.2 ka cooling event. We utilized a combination of radiocarbon ages, adjusted with a new locally derived ΔR, and terrestrial in‐situ cosmogenic nuclide ( TCN ) exposure ages to develop a chronology for early‐Holocene events in the northern Hudson Strait. A marine limit at 192 m a.s.l., dated at 8.1–7.9 cal. ka BP , provides the timing of deglaciation following the 8.2 ka event, confirming that ice persisted at least north of Hudson Bay until then. A moraine complex and esker morphosequence, the Foxe Moraine, relates to glaciomarine outwash deltas and beaches at 160 m a.s.l., and is tightly dated at 7.6 cal. ka BP with a combination of shell dates and exposure ages on boulders. The final rapid collapse of Foxe Peninsula ice occurred by 7.1–6.9 cal. ka BP (radiocarbon dates and TCN depth profile age on an outwash delta), which supports the hypothesis that LIS melting contributed to the contemporaneous global sea‐level rise known as the Catastrophic Rise Event 3 ( CRE ‐3).

ABSTRACTThis paper presents a reconstruction of the geometry, dynamics and flow pattern of the Laurentide and Cordilleran ice sheets along the Albertan portion of the eastern Canadian Rocky Mountains during the last glaciation. Sediment–landform associations relating to the evolution of these ice sheets document their initial convergence and mutual deflection across west‐central Alberta. The continued advance of Laurentide ice locally displaced Cordilleran ice westward into the Front Ranges, and deflected trunk ice emerging from the Athabasca River valley south‐eastwards along the Foothills. These convergent flow patterns remained major features in the regional geometry of both ice sheets throughout the Last Glacial Maximum. The onset of deglaciation was characterized by the northward retreat of the Laurentide Ice Sheet, which progressively unblocked valleys along the mountain front. The configuration of buttress removal initiated the Lateglacial expansion and sequential drawdown of the eastern Cordilleran Ice Sheet along fast‐flowing outlet glaciers. Subsequent regional retreat of Laurentide and Cordilleran ice was associated with the development of large proglacial lakes which amplified deglaciation by destabilizing the margins of both ice sheets, triggering extensive readvances due to periodic changes in water level associated with the opening and closing of outlets by ice margin fluctuations.

Recent proxy measurements reveal that subglacial lakes beneath modern ice sheets periodically store and release large volumes of water, providing an important but poorly understood influence on contemporary ice dynamics and mass balance. This is because direct observations of how lake drainage initiates and proceeds are lacking. Here we present physical evidence of the mechanism and geometry of lake drainage from the discovery of relict subglacial lakes formed during the last glaciation in Canada. These palaeo-subglacial lakes comprised shallow (<10 m) lenses of water perched behind ridges orientated transverse to ice flow. We show that lakes periodically drained through channels incised into bed substrate (canals). Canals sometimes trend into eskers that represent the depositional imprint of the last high-magnitude lake outburst. The subglacial lakes and channels are preserved on top of glacial lineations, indicating long-term re-organization of the subglacial drainage system and coupling to ice flow.

ABSTRACTLandform mapping near the Athabasca valley was fundamental to determining whether the Late Wisconsinan Laurentide and Cordilleran Ice Sheets (LIS, CIS) coalesced. In this paper we examine the detailed landform record using new LiDAR digital elevation and hillshade models of the area of confluence and eventual desuturing of these ice sheets. This work reveals an earlier more extensive Cordilleran advance before coalescence with the LIS. When the ice sheets coalesced, the flow pattern was dominated by ice flow along the mountain front, with Cordilleran ice flowing out of major trunk valleys but Laurentide ice flowing into the Foothills up smaller valleys. This flow pattern implies that when the ice sheets coalesced the CIS was already waning, or at least was not at its maximum. Deglaciation was interrupted by several re‐advances, probably related to the destabilizing effect of proglacial lakes along the ice margin.

Government geological survey maps and research publications have portrayed the distribution of glacial landforms associated with the advance and retreat of the Laurentide and Cordilleran ice sheets across Alberta at a local, regional, and continental scale. To date, this information has not been systematically synthesized into a single compilation at a consistent scale. Although this original work provided valuable information to constrain reconstructions of former ice sheet extent, configuration, and flow geometry, its derivation primarily from the interpretation of aerial photographs and the Shuttle Radar Topography Mission 90 m digital elevation model (DEM) may result in methodological inconsistencies and spatial biases. These biases, together with challenges associated with geomorphic mapping in densely forested areas of western and northern Alberta limit the usefulness of previous mapping when applied to inversion-based ice sheet reconstructions, which have specific input data demands. Recently, light detection and ranging (LiDAR) DEMs have become increasingly available throughout Alberta. Hill-shaded imagery of these data provides unprecedented geomorphic detail beneath the forest cover and reveals that that the glacial geomorphology of northern and western Alberta is more complex than previously recognized. In this paper, we describe the methodology and geomorphic criteria used to produce a glacial landform map of Alberta using previously published data, supplemented by comprehensive new analysis of high-resolution (2–25 m) DEMs. These include 306 624 km 2 of LiDAR imagery, with which it is now possible to verify and where necessary augment previous mapping, particularly across areas with a dense forest cover.

Radiocarbon ages from Southampton Island (Canada) provide new chronological control on the deglaciation of Foxe Channel and northern Hudson Bay, a strategic area for understanding the demise of a marine-based portion of the Laurentide Ice Sheet. A regional marine reservoir age of 630±45yr and a reservoir offset (ΔR) of 263±48yr were calculated from two early to mid-Holocene terrestrial/marine radiocarbon age pairs. These values are consistent with corrections based on early 20th century mollusks suggesting that following deglaciation the oceanic conditions controlling the regional reservoir effect rapidly became similar to those of modern times. However, our ΔR value is 352±52yr less than another correction from eastern Foxe Basin, which may be affected by 14C dilution from carbonate rocks. Our ΔR value is used to calibrate new marine radiocarbon ages which help further develop the deglaciation history of Southampton Island, especially along the north coast where deglaciation of Foxe Channel appears to have been completed by 8100–7800calyr BP (2σ). This provides key chronological constraints on the development of a long marine ice margin in southern Foxe Basin prior to the final breakup of the Foxe ice dome.

Large, local ice centers, collectively termed the Appalachian Glacier Complex, developed in Maritime Canada during the Wisconsinan glaciation, effectively barring Laurentide ice from the region. These ice centers shifted in time and space producing palimpsest glacial landforms including cross-striated bedrock outcrops and lobate drumlins. The glacial flow history of the region was deciphered from erosional glacier features (erosional stratigraphy) and till sheet provenace (depositional stratigraphy). Based on provenance, offshore glacial sediments were correlated to their terrestrial counterparts, establishing crucial links between ice flow patterns on land and ice margins offshore. From these empirical data five major flow patterns or ‘phases’ were defined along with their offshore margins. During the Caledonia Phase in the Early to Middle Wisconsinan, eastward to southeastwardflowing ice from Appalachian upland sources crossed Nova Scotia and extended out to the continental shelf edge where a calving margin was established. Submarine mass-wasting at this margin produced wedge-shaped bodies of diamicton that interfingered with glaciomarine sediments. The Caledonia Phase glacier retreated during the Middle Wisconsinan to the inner Scotian Shelf. During the Escuminac Phase in the Late Wisconsinan (22–19 14C ka) an ice centre formed over the Magdalen Shelf (Escuminac Ice Centre) and transported large quantities of local red bed material southward to the outer shelf/slope margin. Just after 18 14C ka, (∼21 CAL) the Escuminac Phase glacier configuration was re-organized by northward-flowing ice streams into marine channels bordering the Magdalen Shelf and an ice stream draining the Bay of Fundy into a divide over Nova Scotia (Scotian Ice, Divide-Scotian Phase). The Scotian Phase glacier margin offshore is marked by the Scotian Shelf End Moraine Complex dated between 17 and 14 14C ka (∼20 and 16.7 CAL). Between 13.5 (∼16 CAL) and 12.0 14C ka (∼14 CAL) the Scotian Ice Divide segmented into local terrestrial ice centers (Chignecto Phase) as a calving bay developed in the Bay of Fundy. These local centers advanced and retreated, responding to mass balance changes during a time of rapidly changing Northern Hemisphere climates. Responding to increasing climatic warming after 12 14C ka (∼14 CAL), the Chignecto Phase glaciers dissipated and their margins retreated landward. Isolated marine and terrestrial remmants of the Chignecto Phase glaciers re-advanced significantly during the Collins Pond Phase (Younger Dryas) ca. 10.8 14C ka (∼13 CAL).

Marine reservoir age is reported for Foxe Basin, Canada, during deglaciation of the Laurentide Ice Sheet. Radiocarbon (C-14) measurements were made on pairs of contemporaneous molluscs and Salix (willow) macrofossils, yielding a mean marine reservoir age of 985 +/- 10 C-14 a. Mean regional Delta R is calculated to be 615 +/- 20 C-14 a relative to a mean global reservoir age of ca. 400 C-14 a. Previous studies in the Canadian Arctic that have been conducted on modern pre-bomb molluscs give Delta R values of ca. 100-300 C-14 a. The difference between modern and deglacial reservoir ages indicates that Delta R in Foxe Basin has changed over time, which has implications for the reconstruction of past events based on marine C-14 dates. We recalculate the timing of deglaciation of Hudson Strait with this new reservoir age and relate this to the 8.2 cal. ka event. We recommend that local Delta R be determined for the time period being examined whenever possible, and that in cases where this is impossible a Delta R value of 615 C-14 a be applied when calibrating marine samples from Foxe Basin and vicinity during deglaciation. Copyright (C) 2010 John Wiley & Sons, Ltd.

Nova Scotia has a long history of gold mining dating back to the 1800s. Most gold mines operated prior to modern environmental standards, which often resulted in unsecured abandoned mine openings and tailings that were not reclaimed following recovery of gold from ore. These vestiges may be a hazard to the local environment or to visitors to the sites. A critical component for evaluation of hazardous conditions is having an accurate location of these features.