Holocene glacier activity reconstructed from proglacial lake Gjøavatnet on Amsterdamøya, NW Svalbard

Abstract

Well-dated and highly resolved paleoclimate records from high latitudes allow for a better understanding of past climate change. Lake sediments are excellent archives of environmental change, and can record processes occurring within the catchment, such as the growth or demise of an upstream glacier. Here we present a Holocene-length, multi-proxy lake sediment record from proglacial lake Gjøavatnet on the island of Amsterdamøya, northwest Svalbard. Today, Gjøavatnet receives meltwater from the Annabreen glacier and contains a record of changes in glacier activity linked to regional climate conditions. We measured changes in organic matter content, dry bulk density, bulk carbon isotopes, elemental concentrations via Itrax core-scanning, and diatom community composition to reconstruct variability in glacier extent back through time. Our reconstruction indicates that glacially derived sedimentation in the lake decreased markedly at ~11.1 cal kyr BP, although a glacier likely persisted in the catchment until ~8.4 cal kyr BP. During the mid-Holocene (~8.4-1.0 cal kyr BP) there was significantly limited glacial influence in the catchment and enhanced deposition of organic-rich sediment in the lake. The deposition of organic rich sediments during this time was interrupted by at least three multi-centennial intervals of reduced organic matter accumulation (~5.9-5.0, 2.7-2.0, and 1.7-1.5 cal kyr BP). Considering the chronological information and a sedimentological comparison with intervals of enhanced glacier input, we interpret these intervals not as glacial advances, but rather as cold/dry episodes that inhibited organic matter production in the lake and surrounding catchment. At ~1.0 cal kyr BP, input of glacially derived sediment to Gjøavatnet abruptly increased, representing the rapid expansion of the Annabreen glacier.

Early Holocene Temperature Oscillations Exceed Amplitude of Observed and Projected Warming in Svalbard Lakes

“…Reconstructed temperatures are colder and vary less in Gjøavatnet, especially during the warming peak centered on 10 ka BP (Figure g). The most apparent difference between these sites that could explain this offset is the sizable glacier (Annabreen) that drained into Gjøavatnet between ~11 and 10 ka BP (De Wet et al, ), while the Hakluyt catchment remained unglaciated (Gjerde et al, ). Input of cold meltwater from this body of ice might explain the suppressed temperature response in lake Gjøa.…”

“…Input of cold meltwater from this body of ice might explain the suppressed temperature response in lake Gjøa. We evaluate this possibility using the XRF Ti‐based reconstruction of Annabreen by De Wet et al () from the same core. Meltwater fluxes are determined by ice volume as well as ablation rates.…”

“…To investigate whether Annabreen meltwater fluxes can account for the offset between the Gjøa and Hakluyt reconstructions, we use the rate of glacier change (RoC, ΔTi counts per century; Text S4) as a surrogate for meltwater input. We account for the fact that larger glaciers produce greater volumes of meltwater by normalizing RoCs to (relative) size as reflected by XRF Ti counts after De Wet et al (). As seen in Figures and , the offset between the Gjøa and Hakluyt temperature reconstructions is largest when Annabreen is in retreat.…”

Reconstructing Holocene Glacier and Climate Fluctuations From Lake Sediments in Vårfluesjøen, Northern Spitsbergen

“…that many glaciers were small or completely melted during this time interval (Svendsen and Mangerud, 1997;Røthe et al, 2015;van der Bilt et al, 2015;de Wet et al, 2018). Mangerud and Svendsen (2018) postulated that August temperatures in Svalbard were 6 • C warmer from 10000 to 9000 cal.…”

“…This approach has been shown to be highly successful in reconstructing past glacier variability in Scandinavia and the Arctic (Karlèn, 1976;Karlén, 1981;Nesje et al, 1991;Matthews et al, 2000;Dahl et al, 2003;Bakke et al, 2005Bakke et al, , 2009Bakke et al, , 2010Røthe et al, 2015;van der Bilt et al, 2015;de Wet et al, 2018), as well as in reconstructing flood frequency (Bøe et al, 2006;Støren et al, 2008Støren et al, , 2010 and avalanche history (Sletten et al, 2003;Nesje et al, 2007;Vasskog et al, 2011). A possible limitation of this method is that it is challenging to distinguish between sediments deposited by different processes occurring in the catchment, due to common sediment sources and transport distance and paraglacial modification (Ballantyne, 2002;Rubensdotter and Rosqvist, 2009;van der Bilt et al, 2016).…”