1,721,276 research outputs found

    A gradient-boosted tree framework to model the ice thickness of the world's glaciers (IceBoost v1.1)

    Full text link
    Knowledge of glacier ice volumes is crucial for constraining future sea level potential, evaluating freshwater resources, and assessing impacts on societies, from regional to global. Motivated by the disparity in existing ice volume estimates, we present IceBoost, a global machine learning framework trained to predict ice thickness at arbitrary coordinates, thereby enabling the generation of spatially distributed thickness maps for individual glaciers. IceBoost is an ensemble of two gradient-boosted trees trained with 3.7 million globally available ice thickness measurements and an array of 39 numerical features. The model error is similar to those of existing models outside polar regions and is up to 30 %-40 % lower at high latitudes. Providing supervision by exposing the model to available glacier thickness measurements reduces the error by a factor of up to 2 to 3. A feature-ranking analysis reveals that geodetic data are the most informative variables, while ice velocity can improve the model performance by 6 % at high latitudes. A major feature of IceBoost is a capability to generalize outside the training domain, i.e. producing meaningful ice thickness maps in all regions of the world, including on the ice sheet peripheries

    A 75,000-y-old Scandinavian Arctic cave deposit reveals past faunal diversity and paleoenvironment

    Full text link
    During the last glacial period (~118 to 11.7 ka), the Arctic has been characterized by a major redistribution of flora and fauna as a consequence of extreme climatic fluctuations, with associated glacial advances and retreats, sea-level changes, and shifting sea ice extent. In the high-latitude regions of Northern Europe that are currently subject to rapid climate warming, we lack a comprehensive understanding of faunal biodiversity in the last glacial period due to the extreme rarity of preserved organic remains. Here, we present a stratified sediment deposit with a diverse faunal composition preserved in a bone-bearing layer in Arne Qvamgrotta, part of the Storsteinhola cave system (68.10° N 16.38° E) in Northern Norway. Chronological analyses of sediments and bones including radiocarbon, optically stimulated luminescence, uranium–thorium, and phylogenetic dating place the faunal assemblage in Marine Isotope Stage 5a (MIS 5a, Odderade interstadial, ~85 to 71 ka). Combining comparative osteology and bulk-bone metabarcoding, we identify 46 taxa, including mammals, birds, and fish, with several species not previously found in Fennoscandia. The fauna implies a nonanalogous cold-adapted coastal community, with close proximity to sea ice and nearby freshwater bodies. Mitogenome analyses of a subset of taxa identify extinct lineages which attest to a lack of habitat tracking and the absence of a local refugium during the subsequent fully glaciated periods. This faunal record demonstrates long-term faunal dynamics and coastal environmental conditions during MIS 5a in the European Arctic
    corecore