ABSTRACT Major shifts in hydroclimate have been documented during the last deglacial period and the Holocene in south‐central Alaska. Rare freshwater calcium carbonate (marl) deposits in lakes on the Kenai Peninsula can be used to reconstruct past changes in hydroclimate, including the influence of groundwater inflow to lakes. Here, the postglacial sediment sequence from groundwater‐fed Kelly Lake (60.514°N, 150.374°W) was analyzed for multiple proxies including isotopes of carbon and oxygen in marl calcite (δ 13 C marl and δ 18 O marl ), and isotopes of carbon (ẟ 13 C OM ) and abundances of C and N in organic matter. Bulk sediment analyses include organic matter and calcium carbonate (CaCO 3 ) contents, visual stratigraphy and sediment flux. These analyses extend those of a previous paleoenvironmental reconstruction from Kelly Lake, which focused on sedimentary diatom oxygen isotopes and mass balance modeling over the past 10 000 years. Here, we show that Kelly Lake was deglaciated prior to 14.6 ka, and that by 14.0 ka marl dominated the sediments, with CaCO 3 precipitation probably driven by groundwater input and mediated by shallow‐water charophytes. Marl accumulation decreased as organic and clastic inputs increased between ~12.2 and 11.5 ka. This shift, together with an increase in both δ 13 C marl and δ 18 O marl values and a decrease in CaCO 3 content, indicates an increase in the influence of meteoric water on the hydrologic budget under wet conditions, possibly driven by a strengthened Aleutian Low atmospheric pressure cell. A shift to lower δ 13 C marl and δ 18 O marl values at ~11.5 ka is interpreted as an increase in the proportion of groundwater relative to meteoric water in the lake. Beginning around 9 ka, the proportion of meteoric water input continued to increase, the surrounding coniferous forest was established, and by 8 ka, CaCO 3 accumulation ended. Our results elucidate the environmental conditions under which marl was deposited during the Lateglacial and early Holocene in this part of Alaska, and demonstrate how a variety of synoptic‐ and local‐scale climatic variables can converge to influence sedimentation in a groundwater‐fed lake.

Sediment yield can be a sensitive indicator of catchment dynamics and environmental change. For a glacierized catchment in the High Arctic, we compiled and analyzed diverse sediment transfer data, spanning a wide range of temporal scales, to quantify catchment yields and explore landscape response to past and ongoing hydroclimatic variability. The dataset integrates rates of lake sedimentation from correlated varve records and repeated annual and seasonal sediment traps, augmented by multi-year lake and fluvial monitoring. Consistent spatial patterns of deposition enabled reconstruction of catchment yields from varve- and trap-based fluxes. We used hydroclimatic data and multivariate modeling to examine annual controls of sediment delivery over almost a century, and to examine shorter-term controls of sediment transfer during peak glacier melt. Particle-size analyses, especially for annual sediment traps, were used to further infer sediment transfer mechanisms and timing. Through the Medieval Warm Period and Little Ice Age, there were no apparent multi-century trends in lake sedimentation rates, which were over three times greater than those during the mid-Holocene when glaciers were diminished. Twentieth-century sedimentation rates were greater than those of previous millennia, with a mid-century step increase in mean yield from 240 to 425Mg km(-2)yr(-1). Annual yields through the twentieth century showed significant positive relations with spring/summer temperature, rainfall, and peak discharge conditions. This finding is significant for the future of sediment transfer at Linnevatnet, and perhaps more broadly in the Arctic, where continued increases in temperature and rainfall are projected. For 2004-2010, annual yields ranged from 294 to 1330Mg km(-2)yr(-1). Sediment trap volumes and particle-size variations indicate that recent annual yields were largely dominated by spring to early summer transfer of relatively coarse-grained sediment. Fluvial monitoring showed daily to hourly sediment transfer to be related to current and prior discharge, diurnal hysteresis, air temperature, and precipitation. Copyright (c) 2017 John Wiley & Sons, Ltd.

Reconstructing climates of the past relies on a variety of evidence from a large number of sites to capture the varied features of climate and the spatial heterogeneity of climate change. This review summarizes available information from diverse Holocene paleoenvironmental records across eastern Beringia (Alaska, westernmost Canada and adjacent seas), and it quantifies the primary trends of temperature- and moisture-sensitive records based in part on midges, pollen, and biogeochemical indicators (compiled in the recently published Arctic Holocene database, and updated here to v2.1). The composite time series from these proxy records are compared with new summaries of mountain-glacier and lake-level fluctuations, terrestrial water-isotope records, sea-ice and sea-surface-temperature analyses, and peatland and thaw-lake initiation frequencies to clarify multi-centennial- to millennial-scale trends in Holocene climate change. To focus the synthesis, the paleo data are used to frame specific questions that can be addressed with simulations by Earth system models to investigate the causes and dynamics of past and future climate change. This systematic review shows that, during the early Holocene (11.7–8.2 ka; 1 ka = 1000 cal yr BP), rather than a prominent thermal maximum as suggested previously, temperatures were highly variable, at times both higher and lower than present (approximate mid-20th-century average), with no clear spatial pattern. Composited pollen, midge and other proxy records average out the variability and show the overall lowest summer and mean-annual temperatures across the study region during the earliest Holocene, followed by warming over the early Holocene. The sparse data available on early Holocene glaciation show that glaciers in southern Alaska were as extensive then as they were during the late Holocene. Early Holocene lake levels were low in interior Alaska, but moisture indicators show pronounced differences across the region. The highest frequency of both peatland and thaw-lake initiation ages also occurred during the early Holocene. During the middle Holocene (8.2–4.2 ka), glaciers retreated as the regional average temperature increased to a maximum between 7 and 5 ka, as reflected in most proxy types. Following the middle Holocene thermal maximum, temperatures decreased starting between 4 and 3 ka, signaling the onset of Neoglacial cooling. Glaciers in the Brooks and Alaska Ranges advanced to their maximum Holocene extent as lakes generally rose to modern levels. Temperature differences for averaged 500-year time steps typically ranged by 1–2 °C for individual records in the Arctic Holocene database, with a transition to a cooler late Holocene that was neither abrupt nor spatially coherent. The longest and highest-resolution terrestrial water isotope records previously interpreted to represent changes in the Aleutian low-pressure system around this time are here shown to be largely contradictory. Furthermore, there are too few records with sufficient resolution to identify sub-centennial-scale climate anomalies, such as the 8.2 ka event. The review concludes by suggesting some priorities for future paleoclimate research in the region.

The Arctic region is subject to a great amplitude of climate variability and is currently undergoing large-scale changes due in part to anthropogenic global warming. Accurate projections of future change depend on anticipating the response of the Arctic climate system to forcing, and understanding how the response to human forcing will interact with natural climate variations. The Svalbard Archipelago occupies an important location for studying patterns and causes of Arctic climate variability; however, available paleoclimate records from Svalbard are of restricted use due to limitations of existing climate proxies. Here we present a sub-decadal- to multidecadal-scale record of summer temperature for the past 1800 yr from lake sediments of Kongressvatnet on West Spitsbergen, Svalbard, based on the first instrumental calibration of the alkenone paleothermometer. The age model for the High Arctic lake sediments is based on 210 Pb, plutonium activity, and the first application of tephrochronology to lake sediments in this region. We find that the summer warmth of the past 50 yr recorded in both the instrumental and alkenone records was unmatched in West Spitsbergen in the course of the past 1800 yr, including during the Medieval Climate Anomaly, and that summers during the Little Ice Age (LIA) of the 18 th and 19 th centuries on Svalbard were not particularly cold, even though glaciers occupied their maximum Holocene extent. Our results suggest that increased wintertime precipitation, rather than cold temperatures, was responsible for LIA glaciations on Svalbard and that increased heat transport into the Arctic via the West Spitsbergen Current began ca. A.D. 1600.

Sediment cores from Lone Spruce Pond (60.007°N, 159.143°W), southwestern Alaska, record paleoenvironmental changes during the global Last Glacial Maximum (LGM), and during the last 14,500 calendar years BP (14.5 cal ka). We analyzed the abundance of organic matter, biogenic silica, carbon, and nitrogen, and the isotope ratios of C and N, magnetic susceptibility, and grain-size distribution of bulk sediment, abundance of alder shrub (Alnus) pollen, and midge (Chironomidae and Chaoboridae) assemblages in a 4.7-m-long sediment sequence from the depocenter at 22 m water depth. The basal unit contains macrofossils dating to 25–21 cal ka (the global LGM), and is interpreted as glacial-lacustrine sediment. The open water requires that the outlet of the Ahklun Mountain ice cap had retreated to within 6 km of the range crest. In addition to cladocerans and diatoms, the glacial-lacustrine mud contains chironomids consistent with deep, oligotrophic conditions; several taxa associated with relatively warm conditions are present, suggestive of relative warmth during the global LGM. The glacial-lacustrine unit is separated from the overlying non-glacial lake sediment by a possible disconformity, which might record a readvance of glacier ice. Non-glacial sediment began accumulating around 14.5 cal ka, with high flux of mineral matter and fluctuating physical and biological properties through the global deglacial period, including a reversal in biogenic-silica (BSi) content during the Younger Dryas (YD). During the global deglacial interval, the δ13C values of lake sediment were higher relative to other periods, consistent with low C:N ratios (8), and suggesting a dominant atmospheric CO2 source of C for phytoplankton. Concentrations of aquatic faunal remains (chironomids and Cladocera) were low throughout the deglacial interval, diversity was low and warm-indicator taxa were absent. Higher production and air temperatures are inferred following the YD, when bulk organic-matter (OM) content (LOI 550 °C) increased substantially and permanently, from 10 to 30 %, a trend paralleled by an increase in C and N abundance, an increase in C:N ratio (to about 12), and a decrease in δ13C of sediment. Post-YD warming is marked by a rapid shift in the midge assemblage. Between 8.9 and 8.5 cal ka, Alnus pollen tripled (25–75 %), followed by the near-tripling of BSi (7–19 %) by 8.2 cal ka, and δ15N began a steady rise, reflecting the buildup of N and an increase in denitrification in soils. Several chironomid taxa indicative of relatively warm conditions were present throughout the Holocene. Quantitative chironomid-based temperature inferences are complicated by the expansion of Alnus and resulting changes in lake nutrient status and production; these changes were associated with an abrupt increase in cladoceran abundance and persistent shift in the chironomid assemblage. During the last 2,000 years, chironomid-assemblage changes suggest cooler temperatures, and BSi and OM values were generally lower than their maximum Holocene values, with minima during the seventh and eighth centuries, and again during the eighteenth century.

Abstract Directly measured growth rates of two lichens (Pseudephebe minuscula and Rhizocarpon sections Rhizocarpon and Superficiale) from Svalbard made over a two-decade interval (1984–2007) are presented. Growth rates were determined by measuring the change in area of the lichen thalli from digital images and converting area to diameter. Pseudephebe diameter growth rates ranged from 0.2 to 1.5 mm yr−1 and Rhizocarpon grew 0.05 to 0.30 mm yr−1. Growth rates of both are a function of thalli size—growth rates increase with increasing thallus size up to 70 mm diameter for Pseudephebe and 30 mm diameter for Rhizocarpon. While these directly measured growth rate results are consistent with other recent directly measured lichen growth studies, they are not consistent with indirectly determined age-size curves that show a negative correlation between size and growth rate (i.e., rapid “great growth” followed by slower “linear growth”). We explore several reasons to explain the apparent discrepancy between directly measured and indirectly determined growth rates, including climate change, increased nutrient fluxes, and population sampling differences between the two methods. We argue that indirectly determined growth curves, which integrate the effects of changing growing conditions over time, remain the best basis for lichenometric dating.

The Svalbard Research Experience for Undergraduates (REU) program, initiated in 2003, provides a unique, field-based research experience for US undergraduates in Arctic Quaternary geology and climate change. The Svalbard archipelago, between 74° and 81°N latitude in the North Atlantic, lies at the northern end of the warm Gulf Stream current and therefore is sensitive to subtle climate and oceanographic changes. Svalbard has warmed considerably during the last 90 years, and climate proxies indicate even greater Holocene climate variability. Our program has two main purposes: to train young scientists in cutting-edge, Arctic field research methods, and to reconstruct climate changes of the past 5,000 years from layered sediments in lakes and fjords.

Pollen, macro- and micro-fossils, and sedimentologic indicators in sediment cores from Discovery Pond (DP) in south-central Alaska indicate that the coldest interval of the last deglacation was coincident with the onset of the Younger Dryas (YD), around 12.8calka. The multi-proxy record from DP together with a compilation of recently published YD records from southern Alaska and the adjacent northern Pacific Ocean shows that, during the course of the YD, temperatures increased, then reached a maximum sometime around 11calka. At DP, a pronounced increase in the abundance of Isoëtes and Pediastrum, including species associated with oligotrophic lakes and known to respond to increased precipitation, combined with a reduction in wetland aquatics and an increase in the minerogenic component of the sediment, all indicate a shift from wetland to open-water conditions at around 12.2calka. Similar to other evidence from southern Alaska, our proxy record from DP indicates an increase in temperature and effective moisture during the second half of the YD. An increase in winter precipitation might be associated with a deepening of the Aleutian low-pressure system and a northward shift in winter storm tracks, consistent with recent simulations by climate system models.

A 5.6-m-long lake sediment core from Bear Lake, Alaska, located 22 km southeast of Redoubt Volcano, contains 67 tephra layers deposited over the last 8750 cal yr, comprising 15% of the total thickness of recovered sediment. Using 12 AMS 14C ages, along with the 137Cs and 210Pb activities of recent sediment, we evaluated different models to determine the age–depth relation of the core, and to determine the age of each tephra deposit. The selected age model is based on a mixed-effect regression that was passed through the adjusted tephra-free depth of each dated layer. The estimated age uncertainty of the 67 tephras averages ±105 yr (95% confidence intervals). Tephra-fall frequency at Bear Lake was among the highest during the past 500 yr, with eight tephras deposited compared to an average of 3.7/500 yr over the last 8500 yr. Other periods of increased tephra fall occurred 2500–3500, 4500–5000, and 7000–7500 cal yr. Our record suggests that Bear Lake experienced extended periods (1000–2000 yr) of increased tephra fall separated by shorter periods (500–1000 yr) of apparent quiescence. The Bear Lake sediment core affords the most comprehensive tephrochronology from the base of the Redoubt Volcano to date, with an average tephra-fall frequency of one every 130 yr.

Tephra-fall deposits from Cook Inlet volcanoes were detected in sediment cores from Tustumena and Paradox Lakes, Kenai Peninsula, Alaska, using magnetic susceptibility and petrography. The ages of tephra layers were estimated using 21 14C ages on macrofossils. Tephras layers are typically fine, gray ash, 1–5 mm thick, and composed of varying proportions of glass shards, pumice, and glass-coated phenocrysts. Of the two lakes, Paradox Lake contained a higher frequency of tephra (0.8 tephra/100 yr; 109 over the 13,200-yr record). The unusually large number of tephra in this lake relative to others previously studied in the area is attributed to the lake's physiography, sedimentology, and limnology. The frequency of ash fall was not constant through the Holocene. In Paradox Lake, tephra layers are absent between ca. 800–2200, 3800–4800, and 9000–10,300 cal yr BP, despite continuously layered lacustrine sediment. In contrast, between 5000 and 9000 cal yr BP, an average of 1.7 tephra layers are present per 100 yr. The peak period of tephra fall (7000–9000 cal yr BP; 2.6 tephra/100 yr) in Paradox Lake is consistent with the increase in volcanism between 7000 and 9000 yr ago recorded in the Greenland ice cores.