ABSTRACT We defined the timing of surface abandonment for 10 alluvial and debris-flow fans across contrasting climatic settings in the NW Himalaya of northern India using cosmogenic 10 Be surface exposure dating. Debris-flow fans in the Garhwal, Kullu, and Lahul-Spiti regions of the monsoon-influenced Greater Himalaya were largely abandoned during the Mid- to Late Holocene. Large alluvial fans and smaller debris-flow fans in the semiarid Ladakh region of the Greater and Tethyan Himalaya have surface ages that extend throughout the last glacial. Regional events of landform abandonment and incision were defined for the monsoon-influenced western Himalaya ranges and the semiarid western Himalaya ranges over the past ~120 k.y. In the monsoon-influenced and semiarid western Himalaya ranges, these regional events were limited to the Holocene and from ca. 40 ka, respectively. The timing of fan surface abandonment and regional landform abandonment events coincided with periods of weakening monsoon strength and cooling, and local and regional glacier advances. Regional incision events from the monsoon-influenced and semiarid western Himalaya regions were recognized across various climatic conditions due to the ubiquitous nature of erosion in mountain settings. This study showed that climate-driven processes and glaciation were important drivers in fan sedimentation, catchment sediment flux, and the topographic evolution of the NW Himalaya during the late Quaternary.

ABSTRACT This paper presents a detailed sedimentologic data set of minor moraines (heights ≤2.0 m, widths ≤14 m, lengths ≤108 m) that formed beginning near the end of the Little Ice Age by Schwarzensteinkees, a valley glacier in Austria. Sorted sediment and stratified diamict dominate five exposures, and compact massive diamict exists in one exposure. This sediment is interpreted as proglacial outwash and subglacial till. Most moraine sediment shows deformation structures (e.g., smaller and larger folds), and some units contain evidence of water escape. Other units maintain their original subhorizontality. All moraines contain unequivocal evidence of having formed through deformation by pushing during ice-margin fluctuations. Minor moraines formed more specifically by three identified processes: (1) pushing of outwash sediment; (2) stacking and pushing of outwash sediment; and (3) pushing of outwash sediment and freezing-on of subglacial till. Our data suggest that the sedimentologic composition of the valley fill influences the style of push-moraine formation. In this case, the friable nature of outwash sediments can increase the efficiency of the pushing ice front and the likelihood of sediment collapse down the proximal ice-contact slope after ice retreat. This study contributes to our understanding of sediment transport and deposition in high-mountain environments.

ABSTRACT Uplift of the central Andes during the Miocene was followed by large-scale reorganization of Atlantic-draining rivers in Argentine Patagonia. Here, we document the abandonment of one large river in the late Pliocene and the establishment of the modern drainage in the Early Pleistocene. A chronology for these events is provided by 40 Ar/ 39 Ar ages on basalt flows. Remnants of the Pliocene paleovalley system are well preserved in the Lago Cardiel–Gobernador Gregores area, where they are eroded into flat-lying basalt flows dated from ca. 13.9 Ma to 8.6 Ma. Younger basalts that erupted onto the abandoned floor of the paleovalley are as young as 3.7 Ma. Abandonment of the Pliocene paleovalley and establishment of the modern Río Chico and Río Shehuen catchments happened near the close of the Pliocene when Andean glaciers incised the east-sloping pediment on which the late Miocene drainage was established. Lago Cardiel sits within a large endorheic basin that is inset into the late Pliocene paleovalley. The basin began to develop just before 4 Ma, after the paleovalley was abandoned. It became larger and deeper during the Pleistocene due to mass movements along its margins, deflation of the basin floor during times when Lago Cardiel was dry or nearly dry, and possibly lowering along bounding faults. The Pliocene–Pleistocene landscape and drainage changes that we have documented are not unique to the Lago Cardiel–Gobernador Gregores area; similar changes are apparent elsewhere in Patagonia east of the crest of the Andes.

ABSTRACT The northern Rocky Mountain Trench of eastern British Columbia is a broad valley mantled by glaciolacustrine terraces supporting a complex mix of mesic-temperate (“interior wet belt”) forests that are strongly affected by terrain and substrate. Neither the geomorphic history during early Holocene deglaciation nor the vegetation history of the origin of the Tsuga heterophylla (western hemlock) and Thuja plicata (western redcedar) populations in the interior wet-belt forest is well understood. Sediment cores were obtained from two lakes, 10 km apart and occupying different terraces (83 m elevational difference), and these were compared to existing fire-history and paleoclimate reconstructions. Radiocarbon dates and a mapped terrain classification indicate the upper terrace formed as a lacustrine and glaciofluvial kame terrace hundreds of years prior to the lower terrace, which was formed by glaciolacustrine sediments of a proglacial lake. The minimum limiting ages of these terraces correlate with dated jökulhlaup deposits of the Fraser River. The upper site’s first detectable pollen at >11.0 ka was dominated by light-seeded pioneer taxa (Poaceae [grasses], Artemisia [sagebrush], and Populus [aspen]) followed by a peak in Pinus (pine) and finally dominance by Betula (birch) at 10.2 ka. Pollen data suggest an earlier invasion of T. heterophylla (western hemlock) (by ca. 8 ka) than previously understood. Wetlands on extensive, poorly drained, glaciolacustrine soils promoted the persistence of boreal taxa and open forests (e.g., Picea mariana [black spruce]), while the better-drained upper kame terrace promoted development of closed-canopy shade-tolerant taxa. Invasion and expansion of mesic cedar-hemlock taxa progressed since at least the middle Holocene but was highly constrained by edaphic controls.

ABSTRACT Extensive glaciers covered the High Atlas mountains in Morocco during the late Pleistocene. On the northern escarpments of the Marrakech High Atlas, a series of cirques perched at ~3000–3500 m above sea level (asl) fed their valley glaciers that, in some cases, extended to as low as 2000 m asl. Cosmogenic exposure dating with 10 Be and 36 Cl has shown that at least three phases of glaciation are preserved in glacial deposits over the last glacial cycle at 50, 22, and 12 ka, which appear to correlate with marine isotope stage (MIS) 3, the global Last Glacial Maximum (LGM), and the Younger Dryas chronozone. This geochronological framework is sufficiently robust to allow for time-constrained glacier-climate reconstructions. The glaciers associated with these three phases of advance had equilibrium line altitudes (ELAs) of 2761 m asl (ca. 50 ka), 2919 m asl (ca. 22 ka), and 3213 m asl (ca. 12 ka). Glacier-climate modeling suggests that all of these phases were driven by both colder temperatures and wetter conditions than today. The dominant moisture supply to these glaciers in all phases would have been sourced from Atlantic depressions. The influence of an extended and enhanced West African monsoon on glacier development during African Humid Periods is unlikely to have been a significant influence on glacier dynamics. The climate conditions associated with the three glacier phases indicate sustained moisture supply to the highest mountain areas when records from other areas, such as the Middle Atlas lakes and marine sediment cores offshore, indicate marked aridity.

ABSTRACT The mid-Pleistocene transition is a time interval between ca. 1.2 and 0.7 Ma during which a shift occurred from ~41 k.y. glacial-interglacial cycles to ~100 k.y. cycles. Although the mid-Pleistocene transition has been well documented in global marine records, its effects in continental environments, including North America, are incompletely understood owing to the paucity of terrestrial sediment records spanning the entire Quaternary. A notable exception is the ca. 1.4 Ma and younger Blackwater Draw Formation, an extensive eolian sequence on the Southern High Plains of the United States. Intervals of the Blackwater Draw Formation section that are inferred to span the mid-Pleistocene transition can be divided into pre–, syn–, and post–mid-Pleistocene transition parts. Weathering profiles in the pre–mid-Pleistocene transition section are dominated by weakly developed soils formed in arid environments, as evidenced by well-expressed pedogenic carbonate horizons, lack of clay formation during hydrolysis, and magnetically soft, coarse-grained magnetite/maghemite populations. Conversely, the syn– and post–mid-Pleistocene transition intervals demonstrate an increase in weathering intensity by an abrupt increase in clay content formed in part by hydrolysis of feldspars, soil profiles that demonstrate leaching and illuviation, and a fining-upward grain size of the magnetite/maghemite population. Sedimentologic, geochemical, and rock-magnetic data are consistent with a southern and coarser sediment source derived from the Pecos River drainage prior to the mid-Pleistocene transition, followed by a mixture of northern and southern sources during and after the mid-Pleistocene transition. Overall, our results indicate that pre–mid-Pleistocene transition conditions on the Southern High Plains were arid with wind energy sufficient to mobilize sand sheets out of the Pecos River and deposit them on the plateau. The syn– and post–mid-Pleistocene transition environments reflect somewhat wetter conditions and potentially an influx of silt from the north, in addition to continued sand derived from the Pecos River valley. The wetter conditions and silt influx may have resulted from longer-lived and more robust glacial activity in the Northern Hemisphere that characterized the post–mid-Pleistocene transition Earth system.

ABSTRACT Global warming and ice melting in the “tripolar” regions, namely, Antarctica, the Arctic, and the Tibetan Plateau, might lead to a reorganization of the global climate system. Understanding the dynamic links between the “tripolar” environments and the Earth climate system is crucial to improving our capability to project future climate variability. “Tripolar” changes have influenced the onset and evolution of the Asian paleomonsoon system through various atmospheric and/or oceanic mechanisms. Here, we summarize previously reported Asian paleomonsoon variations that can be linked to “tripolar” environmental changes, from tectonic to millennial time scales, and explore linkages between the “tripolar” regions and global climate changes.

ABSTRACT Glaciers in central Asia that developed under a range of climatic conditions from arid to humid provide an excellent opportunity to test glacial responses to changes in climate. To do this, we mapped and dated glacial deposits at 11 sites spread over five mountain ranges in central Asia: the Altai, Tian Shan, Altyn Tagh, Qilian Shan, and Kunlun. The glacial chronologies for these sites were determined from new 10 Be and 26 Al exposure ages for the mapped moraines, in addition to 10 Be ages available in the literature. Paleo–equilibrium-line altitudes were estimated for past glacier extents from the dated moraines. The equilibrium-line altitudes (ELAs) were also estimated for existing glaciers to characterize the spatial pattern in modern climate across the study region. Differences between the modern and paleo-ELAs (∆ELAs) were used to explore the climatic reasons for variations in the glacier sensitivities and responses to past changes in climate. The results show that the glaciers in more humid regions advanced to their maximum during marine oxygen-isotope stage (MIS) 3–2 with ΔELAs of ~1100–600 m. However, glaciers in the arid interior of central Asia, in the rain shadows of the Karakorum and Pamir ranges and in the Gobi Desert ranges, reached their maximum between MIS 6 and 4, and glacier extents during the subsequent colder/drier MIS 3–2 were significantly smaller or did not extend beyond their cirques. Comparisons of our results and the sensitivity analysis of modern glaciers suggest that depression of air temperature was the primary driver of glacier advances in central Asia but that precipitation played a major role in shaping the spatial and temporal heterogeneity of glacier advances. Precipitation was especially important in hyperarid conditions. Therefore, inferences about paleoclimate parameters from past glacial extents must be made after careful consideration of the climatic setting in which the glaciers are found, as well as their sensitivity to climatic factors.

ABSTRACT The last-glacial megaflood Kankakee Torrent streamlined hills and the remarkably straight backslope of the Kalamazoo moraine (Lake Michigan lobe of the Laurentide ice sheet) in southwestern Michigan. Flooding ensued as proglacial Lake Dowagiac overflowed across remnants of the Lake Michigan lobe at the position of the inner margin of the Kalamazoo moraine as glacial debris and ablating ice were pinned against Portage Prairie. Proglacial Lake Dowagiac developed in the Dowagiac River valley as the lobe retreated to form the Valparaiso moraine. A minimum age of the Kankakee Torrent (18.7 ± 0.6 k.y. B.P) is indicated by the weighted mean value of six optically stimulated luminescence ages determined from quartz sand in glaciofluvial sediment on the Kalamazoo moraine (Lake Michigan and Saginaw lobes). This value is consistent with tighter age control based on radiocarbon ages of tundra plants within silty sediment forming ice-walled lake plains and in a torrent-scoured lake basin (Oswego channel) in Illinois. Crosscutting relationships of well-dated moraines indicate the Kankakee Torrent occurred sometime between 19.7 and 18.9 calibrated (cal.) k.y. B.P. as it skirted the south margin of the Valparaiso Morainic System.

ABSTRACT Mountain glacier moraine sequences and their chronologies allow us to evaluate the timing and climate conditions that underpin changes in the equilibrium line altitudes (ELAs), which can provide valuable information on the paleoclimatology of understudied regions such as tropical East Africa. However, moraine sequences are inherently discontinuous, and the precise climate conditions that they represent can be ambiguous due to the sensitivity of mountain glaciers to temperature, precipitation, and other environmental variables. Here, we used a two-dimensional (2-D) ice-flow and mass-balance model to simulate glacier extents and ELAs in the Rwenzori Mountains in East Africa over the past 31,000 yr (31 k.y.), including the Last Glacial Maximum (LGM), late glacial period, and the Holocene Epoch. We drove the glacier model with two independent, continuous temperature reconstructions to simulate possible glacier length changes through time. Model input paleoclimate values came from branched glycerol dialkyl glycerol tetraether (brGDGT) temperature reconstructions from alpine lakes on Mount Kenya for the last ~31 k.y., and precipitation reconstructions for the LGM came from various East African locations. We then compared the simulated fluctuations with the positions and ages (where known) of the Rwenzori moraines. The simulated glacier extents reached within 1.1 km of the dated LGM moraines in one valley (93% of the full LGM extent) when forced by the brGDGT temperature reconstructions (maximum cooling of 6.1 °C) and a decrease in precipitation (−10% than modern amounts). These simulations suggest that the Rwenzori glaciers required a cooling of at least 6.1 °C to reach the dated LGM moraines. Based on the model output, we predict an age of 12–11 ka for moraines located halfway between the LGM and modern glacier extents. We also predict ice-free conditions in the Rwenzori Mountains for most of the early to middle Holocene, followed by a late Holocene glacier readvance within the last 2000 yr.

ABSTRACT Climate during the Last Glacial Maximum (LGM) varied substantially across North America, strongly influencing changes in plant and animal distributions and causing variations in the timing and relative magnitude of ice expansion and recession. The Olympic Peninsula is a mountainous maritime terrain in northwestern Washington, where the climate today is most strongly influenced by Pacific weather systems. However, what about during the LGM, when ice sheets covered most of northern North America? Fossil beetle assemblages of LGM age contain species that currently inhabit riparian and lacustrine habitats in the boreal zone of Canada and Alaska, and in higher elevations in the Cascades and Rocky Mountains. They include three Olophrum species that today are unknown from the Olympic Peninsula. Olophrum consimile is especially well represented, and its occurrence today above 1000 m elevation in the Cascades of northern Washington State indicates summers during the LGM would have been at least 4 °C cooler than today. The absence of wood-boring beetles, in contrast to assemblages from deposits correlating with marine isotope stage (MIS) 3, supports an open rather than a forested landscape. The insect fossils also include an undescribed species of a blind trechine ground beetle, likely endemic to the Pacific Northwest with biogeographic affinities to Asia. Pollen and plant macrofossil evidence for a Sitka spruce and mountain hemlock parkland with similarities to the vegetation of modern southeast Alaska also supports an interpretation of a climate with summer temperatures ~4 °C cooler than today. Both the vegetation and the insects provide evidence that the climate was wet with persistent snow cover and not as dry as has been reported from the Puget Lowland to the east. Glacial geology provides evidence that during the colder climate of the LGM, mountain glaciers advanced down the western valleys of the Olympic Peninsula to the lowlands but not as far as they had extended during MIS 3. The amount of climatic cooling on the Olympic Peninsula during the LGM was less than at similar latitudes in midcontinental or eastern North America, indicating a strong modulation of climate by the Pacific Ocean.

ABSTRACT The Tuya-Kawdy region of northern British Columbia is well established as a place where glaciation and volcanism overlapped in space. However, no modern work has integrated observations from the region’s volcanic and glacial deposits with geochronologic constraints to summarize how they might overlap in time. Here, we provide a general overview of such characteristics and 23 new 40 Ar/ 39 Ar eruption ages of glaciovolcanic deposits ranging from 4.3 Ma to 63 ka to constrain the timing, location, and minimum thicknesses and distributions of coincident ice. Subaerial lava fields interspersed with glaciovolcanism record periods of ice-sheet absence in presumably warmer climate conditions. These generally coincide with interglacial marine isotope stages. Many of the volcanoes have a secondary record of posteruption glacial modification, cirques, erratics, and mega-lineations, which document later climate changes up to the present. We used edifice-based terrain analysis to reconstruct changes to local minimum Cordilleran ice-sheet thicknesses, extents, and flow directions at specific locations and times during the late Pliocene and the Pleistocene.

ABSTRACT High-resolution light detection and ranging (lidar) data and new stratigraphic, lake sediment, and radiocarbon constraints help to resolve a long-standing dispute regarding the timing and nature of the Everson interstade and the Sumas stade, the last major events of the Cordilleran ice sheet in the Fraser Lowland. The new data indicate that: (1) an early, maximum Sumas advance occurred roughly 14,500 cal yr B.P. (calibrated 14 C years before 1950), extending into the Salish Sea near Bellingham, Washington; (2) ice retreated north of the International Boundary long enough for forests to establish in deglaciated lowland sites; (3) a rapid, short-lived rise in local relative sea level (RSL) of ~20–30 m, possibly related to meltwater pulse 1A or the collapse of a glacio-isostatic forebulge, inundated the U.S. portion of the lowlands up to ~130 m above modern sea level; and (4) directly following this transgression at ca. 14,000 cal yr B.P., ice readvanced across the border to nearly the same limit as reached during the early Sumas period. Distinct crosscutting marine strandlines (erosional and depositional remains of emerged marine shorelines), subaerial moraines, and till plains imaged in lidar data indicate that following the maximum extent of the second Sumas advance, local RSL progressively lowered as the glacier fluctuated and gradually thinned. By ca. 13,000 cal yr B.P., ice had retreated north of the border, and local RSL had fallen to within ~4 m of modern. A layer of possible loess in sediments in Squalicum Lake suggests a possible third and final Sumas readvance between 13,000 and 11,150 cal yr B.P., at which time a moraine was constructed ~8 km south of the border near the town of Sumas, Washington. Together, our results suggest that the concept of a distinct Everson interstade and Sumas stade should be abandoned in favor of a more nuanced “Sumas episode” that encompasses the sequence of events recorded in the Fraser Lowland.

ABSTRACT Recently obtained radiocarbon ages from the southern Puget Lowland and reevaluation of limiting ages from the Olympic Peninsula in the light of new light detection and ranging (LiDAR) data suggest that the Juan de Fuca and Puget lobes of the Cordilleran ice sheet reached their maximum extents after 16,000 calibrated yr B.P. Source areas for both lobes fed through a common conduit, likely requiring that downstream responses to changes in either source area were similar. Dates for ice-sheet retreat are sparse and contradictory, but they suggest that retreat was rapid. Depositional and geomorphic evidence shows that retreat of the Juan de Fuca lobe predated retreat of the Puget lobe. No recessional end moraines have been identified in the Puget Lowland, in contrast to numerous recessional end moraines constructed by the Okanogan lobe east of the Cascade Range, and in contrast to later ice-sheet retreat in western Whatcom County north of the Puget Lowland. These observations lead to the hypothesis that collapse of the Juan de Fuca lobe, hastened by the instability of a marine-based ice sheet, steepened the ice-sheet surface over the eastern Strait of Juan de Fuca and diverted ice flow upstream of the Puget lobe to the west. Starved of ice, the Puget lobe retreated quickly.

ABSTRACT An extensive kame-terrace sequence in the middle Rangitata Valley reveals ice-volume fluctuations spanning the last (Otiran) glaciation. Stratigraphic and sedimentologic characteristics document lateral ice-marginal processes and provide context for luminescence dating. The sequence provides novel and complementary data on glacier ice thickness, which fluctuated substantially throughout the Otiran glaciation. Thick ice constructed one of the highest kame terraces (540 m above the valley floor) ca. 68 ka and thinned nearly 500 m to the valley floor by ca. 53 ka. Following an episode of ice thickening to an unknown elevation, ice again thinned to the valley floor by ca. 44 ka. Ice thickened to its greatest late marine oxygen isotope stage (MIS) 3 extent of 480 m by ca. 37 ka, and thinned to 230 m by ca. 31 ka. The final ice expansion, to 260 m, occurred by ca. 25.5 ka, and the ice fluctuated and thinned to 240 m at ca. 22–20 ka and to 170 m at ca. 21–17 ka. Published cosmogenic radionuclide (CRN) ages indicate surface stabilization near the valley floor (55 m) by ca. 18 ka. This ice-thickness chronology provides an independently derived ice-volume record that is consistent with local and regional glacial chronologies. The site, lying between the Mackenzie Basin and the northern Canterbury Plains drainages, displays a chronology with advances correlative in part with each of those regions. Maximum ice extent occurred 70–65 ka in the Rangitata Valley and the Mackenzie Basin, while the subsequent ice expansion ca. 37 ka is similar in timing to chronologies in both the Rakaia Valley to the north and the Mackenzie Basin to the south.

ABSTRACT The northward retreat history of the Laurentide ice sheet through the lowlands of the northeastern United States during the last deglaciation is well constrained, but its vertical thinning history is less well known because of the lack of direct constraints on ice thickness through time and space. In addition, the highest elevations in New England are characterized by gently sloping upland surfaces and weathered block fields, features with an uncertain history. To better constrain ice-sheet history in this area and its relationship to alpine geomorphology, we present 20 new 10 Be and seven in situ 14 C cosmogenic nuclide measurements along an elevation transect at Mount Washington, New Hampshire, the highest mountain in the northeastern United States (1917 m above sea level [a.s.l.]). Our results suggest substantially different exposure and erosion histories on the upper and lower parts of the mountain. Above 1600 m a.s.l., 10 Be and in situ 14 C measurements are consistent with upper reaches of the mountain deglaciating by 18 ka. However, some 10 Be ages are up to several times greater than the age of the last deglaciation, consistent with weakly erosive, cold-based ice that did not deeply erode preglacial surfaces. Below 1600 m a.s.l., 10 Be ages are indistinguishable over a nearly 900 m range in elevation and imply rapid ice-surface lowering ca. 14.1 ± 1.1 ka (1 standard deviation; n = 9). This shift from slow thinning early in the deglaciation on the upper part of the mountain to abrupt thinning across the lower elevations coincided with accelerated ice-margin retreat through the region recorded by Connecticut River valley varve records during the Bølling interstadial. The Mount Washington cosmogenic nuclide vertical transect and the Connecticut River valley varve record, along with other New England cosmogenic nuclide records, suggest rapid ice-volume loss in the interior northeastern United States in response to Bølling warming.

ABSTRACT In late Wisconsin time, the Purcell Trench lobe of the Cordilleran ice sheet dammed the Clark Fork of the Columbia River in western Montana, creating glacial Lake Missoula. During part of this epoch, the Okanogan lobe also dammed the Columbia River downstream, creating glacial Lake Columbia in northeast Washington. Repeated failure of the Purcell Trench ice dam released glacial Lake Missoula, causing dozens of catastrophic floods in eastern Washington that can be distinguished by the geologic record they left behind. These floods removed tens of meters of pale loess from dark basalt substrate, forming scars along flowpaths visible from space. Different positions of the Okanogan lobe are required for modeled Missoula floods to inundate the diverse channels that show field evidence for flooding, as shown by accurate dam-break flood modeling using a roughly 185 m digital terrain model of existing topography (with control points dynamically varied using automatic mesh refinement). The maximum extent of the Okanogan lobe, which blocked inundation of the upper Grand Coulee and the Columbia River valley, is required to flood all channels in the Telford scablands and to produce highest flood stages in Pasco Basin. Alternatively, the Columbia River valley must have been open and the upper Grand Coulee blocked to nearly match evidence for high water on Pangborn bar near Wenatchee, Washington, and to flood Quincy Basin from the west. Finally, if the Columbia River valley and upper Grand Coulee were both open, Quincy Basin would have flooded from the northeast. In all these scenarios, the discrepancy between modeled flood stages and field evidence for maximum flood stages increases in all channels downstream, from Spokane to Umatilla Basin. The pattern of discrepancies indicates that bulking of floods by loess increased flow volume across the scablands, but this alone does not explain low ­modeled flow stages along the Columbia River valley near Wenatchee. This latter discrepancy between modeled flood stages and field data requires either additional bulking of flow by sediment along the Columbia reach downstream of glacial Lake Columbia, or coincident dam failures of glacial Lake Columbia and glacial Lake Missoula.

ABSTRACT The Matanuska lowland north of Anchorage, Alaska, was episodically glaciated during the Pleistocene by the merged westward flow of the Matanuska and Knik glaciers. During the late Wisconsin glaciation, glacial Lake Atna filled the Copper River Basin, impounded by an ice dam blocking the Matanuska drainage divide at Tahneta Pass and the adjacent Squaw Creek headwaters and ice dams at other basin outlets, including the Susitna and Copper rivers. On the Matanuska lowland floor upvalley from the coalesced glacier’s late-Wisconsin terminus, a series of regularly spaced, symmetrical ridges with 0.9-km wavelengths and heights to 36 m are oriented normal to oblique to the valley and covered by smaller subparallel ridges with wavelengths typically ~80 m and amplitudes to 3 m. These and nearby drumlins, eskers, and moraines were previously interpreted to be glacial in origin. Borrow-pit exposures in the large ridges, however, show sorting and stratification, locally with foreset bedding. A decade ago we reinterpreted such observations as evidence of outburst flooding during glacial retreat, driven by water flushing from Lake Atna through breaches in the Tahneta Pass and Squaw Creek ice dam. In this view, the ridges once labeled Rogen and De Geer moraines were reinterpreted as two scales of fluvial dunes. New observations in the field and from meter-scale light detection and ranging (LiDAR) and interferometric synthetic aperture radar (IfSAR) digital elevation models, together with grain-size analyses and ground-penetrating radar profiles, provide further evidence that portions of the glacial landscape of the Matanuska lowlands were modified by megaflooding after the Last Glacial Maximum, and support the conclusion that the Knik Glacier was the last active glacier in the lowland.

ABSTRACT The Pleistocene Okanogan lobe of Cordilleran ice in north-central Washington State dammed Columbia River to pond glacial Lake Columbia and divert the river south across one or another low spot along a 230-km-long drainage divide. When enormous Missoula floods from the east briefly engulfed the lake, water poured across a few such divide saddles. The grandest such spillway into the Channeled Scabland became upper Grand Coulee. By cutting headward to Columbia valley, upper Grand Coulee’s flood cataract opened a valve that then kept glacial Lake Columbia low and limited later floods into nearby Moses Coulee. Indeed few of the scores of last-glacial Missoula floods managed to reach it. Headward cutting of an inferred smaller cataract (Foster Coulee) had earlier lowered glacial Lake Columbia’s outlet. Such Scabland piracies explain a variety of field evidence assembled here: apparently successive outlets of glacial Lake Columbia, and certain megaflood features downcurrent to Wenatchee and Quincy basin. Ice-rafted erratics and the Pangborn bar of foreset gravel near Wenatchee record late Wisconsin flood(s) down Columbia valley as deep as 320 m. Fancher bar, 45 m higher than Pangborn bar, also has tall foreset beds—but its gravel is partly rotted and capped by thick calcrete, thus pre-Wisconsin age, perhaps greatly so. In western Quincy basin foreset beds of basaltic gravel dip east from Columbia valley into the basin—gravel also partly rotted and capped by thick calcrete, also pre-Wisconsin. Yet evidence of late Wisconsin eastward flow to Quincy basin is sparse. This sequence suggests that upper Grand Coulee had largely opened before down-Columbia megaflood(s) early in late Wisconsin time. A drift-obscured area of the Waterville Plateau near Badger Wells is the inconspicuous divide saddle between Columbia tributary Foster Creek drainage and Moses Coulee drainage. Before flood cataracts had opened upper Grand Coulee or Foster Coulee, and while Okanogan ice blocked the Columbia but not Foster Creek, glacial Lake Columbia (diverted Columbia River) drained over this saddle at about 654 m and down Moses Coulee. When glacial Lake Columbia stood at this high level so far west, Missoula floods swelling the lake could easily and deeply flood Moses Coulee. Once eastern Foster Coulee cataract had been cut through, and especially once upper Grand Coulee’s great cataract receded to Columbia valley, glacial Lake Columbia stood lower, and Moses Coulee became harder to flood. During the late Wisconsin (marine isotope stage [MIS] 2), only when Okanogan-lobe ice blocked the Columbia near Brewster to form a high lake could Missoula floodwater from glacial Lake Missoula rise enough to overflow into Moses Coulee—and then only in a few very largest Missoula floods. Moses Coulee’s main excavation must lie with pre-Wisconsin outburst floods (MIS 6 or much earlier)—before upper Grand Coulee’s cataract had receded to Columbia valley.