Search

The Southern Ocean overturning circulation is driven by winds, heat fluxes, and freshwater sources. Among these sources of freshwater, Antarctic sea ice formation and melting play the dominant role. Even though ice-shelf melt is relatively small in magnitude, it is located close to regions of convection, where it may influence dense water formation. Here, we explore the impacts of ice-shelf melting on Southern Ocean water-mass transformation (WMT) using simulations from the Energy Exascale Earth System Model (E3SM) both with and without the explicit representation of melt fluxes from beneath Antarctic ice shelves. We find that ice-shelf melting enhances transformation of Upper Circumpolar Deep Water, converting it to lower density values. While the overall differences in Southern Ocean WMT between the two simulations are moderate, freshwater fluxes produced by ice-shelf melting have a further, indirect impact on the Southern Ocean overturning circulation through their interaction with sea ice formation and melting, which also cause considerable upwelling. We further find that surface freshening and cooling by ice-shelf melting cause increased Antarctic sea ice production and stronger density stratification near the Antarctic coast. In addition, ice-shelf melting causes decreasing air temperature, which may be directly related to sea ice expansion. The increased stratification reduces vertical heat transport from the deeper ocean. Although the addition of ice-shelf melting processes leads to no significant changes in Southern Ocean WMT, the simulations and analysis conducted here point to a relationship between increased Antarctic ice-shelf melting and the increased role of sea ice in Southern Ocean overturning.

During 2012 to 2014 five expeditions collected surface water samples for plastic pollution analysis representing the first data within Lake Ontario and the first multi-year dataset for Lake Erie. Lake Ontario had the highest abundances of any Great Lake to date with an average of over 230,000 particles/km2. Though having a considerable smaller average of ~45,000 particles/km2, Lake Erie remains second only to Lake Ontario based on studies to date and averaged across all samples and years. The high levels of pelagic plastic pollution is likely owing to their position as the last two lakes in the Laurentian Great Lakes ecosystem, as well as the prominence of population centers along their shorelines. As with previous studies, most particles were found within the smallest size classification (0.355–0.999 mm; 73%), with fragments (63%) and pellets (26%) forming the dominant morphologies. The minor contribution of fibers/lines (4%) is consistent with previous Great Lakes studies, though not with studies within other environmental compartments (e.g., sediment, fish, atmospheric). This could be due to the negative buoyancy of polymeric fibrous materials, a hypothesis consistent with the dominance of the less dense polymers polyethylene (46%) and polypropylene (43%) (based on FTIR analysis). For the first time, the multiyear Lake Erie samples were compared to modeled plastic distributions and found to fit reasonably well. Using the sample data to calibrate the model we estimate that there are 475 million plastic particles, with a total mass of 6.45 metric tons, floating on the surface of Lake Erie alone.

Individual tennis points evolve over time and space, as each of the two opposing players are constantly reacting and positioning themselves in response to strikes of the ball. However, these reactions are diminished into simple tally statistics such as the amount of winners or unforced errors a player has. In this paper, a new way is proposed to evaluate how an individual tennis point is evolving, by measuring how many points a player can expect from each shot, given who struck the shot and where both players are located. This measurement, named “Expected Shot Value” (ESV), derives from stochastically modeling each shot of individual tennis points. The modeling will take place on multiple resolutions, differentiating between the continuous player movement and discrete events such as strikes occurring and duration of shots ending. Multi-resolution stochastic modeling allows for the incorporation of information-rich spatiotemporal player-tracking data, while allowing for computational tractability on large amounts of data. In addition to estimating ESV, this methodology will be able to identify the strengths and weaknesses of specific players, which will have the ability to guide a player’s in-match strategy.

Biological interpretation of a large amount of gene or protein data is complex. Ontology analysis tools are imperative in finding functional similarities through overrepresentation or enrichment of terms associated with the input gene or protein lists. However, most tools are limited by their ability to do ontology-specific and species-limited analyses. Furthermore, some enrichment tools are not updated frequently with recent information from databases, thus giving users inaccurate, outdated or uninformative data. Here, we present MOET or the Multi-Ontology Enrichment Tool (v.1 released in April 2019 and v.2 released in May 2021), an ontology analysis tool leveraging data that the Rat Genome Database (RGD) integrated from in-house expert curation and external databases including the National Center for Biotechnology Information (NCBI), Mouse Genome Informatics (MGI), The Kyoto Encyclopedia of Genes and Genomes (KEGG), The Gene Ontology Resource, UniProt-GOA, and others. Given a gene or protein list, MOET analysis identifies significantly overrepresented ontology terms using a hypergeometric test and provides nominal and Bonferroni corrected P-values and odds ratios for the overrepresented terms. The results are shown as a downloadable list of terms with and without Bonferroni correction, and a graph of the P-values and number of annotated genes for each term in the list. MOET can be accessed freely from https://rgd.mcw.edu/rgdweb/enrichment/start.html.

Subglacial hydrology is a leading control on basal friction and the dynamics of glaciers and ice sheets. At low discharge, subglacial water flows through high-pressure, sheet-like systems that lead to low effective pressures. However, at high discharge, subglacial water melts the overlying ice into localized channels that efficiently remove water from the bed, thereby increasing effective pressure and basal friction. Recent observations suggest channelized subglacial flow exists beneath Thwaites Glacier, yet it remains unclear if stable channelization is feasible in West Antarctica, where surface melting is nonexistent and water at the bed is limited. Here, we use the MPAS-Albany Land Ice model to run a suite of over 130 subglacial hydrology simulations of Thwaites Glacier across a wide range of physical parameter choices to assess the likelihood of channelization. We then narrow our range of viable simulations by comparing modeled water thicknesses to previously observed radar specularity content, which indicates flat, spatially extensive water bodies at the bed. In all of our data-compatible simulations, stable channels reliably form within 100–200 km of the grounding line and reach individual discharge rates of 35–110 m3 s−1 at the ice–ocean boundary. While only one to two channels typically form across the 200 km width of the glacier in our simulations, their high efficiency drains water across the entire lateral extent of the glacier. We posit the large catchment size of Thwaites Glacier, its funnel-like geometry, and high basal melt rates together accumulate enough water to form stable channels. No simulations resembled observed specularity content when channelization was disabled. Our results suggest channelized subglacial hydrology has two consequences for Thwaites Glacier dynamics: (i) amplifying submarine melting of the terminus and ice shelf while (ii) simultaneously raising effective pressure within 100 km of the grounding line and increasing basal friction. The distribution of effective pressure implied from our modeling differs from parameterizations typically used in large-scale ice sheet models, suggesting the development of more process-based parameterizations may be necessary.

The Ensemble Mars Atmosphere Reanalysis System (EMARS) dataset version 1.0 contains hourly gridded atmospheric variables for the planet Mars, spanning Mars Year (MY) 24 through 33 (1999 through 2017). A reanalysis represents the best estimate of the state of the atmosphere by combining observations that are sparse in space and time with a dynamical model and weighting them by their uncertainties. EMARS uses the Local Ensemble Transform Kalman Filter (LETKF) for data assimilation with the GFDL/NASA Mars Global Climate Model (MGCM). Observations that are assimilated include the Thermal Emission Spectrometer (TES) and Mars Climate Sounder (MCS) temperature retrievals. The dataset includes gridded fields of temperature, wind, surface pressure, as well as dust, water ice, CO2 surface ice and other atmospheric quantities. Reanalyses are useful for both science and engineering studies, including investigations of transient eddies, the polar vortex, thermal tides and dust storms, and during spacecraft operations.

The Energy Exascale Earth System Model (E3SM) is a new coupled Earth system model sponsored by the U.S Department of Energy. Here we present E3SM global simulations using active ocean and sea ice that are driven by the Coordinated Ocean‐ice Reference Experiments II (CORE‐II) interannual atmospheric forcing data set. The E3SM ocean and sea ice components are MPAS‐Ocean and MPAS‐Seaice, which use the Model for Prediction Across Scales (MPAS) framework and run on unstructured horizontal meshes. For this study, grid cells vary from 30 to 60 km for the low‐resolution mesh and 6 to 18 km at high resolution. The vertical grid is a structured z‐star coordinate and uses 60 and 80 layers for low and high resolution, respectively. The lower‐resolution simulation was run for five CORE cycles (310 years) with little drift in sea surface temperature (SST) or heat content. The meridional heat transport (MHT) is within observational range, while the meridional overturning circulation at 26.5°N is low compared to observations. The largest temperature biases occur in the Labrador Sea and western boundary currents (WBCs), and the mixed layer is deeper than observations at northern high latitudes in the winter months. In the Antarctic, maximum mixed layer depths (MLD) compare well with observations, but the spatial MLD pattern is shifted relative to observations. Sea ice extent, volume, and concentration agree well with observations. At high resolution, the sea surface height compares well with satellite observations in mean and variability.

The land ice contribution to global mean sea level rise has not yet been predicted 1 using ice sheet and glacier models for the latest set of socio-economic scenarios, nor using coordinated exploration of uncertainties arising from the various computer models involved. Two recent international projects generated a large suite of projections using multiple models 2–8, but primarily used previous-generation scenarios 9 and climate models 10, and could not fully explore known uncertainties. Here we estimate probability distributions for these projections under the new scenarios 11,12 using statistical emulation of the ice sheet and glacier models. We find that limiting global warming to 1.5 degrees Celsius would halve the land ice contribution to twenty-first-century sea level rise, relative to current emissions pledges. The median decreases from 25 to 13 centimetres sea level equivalent (SLE) by 2100, with glaciers responsible for half the sea level contribution. The projected Antarctic contribution does not show a clear response to the emissions scenario, owing to uncertainties in the competing processes of increasing ice loss and snowfall accumulation in a warming climate. However, under risk-averse (pessimistic) assumptions, Antarctic ice loss could be five times higher, increasing the median land ice contribution to 42 centimetres SLE under current policies and pledges, with the 95th percentile projection exceeding half a metre even under 1.5 degrees Celsius warming. This would severely limit the possibility of mitigating future coastal flooding. Given this large range (between 13 centimetres SLE using the main projections under 1.5 degrees Celsius warming and 42 centimetres SLE using risk-averse projections under current pledges), adaptation planning for twenty-first-century sea level rise must account for a factor-of-three uncertainty in the land ice contribution until climate policies and the Antarctic response are further constrained.

Projections of the sea level contribution from the Greenland and Antarctic ice sheets (GrIS and AIS) rely on atmospheric and oceanic drivers obtained from climate models. The Earth System Models participating in the Coupled Model Intercomparison Project phase 6 (CMIP6) generally project greater future warming compared with the previous Coupled Model Intercomparison Project phase 5 (CMIP5) effort. Here we use four CMIP6 models and a selection of CMIP5 models to force multiple ice sheet models as part of the Ice Sheet Model Intercomparison Project for CMIP6 (ISMIP6). We find that the projected sea level contribution at 2100 from the ice sheet model ensemble under the CMIP6 scenarios falls within the CMIP5 range for the Antarctic ice sheet but is significantly increased for Greenland. Warmer atmosphere in CMIP6 models results in higher Greenland mass loss due to surface melt. For Antarctica, CMIP6 forcing is similar to CMIP5 and mass gain from increased snowfall counteracts increased loss due to ocean warming.

Humboldt Glacier drains ~5% of the Greenland Ice Sheet and has retreated and accelerated since the late 1990s. The northern section of the terminus has retreated towards an overdeepening in the glacier bed that extends tens of kilometers towards the ice sheet interior, raising the possibility of a rapid increase in ice discharge and retreat in the near future. Here we investigate the potential 21st century sea-level contribution from Humboldt Glacier with the MPAS-Albany Land Ice (MALI) ice sheet model. First, we optimize the basal friction field using observations of surface velocity and ice surface elevation to obtain an initial condition for the year 2007. Next, we tune parameters for calving, basal friction, and submarine melt to match the observed retreat rates and surface velocity changes. We then simulate glacier evolution to 2100 under a range of climate forcings from the Ice Sheet Model Intercomparison Project for CMIP6 (ISMIP6), using ocean temperatures from the MIROC5 Earth System Model, with surface mass balance and subglacial discharge from MAR3.9/MIROC5. Our simulations predict ~3.5 mm of sea-level rise from the retreat of Humboldt Glacier by 2100 for RCP8.5, and ~1 mm for RCP2.6. The results are insensitive to the choice of calving parameters for grounded ice, but a low stress threshold for calving from floating ice is necessary to initiate retreat. We find that a highly plastic basal friction law is required to reproduce the observed acceleration, but the choice of basal friction law does not have a large effect on the magnitude of sea-level contribution by 2100 because much of the ice is at present close to floatation in the areas that retreat most significantly. Instead, the majority of ice mass loss comes from increasingly negative surface mass balance. Preliminary results from experiments with a subglacial hydrology model suggest that the simple treatment of subglacial discharge used in our 21st century projections (as used in the ISMIP6-Greenland protocol) underestimates spatial variability of melting at the glacier front but gives a reasonable approximation of total melt. When compared to the recent ISMIP6 estimates of 60–140 mm sea-level rise from the entire Greenland Ice Sheet by 2100, our estimate of 3.5 mm from Humboldt Glacier indicates a significant but far from dominant contribution from this single large outlet.

The distribution of plastic in the ocean is poorly constrained, with the mass of floating plastic at the ocean surface being orders of magnitude smaller than estimated plastic inputs. Coastlines likely contain significant amounts of plastic, but inconsistent methodologies between beached plastic observations prevent determining the mass and distribution of globally beached plastic. We present Lagrangian model sensitivity experiments to estimate the beached fraction of marine plastic and to investigate the global distribution of beached plastic on coastlines.We perform simulations where particles, representing masses of floating plastic, are inserted at the ocean coasts. The particles are then advected by surface currents (HYCOM/NCODA global reanalysis and surface Stokes drift from the WaveWatch III global reanalysis) for 5 years. Beaching is parametrized stochastically using exponentional probability. Here, we test the sensitivity to e-folding time scales between 1 and 100 days, applied when plastic is within the coastal zone, within 10km of the nearest coastline. Resuspension of beached plastic is parameterised exponentially with an e-folding timescale between 69 and 273 days. No other loss processes are implemented.Between 39-95% of floating plastic mass is beached after 5 years, with the beached fraction depending on the ratio between the beaching and resuspension timescales. In all simulations, at least 77% of floating plastic mass is found either beached or within the coastal zone, indicating coastal regions are a significant reservoir of mismanaged terrestrial plastic. However, plastic entering the ocean from islands or near energetic boundary currents is more likely to reach the open ocean. The distribution of beached plastic is closely related to the input distribution, with the highest concentrations found in Southeast Asia and the Mediterranean.Our results highlight coastlines and coastal waters as important reservoirs of marine plastic debris and indicate a need for greater understanding of plastic transport near and at the coastlines. Furthermore, improved representation of plastic beaching can help study marine plastic fragmentation, as mechanical stress during the transitions between coastlines and coastal waters and the increased UV exposure of beached plastic likely contribute to the fragmentation.

Projections of the sea level contribution from the Greenland and Antarctic ice sheets (GrIS and AIS) rely on atmospheric and oceanic drivers obtained from climate models. The Earth System Models participating in the Coupled Model Intercomparison Project phase 6 (CMIP6) generally project greater future warming compared with the previous Coupled Model Intercomparison Project phase 5 (CMIP5) effort. Here we use four CMIP6 models and a selection of CMIP5 models to force multiple ice sheet models as part of the Ice Sheet Model Intercomparison Project for CMIP6 (ISMIP6). We find that the projected sea level contribution at 2100 from the ice sheet model ensemble under the CMIP6 scenarios falls within the CMIP5 range for the Antarctic ice sheet but is significantly increased for Greenland. Warmer atmosphere in CMIP6 models results in higher Greenland mass loss due to surface melt. For Antarctica, CMIP6 forcing is similar to CMIP5 and mass gain from increased snowfall counteracts increased loss due to ocean warming.

Global coastlines potentially contain significant amounts of plastic debris, with harmful implications for marine and coastal ecosystems, fisheries and tourism. However, the global amount, distribution and origin of plastic debris on beaches and in coastal waters is currently unknown. Here we analyze beaching and resuspension scenarios using a Lagrangian particle transport model. Throughout the first 5 years after entering the ocean, the model indicates that at least 77% of positively buoyant marine plastic debris (PBMPD) released from land-based sources is either beached or floating in coastal waters, assuming no further plastic removal from beaches or the ocean surface. The highest concentrations of beached PBMPD are found in Southeast Asia, caused by high plastic inputs from land and limited offshore transport, although the absolute concentrations are generally overestimates compared to field measurements. The modeled distribution on a global scale is only weakly influenced by local variations in resuspension rates due to coastal geomorphology. Furthermore, there are striking differences regarding the origin of the beached plastic debris. In some exclusive economic zones (EEZ), such as the Indonesian Archipelago, plastic originates almost entirely from within the EEZ while in other EEZs, particularly remote islands, almost all beached plastic debris arrives from remote sources. Our results highlight coastlines and coastal waters as important reservoirs of marine plastic debris and limited transport of PBMPD between the coastal zone and the open ocean.

We investigate the task of entity counting in overhead imagery from the perspective of re-purposing representations learned from ground imagery, e.g., ImageNet, via feature adaptation. We explore two directions of feature adaptation and analyze their performances using two popular aerial datasets for vehicle counting: PUCPR+ and CARPK. First, we explore proxy self-supervision tasks such as RotNet, jigsaw, and image inpainting to re-fine the pretrained representation. Second, we insert additional network layers to adaptively select suitable features (e.g., squeeze and excitation blocks) or impose desired properties (e.g., using active rotating filters for rotation invariance). Our experimental results show that different adaptations produce different amounts of performance improvements depending on data characteristics. Overall, we achieve a mean absolute error (MAE) of 3.71 and 5.93 on the PUCPR+ and CARPK datasets, respectively, outperforming the previous state of the art: MAEs of 5.24 for PUCPR+ and 7.48 for CARPK.

Sub-ice-shelf melting plays a critical role in the dynamics of the Antarctic Ice Sheet and also feeds back on the regional climate, transforming ocean properties (e.g., affecting deep-water production and sea-ice formation). A full understanding of these processes, as well as the ability to project their response to a changing climate, requires Earth System Models (ESMs) that include coupling with ice-sheet processes. However, biases in regional Antarctic climate can be amplified through sub-ice-shelf melting, and biased melt rates can have significant adverse effects on ice-sheet model initialization and evolution. In preparation for inclusion of dynamic ice sheets in ESMs, this presentation discusses our recent experience in understanding the causes of biases in ocean properties on the Antarctic continental shelf and their relationship to ice-shelf melting. Differences in model behavior across configurations and simulations using the Energy Exascale Earth System Model (E3SM) demonstrates a sensitivity of melt rates to climate. We assess the sensitivity of those melt rates to changes in the region’s climate, including freshening on the continental shelf and shoaling of the thermocline. We also show that ice-shelf meltwater feeds back onto the climate, for example, by affecting melting under neighboring ice shelves, sometimes dramatically so. We demonstrate that significant reductions in melt-rate biases can be achieved through modifications to ocean model mixing parameterizations. This work charts a path forward for configuring ESMs to produce realistic Antarctic melt rates.

As a step in a Dynamic Data-Driven Applications Systems (DDDAS) method to characterize the background in a vehicle tracking problem, we extend the application of deep learning to a hyperspectral dataset (the AeroRIT dataset) to evaluating network uncertainty. Expressing uncertainty information is crucial for evaluating what additional information is needed in the DDDAS algorithm and where more resources are required. Hyperspectral signatures tend to be very noisy, when captured from an aerial flight and a slight shift in the atmospheric conditions can alter the signals significantly, which in turn may affect the trained network’s classifications. In this work, we apply Deep Ensembles, Monte Carlo Dropout and Batch Ensembles and study their effects with respect to achieving robust pixel-level identifications by expressing the uncertainty within the trained networks on the task of semantic segmentation. We modify the U-Net-m architecture from the AeroRIT paper to account for the frameworks and present our results as a step towards accounting for sensitive changes in hyperspectral signals.

We introduce MPAS-Albany Land Ice (MALI) v6.0, a new variable-resolution land ice model that uses unstructured Voronoi grids on a plane or sphere. MALI is built using the Model for Prediction Across Scales (MPAS) framework for developing variable-resolution Earth system model components and the Albany multi-physics code base for the solution of coupled systems of partial differential equations, which itself makes use of Trilinos solver libraries. MALI includes a three-dimensional first-order momentum balance solver (Blatter–Pattyn) by linking to the Albany-LI ice sheet velocity solver and an explicit shallow ice velocity solver. The evolution of ice geometry and tracers is handled through an explicit first-order horizontal advection scheme with vertical remapping. The evolution of ice temperature is treated using operator splitting of vertical diffusion and horizontal advection and can be configured to use either a temperature or enthalpy formulation. MALI includes a mass-conserving subglacial hydrology model that supports distributed and/or channelized drainage and can optionally be coupled to ice dynamics. Options for calving include “eigencalving”, which assumes that the calving rate is proportional to extensional strain rates. MALI is evaluated against commonly used exact solutions and community benchmark experiments and shows the expected accuracy. Results for the MISMIP3d benchmark experiments with MALI's Blatter–Pattyn solver fall between published results from Stokes and L1L2 models as expected. We use the model to simulate a semi-realistic Antarctic ice sheet problem following the initMIP protocol and using 2 km resolution in marine ice sheet regions. MALI is the glacier component of the Energy Exascale Earth System Model (E3SM) version 1, and we describe current and planned coupling to other E3SM components.

We propose a new method for estimating gill net selectivity which estimates the probabilities leading to retention by analyzing both the fish morphology and the mesh geometry. This method estimates the number of fish approaching and contacting gill nets of different mesh sizes as an intermediate step towards computing the selectivity. Instead of assuming an underlying probability distribution as in indirect methods, we split the entire interaction between a fish and the gill net into several stages, each with its own probability. All the necessary parameters to compute these probabilities can be obtained from measurements of the fish, knowledge of the mesh geometry, and catch data from different mesh sizes. The framework offers three pathways for computing the total number of fish contacting the gill nets and has the capability to use both wedged and entangled fish in the analysis. As a proof of concept, the method is applied to catch data for cod (G. morhua) and Dolly Varden (S. malma) to estimate the number of fish contacting the gill nets in both cases. By estimating the number of fish contacting the gill net in addition to the selectivity, this method provides an important step towards deriving estimates of fish density in a particular fishery from gill net measurement.

The impact of summer surface melt on the dynamics of the Greenland Ice Sheet is modulated by the state of the subglacial hydrologic system. Studies of ice motion indicate that efficiency of the subglacial system increases over the melt season, decreasing the sensitivity of ice motion to surface melt. However, these inferences are based on limited indirect observations of the subglacial hydrologic system that leave many factors poorly constrained, particularly the presence and stability of subglacial channels. Here we use observations from 11 GPS stations, from which we derive ice velocity, longitudinal strain rates, and basal uplift, alongside observations of surface ablation and supraglacial lake drainage events, to explore the coevolution of ice motion and the subglacial hydrologic system in the Pakitsoq region of western Greenland during the 2011 melt season. We observe ice acceleration after the onset of local surface melting, followed by gradual ice deceleration, consistent with the pattern expected from increased subglacial drainage efficiency. Supraglacial lake drainages appear to precipitate ice deceleration and increased basal traction, suggesting that lake drainages effectively reorganize the local subglacial hydrologic system into a more efficient state that persists through the remainder of the melt season. At high elevations, ice velocity and inferred basal uplift suggest that continued cavity growth or sediment behavior, not subglacial channelization, drive the apparent increase in subglacial efficiency. Our results further indicate that these transient perturbations are critical in the seasonal evolution of ice motion.

Moulins permit access of surface meltwater to the glacier bed, causing basal lubrication and ice speedup in the ablation zone of western Greenland during summer. Despite the substantial impact of moulins on ice dynamics, the conditions under which they form are poorly understood. We assimilate a time-series of ice surface velocity from a network of eleven Global Positioning System receivers into an ice sheet model to estimate ice sheet stresses during winter, spring, and summer in an approx. 30 x 10 km region. Surface-parallel von Mises stress increases slightly during spring speedup and early summer, sufficient to allow formation of 16% of moulins mapped in the study area. In contrast, 63% of moulins experience stresses over the tensile strength of ice during a short (hours) supraglacial lake drainage event. Lake drainages appear to control moulin density, which is itself a control on subglacial drainage efficiency and summer ice velocities.

To address the pressing need to better understand the behavior and complex interaction of ice sheets within the global Earth system, significant development of continental-scale, dynamical ice sheet models is underway. Concurrent to the development of the Community Ice Sheet Model (CISM), the corresponding verification and validation (V&V) process is being coordinated through a new, robust, Python-based extensible software package, the Land Ice Verification and Validation toolkit (LIVVkit). Incorporated into the typical ice sheet model development cycle, it provides robust and automated numerical verification, software verification, performance validation, and physical validation analyses on a variety of platforms, from personal laptops to the largest supercomputers. LIVVkit operates on sets of regression test and reference data sets, and provides comparisons for a suite of community prioritized tests, including configuration and parameter variations, bit-for-bit evaluation, and plots of model variables to indicate where differences occur. LIVVkit also provides an easily extensible framework to incorporate and analyze results of new intercomparison projects, new observation data, and new computing platforms. LIVVkit is designed for quick adaptation to additional ice sheet models via abstraction of model specific code, functions, and configurations into an ice sheet model description bundle outside the main LIVVkit structure. Ultimately, through shareable and accessible analysis output, LIVVkit is intended to help developers build confidence in their models and enhance the credibility of ice sheet models overall.

Epithelial-mesenchymal transition (EMT) is a fundamental biological process that plays a central role in embryonic development, tissue regeneration, and cancer metastasis. Transforming growth factor-β (TGFβ) is a potent inducer of this cellular transition, which is composed of transitions from an epithelial state to intermediate or partial EMT state(s) to a mesenchymal state. Using computational models to predict cell state transitions in a specific experiment is inherently difficult for reasons including model parameter uncertainty and error associated with experimental observations. In this study, we demonstrate that a data-assimilation approach using an ensemble Kalman filter, which combines limited noisy observations with predictions from a computational model of TGFβ-induced EMT, can reconstruct the cell state and predict the timing of state transitions. We used our approach in proof-of-concept “synthetic” in silico experiments, in which experimental observations were produced from a known computational model with the addition of noise. We mimic parameter uncertainty in in vitro experiments by incorporating model error that shifts the TGFβ doses associated with the state transitions and reproduces experimentally observed variability in cell state by either shifting a single parameter or generating “populations” of model parameters. We performed synthetic experiments for a wide range of TGFβ doses, investigating different cell steady-state conditions, and conducted parameter studies varying properties of the data-assimilation approach including the time interval between observations and incorporating multiplicative inflation, a technique to compensate for underestimation of the model uncertainty and mitigate the influence of model error. We find that cell state can be successfully reconstructed and the future cell state predicted in synthetic experiments, even in the setting of model error, when experimental observations are performed at a sufficiently short time interval and incorporate multiplicative inflation. Our study demonstrates the feasibility and utility of a data-assimilation approach to forecasting the fate of cells undergoing EMT.

Resources for rat researchers are extensive, including strain repositories and databases all around the world. The Rat Genome Database (RGD) serves as the primary rat data repository, providing both manual and computationally collected data from other databases.

Epithelial-mesenchymal transition (EMT) is a fundamental biological process that plays a central role in embryonic development, tissue regeneration, and cancer metastasis. Transforming growth factor-β(TGFβ) is a major and potent inducer of this cellular transition, which is comprised of transitions from an epithelial state to an intermediate or partial EMT state, then to a mesenchymal state. Using computational models to predict state transitions in a specific experiment is inherently difficult for many reasons, including model parameter uncertainty and the error associated with experimental observations. In this study, we demonstrate that a data-assimilation approach using an ensemble Kalman filter, which combines limited noisy observations with predictions from a computational model of TGFβ-induced EMT, can reconstruct the cell state and predict the timing of state transitions. We used our approach in proof-of-concept “synthetic”in silicoexperiments, in which experimental observations were produced from a known computational model with the addition of noise. We mimic parameter uncertainty inin vitroexperiments by incorporating model error that shifts the TGFβdoses associated with the state transitions. We performed synthetic experiments for a wide range of TGFβdoses to investigate different cell steady state conditions, and we conducted a parameter study varying several properties of the data-assimilation approach, including the time interval between observations, and incorporating multiplicative inflation, a technique to compensate for underestimation of the model uncertainty and mitigate the influence of model error. We find that cell state can be successfully reconstructed in synthetic experiments, even in the setting of model error, when experimental observations are performed at a sufficiently short time interval and incorporate multiplicative inflation. Our study demonstrates a feasible proof-of-concept for a data assimilation approach to forecasting the fate of cells undergoing EMT.Author summaryEpithelial-mesenchymal transition (EMT) is a biological process in which an epithelial cell loses core epithelial-like characteristics, such as tight cell-to-cell adhesion, and gains core mesenchymal-like characteristics, such as an increase in cell motility. EMT is a multistep process, in which the cell undergoes transitions from epithelial state to a partial or intermediate state, and then from a partial state to a mesenchymal state. In this study, we apply data assimilation to improve prediction of these state transitions. Data assimilation is an approach well known in the weather forecasting community, in which experimental observations are iteratively combined with predictions from a dynamical model to provide an improved estimation of both observed and unobserved system states. We show that this data assimilation approach can reconstruct cell state measurements and predict state transition dynamics using noisy observations, while minimizing the error produced by the limitations and imperfections of the dynamical model.

The sea level contribution of the Antarctic ice sheet constitutes a large uncertainty in future sea level projections. Here we apply a linear response theory approach to 16 state-of-the-art ice sheet models to estimate the Antarctic ice sheet contribution from basal ice shelf melting within the 21st century. The purpose of this computation is to estimate the uncertainty that arises from large uncertainty in the external forcing that future warming may exert onto the ice sheet. While ice shelf melting is considered to be a major if not the largest perturbation of the ice sheet's flow into the ocean, the approach is neglecting a number of processes such as surface mass balance related contributions and mechanisms. In assuming linear response theory, we are able to capture complex temporal responses of the ice sheets, but we neglect any dampening or self-amplifying processes. This is particularly relevant in situations where an instability is dominating the ice loss. Results obtained here are thus relevant in particular wherever the ice loss is dominated by the forcing as opposed to an internal instability, for example in strong warming scenarios. In order to allow for comparison the methodology was chosen to be exactly the same as in an earlier study (Levermann et al., 2014), but with 16 instead of 5 ice sheet models. We include uncertainty in the atmospheric warming response to carbon emissions (full range of CMIP-5 climate model sensitivities), uncertainty in the oceanic transport to the Southern Ocean (obtained from the time-delayed and scaled oceanic subsurface warming in CMIP-5 models in relation to the global mean surface warming) and the observed range of responses of basal ice shelf melting to oceanic warming outside the ice shelf cavity. This uncertainty in basal ice shelf melting is then convoluted with the linear response functions of each of the 16 ice sheet models to obtain the ice flow response to the individual global warming path. The model median for the observational period from 1992 to 2017 is 9.6 mm with a likely range between 5.2 mm and 20.3 mm compared to the observed sea-level contribution from Antarctica of 7.4 mm with a standard deviation of 3.7 mm (Shepherd et al., 2018). For the so-called business-as-usual warming path, RCP-8.5, we obtain a median contribution of the Antarctic ice sheet to global mean sea-level rise within the 21st century of 17 cm with a likely range (66-percentile around the mean) between 9 cm and 36 cm and a very likely range (90-percentile around the mean) between 6 cm and 59 cm. For the RCP-2.6 warming path which will keep the global mean temperature below two degrees of global warming and is thus consistent with the Paris Climate Agreement yields a median of 13 cm of global mean sea-level contribution. The likely range for the RCP-2.6 scenario is between 7 cm and 25 cm and the very likely range is between 5 cm and 39 cm. The structural uncertainties in the method do not allow an interpretation of any higher uncertainty percentiles. We provide projections for the five Antarctic regions and for each model and each scenario, separately. The rate of sea level contribution is highest under the RCP-8.5 scenario. The maximum within the 21st century of the median value is 4 cm per decade with a likely range between 2 cm/dec and 8 cm/dec and a very likely range between 1 cm/dec and 13 cm/dec.

The increased sensitivity of modern hyperspectral line-scanning systems has led to the development of imaging systems that can acquire each line of hyperspectral pixels at very high data rates (in the 200–400 Hz range). These data acquisition rates present an opportunity to acquire full hyperspectral scenes at rapid rates, enabling the use of traditional push-broom imaging systems as low-rate video hyperspectral imaging systems. This paper provides an overview of the design of an integrated system that produces low-rate video hyperspectral image sequences by merging a hyperspectral line scanner, operating in the visible and near infra-red, with a high-speed pan-tilt system and an integrated IMU-GPS that provides system pointing. The integrated unit is operated from atop a telescopic mast, which also allows imaging of the same surface area or objects from multiple view zenith directions, useful for bi-directional reflectance data acquisition and analysis. The telescopic mast platform also enables stereo hyperspectral image acquisition, and therefore, the ability to construct a digital elevation model of the surface. Imaging near the shoreline in a coastal setting, we provide an example of hyperspectral imagery time series acquired during a field experiment in July 2017 with our integrated system, which produced hyperspectral image sequences with 371 spectral bands, spatial dimensions of 1600 × 212, and 16 bits per pixel, every 0.67 s. A second example times series acquired during a rooftop experiment conducted on the Rochester Institute of Technology campus in August 2017 illustrates a second application, moving vehicle imaging, with 371 spectral bands, 16 bit dynamic range, and 1600 × 300 spatial dimensions every second.

Reconstructions of excitation patterns in cardiac tissue must contend with uncertainties due to model error, observation error, and hidden state variables. The accuracy of these state reconstructions may be improved by efforts to account for each of these sources of uncertainty, in particular, through the incorporation of uncertainty in model specification and model dynamics. To this end, we introduce stochastic modeling methods in the context of ensemble-based data assimilation and state reconstruction for cardiac dynamics in one- and three-dimensional cardiac systems. We propose two classes of methods, one following the canonical stochastic differential equation formalism, and another perturbing the ensemble evolution in the parameter space of the model, which are further characterized according to the details of the models used in the ensemble. The stochastic methods are applied to a simple model of cardiac dynamics with fast-slow time-scale separation, which permits tuning the form of effective stochastic assimilation schemes based on a similar separation of dynamical time scales. We find that the selection of slow or fast time scales in the formulation of stochastic forcing terms can be understood analogously to existing ensemble inflation techniques for accounting for finite-size effects in ensemble Kalman filter methods; however, like existing inflation methods, care must be taken in choosing relevant parameters to avoid over-driving the data assimilation process. In particular, we find that a combination of stochastic processes-analogously to the combination of additive and multiplicative inflation methods-yields improvements to the assimilation error and ensemble spread over these classical methods.

Tracking vehicles through dense environments is an important and challenging task that is mostly tackled using visible and near IR wavelengths. Hyperspectral imaging is known to improve the robustness of target identification, but the massive increase in data created is usually prohibitive for tracking many targets. We present a persistent real-time aerial target tracking system, taking advantage of an adaptive, multimodal sensor concept and blending the hyperspectral likelihoods with kinematic likelihoods in a multidimensional assignment framework. The adaptive sensor is capable of providing wide field of view panchromatic images as well as the spectra of small number of pixels. The proposed system does not require large amount of hyperspectral data collection as we focus on tracking fewer number of targets with higher persistency. This overcomes the data challenge of hyperspectral tracking by following dynamic data-driven application systems (DDDAS) principles to control hyperspectral data collection where most beneficial. The DDDAS framework for controlling hyperspectral data collection is developed by incorporating prior information from the filter movement predictions and information from motion detection. The proposed multidimensional hyperspectral feature-aided tracker is compared to a 2-D hyperspectral feature-aided tracker and another cascaded hyperspectral data based tracker by generating a synthetic, realistic, aerial video on a dense scene.

The McMurdo Dry Valleys of Antarctica host the coldest and driest ecosystem on Earth, which is acutely sensitive to the availability of water coming from glacial runoff. We modeled the spatial variability in ablation and assessed climate sensitivity of the glacier ablation zones using 16 years of meteorological and surface mass-balance observations collected in Taylor Valley. Sublimation was the primary form of mass loss over much of the ablation zones, except for near the termini where melt, primarily below the surface, dominated. Microclimates in ~10 m scale topographic basins generated melt rates up to ten times higher than over smooth glacier surfaces. In contrast, the vertical terminal cliffs on the glaciers can have higher or lower melt rates than the horizontal surfaces due to differences in incoming solar radiation. The model systematically underpredicted ablation for the final 5 years studied, possibly due to an increase of windblown sediment. Surface mass-balance sensitivity to temperature was ~−0.02 m w.e. K−1, which is among the smallest magnitudes observed globally. We also identified a high sensitivity to ice albedo, with a decrease of 0.02 having similar effects as a 1 K increase in temperature, and a complex sensitivity to wind speed.

Modelling of cardiac electrical behaviour has led to important mechanistic insights, but important challenges, including uncertainty in model formulations and parameter values, make it difficult to obtain quantitatively accurate results. An alternative approach is combining models with observations from experiments to produce a data-informed reconstruction of system states over time. Here, we extend our earlier data-assimilation studies using an ensemble Kalman filter to reconstruct a three-dimensional time series of states with complex spatio-temporal dynamics using only surface observations of voltage. We consider the effects of several algorithmic and model parameters on the accuracy of reconstructions of known scroll-wave truth states using synthetic observations. In particular, we study the algorithm’s sensitivity to parameters governing different parts of the process and its robustness to several model-error conditions. We find that the algorithm can achieve an acceptable level of error in many cases, with the weakest performance occurring for model-error cases and more extreme parameter regimes with more complex dynamics. Analysis of the poorest-performing cases indicates an initial decrease in error followed by an increase when the ensemble spread is reduced. Our results suggest avenues for further improvement through increasing ensemble spread by incorporating additive inflation or using a parameter or multi-model ensemble. This article is part of the theme issue ‘Uncertainty quantification in cardiac and cardiovascular modelling and simulation’.

We model three-dimensional motion of plastic pollution in Lake Erie due to advection, density-driven sinking, and turbulent mixing using a Lagrangian transport model to explore the distribution of plastic in the water column and sediment. Nine polymer types that make up over 75% of predicted worldwide plastic waste were modeled, and the model keeps track of particles that hit the bottom to represent deposition. Modeled spatial distributions are compared with samples from the surface to calibrate the model and derive estimates for the mass of plastic in the lake volume and the flux of plastic into the sediment. The mass estimate of 381 ± 204 metric tons is two orders of magnitude larger than previous surface estimates, though still a fraction of predicted yearly input. The results are a step towards closing the plastic mass balance in Lake Erie as well as understanding the transport of plastic into the sediment.

Ice temperature profiles from the Greenland Ice Sheet contain information on the deformation history, past climates and recent warming. We present full-depth temperature profiles from two drill sites on a flow line passing through Swiss Camp, West Greenland. Numerical modeling reveals that ice temperatures are considerably higher than would be expected from heat diffusion and dissipation alone. The possible causes for this extra heat are evaluated using a Lagrangian heat flow model. The model results reveal that the observations can be explained with a combination of different processes: enhanced dissipation (strain heating) in ice-age ice, temperate paleo-firn, and cryo-hydrologic warming in deep crevasses.

We introduce MPAS-Albany Land Ice (MALI), a new, variable resolution land ice model that uses unstructured Voronoi grids on a plane or sphere. MALI is built using the Model for Prediction Across Scales (MPAS) framework for developing variable resolution Earth System Model components and the Albany multi-physics code base for solution of coupled systems of partial-differential equations, which itself makes use of Trilinos solver libraries. MALI includes a three-dimensional, first-order momentum balance solver ("Blatter-Pattyn") by linking to the Albany-LI ice sheet velocity solver, as well as an explicit shallow ice velocity solver. Evolution of ice geometry and tracers is handled through an explicit first-order horizontal advection scheme with vertical remapping. Evolution of ice temperature is treated using operator splitting of vertical diffusion and horizontal advection and can be configured to use either a temperature or enthalpy formulation. MALI includes a mass-conserving subglacial hydrology model that supports distributed and/or channelized drainage and can optionally be coupled to ice dynamics. Options for calving include "eigencalving", which assumes calving rate is proportional to extensional strain rates. MALI is evaluated against commonly used exact solutions and community benchmark experiments and shows the expected accuracy. We report first results for the MISMIP3d benchmark experiments for a Blatter-Pattyn type model and show that results fall in-between those of models using Stokes flow and L1L2 approximations. We use the model to simulate a semi-realistic Antarctic Ice Sheet problem for 1100 years at 20 km resolution. MALI is the glacier component of the Energy Exascale Earth System Model (E3SM) version 1, and we describe current and planned coupling to other components.

Subglacial hydrology plays a key role in many glaciological processes, including ice dynamics via the modulation of basal sliding. Owing to the lack of an overarching theory, however, a variety of model approximations exist to represent the subglacial drainage system. The Subglacial Hydrology Model Intercomparison Project (SHMIP) provides a set of synthetic experiments to compare existing and future models. We present the results from 13 participating models with a focus on effective pressure and discharge. For many applications (e.g. steady states and annual variations, low input scenarios) a simple model, such as an inefficient-system-only model, a flowline or lumped model, or a porous-layer model provides results comparable to those of more complex models. However, when studying short term (e.g. diurnal) variations of the water pressure, the use of a two-dimensional model incorporating physical representations of both efficient and inefficient drainage systems yields results that are significantly different from those of simpler models and should be preferentially applied. The results also emphasise the role of water storage in the response of water pressure to transient recharge. Finally, we find that the localisation of moulins has a limited impact except in regions of sparse moulin density.

We consider methods to address the optical feature-aided remote sensing tracking problem for vehicles in a challenging environment. Our approach is to apply the dynamic data driven application systems computing paradigm to implement control of an adaptive sensor. This adaptive sensor acquires a panchromatic image while simultaneously allowing the collection of visible-near infrared spectral data at specified pixels. This sensor holds the promise of delivering the increased accuracy of targeted spectral sensing without the enormous data volume of full spectral images. The target of interest is optimally imaged by the sensor based on the target's forecasted location and motion relative to the extracted content of the background. Background context is both extracted from the image and created from the OpenStreetMap road network. We describe the implementation of the tracking framework and testing of some of the components using simulated imagery created with the digital imaging and remote sensing image generation model. The Gaussian sum filter is employed to solve the data assimilation problem by forming a multimodel forecasting set that is used to increase the robustness and flexibility of tracking. For feature matching, we create an efficient sampling strategy that is informed by the viewing conditions to adaptively determine which pixels to measure spectrally in order to distinguish between different targets using a spectral distance measure.