Skip to content

Library

Articles

Every article on Frontiersin is peer reviewed, permanently archived and free to read.

Showing 12 articles

Original ResearchJournal of Climate & Earth Systems

Quantifying feedback pathways in accelerated Arctic sea-ice thinning, 1995–2024

Elena Varga, Rafael Moreno

Arctic sea ice has thinned substantially over the past three decades, yet the relative contributions of individual feedback processes remain debated. We combine a 30-year record of satellite altimetry with an ensemble of coupled model simulations to partition thinning into albedo, cloud and ocean heat transport contributions. Our results indicate that ocean heat convergence explains a larger share of winter thinning than previously estimated, while surface albedo feedback dominates summer losses. These findings refine projections of seasonally ice-free conditions and highlight observational gaps in the Atlantic sector.

Original ResearchComputational Neuroscience Review

Short-term synaptic plasticity stabilises working memory in sparse spiking networks

Daniel Osei, Lucas Ferreira

Persistent activity has long been proposed as the substrate of working memory, but sustaining it in sparse cortical networks is energetically costly. We show that short-term synaptic facilitation allows information to be held in latent synaptic states, reducing the need for continuous firing. Simulations reproduce the activity-silent dynamics observed in primate prefrontal recordings and predict a characteristic reactivation signature after brief non-specific input.

Original ResearchMolecular Medicine & Therapeutics

Metabolic reprogramming reverses T-cell exhaustion in solid tumour microenvironments

Mei Chen, Sara Lindgren

Exhausted CD8+ T cells limit the efficacy of checkpoint blockade in many solid tumours. Using single-cell transcriptomics and metabolic flux analysis, we identify a mitochondrial bottleneck that marks terminally exhausted populations. Pharmacological restoration of fatty-acid oxidation rescued effector function ex vivo and improved tumour control in combination with anti-PD-1 therapy in murine models.

MethodsApplied Artificial Intelligence Letters

Calibrated uncertainty estimates for machine-learning surrogates in scientific simulation

Lucas Ferreira

Machine-learning surrogates are increasingly used to accelerate expensive physical simulations, but their predictive uncertainty is frequently miscalibrated. We introduce a conformal ensemble procedure that provides finite-sample coverage guarantees without retraining the base model. Across five benchmark simulation tasks, the method achieves nominal coverage with intervals up to 35% narrower than existing baselines.

Original ResearchMarine Ecology & Conservation

Coral reef recovery trajectories following compound marine heatwaves in the Arabian Sea

Aisha Rahman

Compound marine heatwaves are becoming more frequent, yet long-term recovery dynamics of reefs exposed to repeated events are poorly documented. We report eight years of benthic surveys from 24 reef sites and find that recovery depends strongly on structural complexity and herbivore biomass. Sites with intact grazer communities regained 70% of pre-disturbance coral cover within five years.

Original ResearchAdvanced Materials & Energy

Interfacial engineering of sulfide solid electrolytes for long-cycle lithium-metal batteries

Jonas Weber

Sulfide solid electrolytes offer high ionic conductivity but suffer from interfacial degradation against lithium metal. We report an atomic-layer-deposited oxyhalide interlayer that suppresses dendrite nucleation and reduces interfacial resistance by an order of magnitude. Symmetric cells cycled stably for over 2,000 hours at practical current densities.

Original ResearchMolecular Medicine & Therapeutics

Population-scale sequencing reveals rare-variant burden in early-onset cardiometabolic disease

Sara Lindgren, Mei Chen

Rare coding variants contribute to cardiometabolic risk, but their aggregate burden is difficult to quantify. Whole-exome sequencing of 180,000 participants identified 14 genes with significant rare-variant associations, five of which are novel. Carriers of high-impact variants showed onset of disease on average nine years earlier than non-carriers.

Brief ReportJournal of Climate & Earth Systems

Orographic modulation of aerosol–cloud interactions in the subtropical Andes

Rafael Moreno

Mountain terrain strongly modulates cloud microphysics, complicating estimates of aerosol radiative forcing. Combining aircraft observations with large-eddy simulations, we show that orographic lifting enhances droplet activation sensitivity to aerosol loading by up to 40% relative to adjacent plains.

Original ResearchApplied Artificial Intelligence Letters

Energy-efficient event-based vision with neuromorphic recurrent architectures

Daniel Osei, Lucas Ferreira

Event cameras generate sparse, asynchronous data streams well suited to neuromorphic processing. We propose a recurrent spiking architecture that achieves competitive accuracy on gesture and optical-flow benchmarks while consuming an order of magnitude less energy than conventional convolutional baselines on equivalent hardware.

Original ResearchMarine Ecology & Conservation

Assisted recolonisation accelerates kelp forest restoration on urchin barrens

Aisha Rahman, Elena Varga

Kelp forests have declined across many temperate coastlines. In a five-year field experiment, combined urchin removal and spore-bag seeding restored canopy density to reference levels within three years, while removal alone produced slower and patchier recovery.

ReviewAdvanced Materials & Energy

Layered sodium-ion cathodes with suppressed phase transitions for grid-scale storage

Jonas Weber

Sodium-ion batteries are a promising low-cost alternative for stationary storage. This review synthesises recent advances in layered oxide cathodes, focusing on dopant strategies that suppress detrimental phase transitions and improve cycle life under realistic grid operating conditions.

Original ResearchComputational Neuroscience Review

Graph neural networks predict individual cognitive scores from structural connectomes

Daniel Osei

Structural connectomes encode rich information about individual brain organisation. We train graph neural networks on diffusion-MRI-derived connectomes from 9,000 participants and show they outperform kernel baselines in predicting fluid intelligence, with attributions concentrated in fronto-parietal hubs.