Here we use polarimetric measurements from an Autonomous phase-sensitive Radio-Echo Sounder (ApRES) to investigate ice fabric within Whillans Ice Stream, West Antarctica. The survey traverse is bounded at one end by the suture zone with the Mercer Ice Stream and at the other end by a basal ‘sticky spot’. Our data analysis employs a phase-based polarimetric coherence method to estimate horizontal ice fabric properties: the fabric orientation and the magnitude of the horizontal fabric asymmetry. We infer an azimuthal rotation in the prevailing horizontal c-axis between the near-surface (z ≈ 10–50 m) and deeper ice (z ≈ 170–360 m), with the near-surface orientated closer to perpendicular to flow and deeper ice closer to parallel. In the near-surface, the fabric asymmetry increases toward the center of Whillans Ice Stream which is consistent with the surface compression direction. By contrast, the fabric orientation in deeper ice is not aligned with the surface compression direction but is consistent with englacial ice reacting to longitudinal compression associated with basal resistance from the nearby sticky spot.

Ice streams are corridors of rapid ice flow draining the ice sheets. They can exhibit astonishing spatial variability on annual to centennial time scales. We propose that changes in the subglacial drainage of meltwater could induce these sudden rearrangements of ice streams. We develop a two‐dimensional, thermomechanical model representing an ice stream cross section and couple it to a plastically deforming bed with spatially variable meltwater influx. We find that where ice flows over deformable sediments and lacks significant topographic control, the efficiency of subglacial water drainage exerts direct control on the velocity, location, and width of ice streams. This implies that meltwater percolation at the meter scale could have a significant effect on the short‐term variability in ice loss from a continental‐scale ice sheet. We verify our model against previous analytical results and validate it against surface observations from the Siple Coast of West Antarctica.

The Subglacial Antarctic Lakes Scientific Access (SALSA) Project accessed Mercer Subglacial Lake using environmentally clean hot-water drilling to examine interactions among ice, water, sediment, rock, microbes and carbon reservoirs within the lake water column and underlying sediments. A ~0.4 m diameter borehole was melted through 1087 m of ice and maintained over ~10 days, allowing observation of ice properties and collection of water and sediment with various tools. Over this period, SALSA collected: 60 L of lake water and 10 L of deep borehole water; microbes >0.2 μm in diameter from in situ filtration of ~100 L of lake water; 10 multicores 0.32–0.49 m long; 1.0 and 1.76 m long gravity cores; three conductivity–temperature–depth profiles of borehole and lake water; five discrete depth current meter measurements in the lake and images of ice, the lake water–ice interface and lake sediments. Temperature and conductivity data showed the hydrodynamic character of water mixing between the borehole and lake after entry. Models simulating melting of the ~6 m thick basal accreted ice layer imply that debris fall-out through the ~15 m water column to the lake sediments from borehole melting had little effect on the stratigraphy of surficial sediment cores.

Characterizing the processes leading to deforestation is critical to the development and implementation of targeted forest conservation and management policies. In this work, we develop a deep learning model called ForestNet to classify the drivers of primary forest loss in Indonesia, a country with one of the highest deforestation rates in the world. Using satellite imagery, ForestNet identifies the direct drivers of deforestation in forest loss patches of any size. We curate a dataset of Landsat 8 satellite images of known forest loss events paired with driver annotations from expert interpreters. We use the dataset to train and validate the models and demonstrate that ForestNet substantially outperforms other standard driver classification approaches. In order to support future research on automated approaches to deforestation driver classification, the dataset curated in this study is publicly available at https://stanfordmlgroup.github.io/projects/forestnet.

An overset grid methodology is developed for the fully coupled analysis of fluid-structure interaction (FSI) problems. The overset grid approach alleviates some of the computational geometry difficulties traditionally associated with Arbitrary-Lagrangian-Eulerian (ALE) based, moving mesh methods for FSI. Our partitioned solution algorithm uses separate solvers for the fluid (finite volume method) and the structure (finite element method), with mesh motion computed only on a subset of component grids of our overset grid assembly. Our results indicate a significant reduction in computational cost for the mesh motion, and element quality is improved. Numerical studies of the benchmark test demonstrate the benefits of our overset mesh method over traditional approaches.