Papers

In this study, we demonstrate that compared with traditional centroid-based methods, machine learning methods (particularly transformer-based architectures) achieve superior subpixel position and therefore angular resolution in discretized particle detectors. Using Geant4 Monte Carlo simulated cosmic ray muon data from an 8x8 segmented scintillator detector array, we compare four reconstruction approaches: transformer neural networks, convolutional neural networks, linear regression, and energy-weighted centroids. The transformer architecture achieves the best angular reconstruction with a root mean square error of 1.14{\deg} and a position mean absolute error of 0.24 cm, representing improvements of 2.22x and 6.33x, respectively, over the centroid method. These results enable precise particle trajectory reconstruction for applications in muon tomography and cosmic ray detection.

We show that machine learning methods produce superior particle position reconstruction accuracy in scintillation-based detectors.

The n body problem, fundamental to astrophysics, simulates the motion of n bodies acting under the effect of their own mutual gravitational interactions. Traditional machine learning models that are used for predicting and forecasting trajectories are often data intensive black box models, which ignore the physical laws, thereby lacking interpretability. Whereas Scientific Machine Learning ( Scientific ML ) directly embeds the known physical laws into the machine learning framework. Through robust modelling in the Julia programming language, our method uses the Scientific ML frameworks: Neural ordinary differential equations (NODEs) and Universal differential equations (UDEs) to predict and forecast the system dynamics. In addition, an essential component of our analysis involves determining the forecasting breakdown point, which is the smallest possible amount of training data our models need to predict future, unseen data accurately. We employ synthetically created noisy data to simulate real-world observational limitations. Our findings indicate that the UDE model is much more data efficient, needing only 20% of data for a correct forecast, whereas the Neural ODE requires 90%.

Particle image velocimetry (PIV) and related optical-flow methods are widely used to quantify fluid motion, but their development and evaluation are often hindered by fragmented software, inconsistent interfaces, and limited reproducibility. To address these challenges, we present Flow Gym, a framework for developing, benchmarking, training, and deploying flow-field quantification methods, with a primary focus on PIV. Its core contribution is a standardized interface that allows classical and learning-based algorithms to be integrated, compared, and deployed within a common pipeline. The framework includes JAX implementations and wrappers for existing methods, modular pre-processing and post-processing components, and utilities for training and benchmarking. By leveraging JAX, Flow Gym supports hardware-accelerated execution while remaining interoperable with external implementations from libraries such as OpenCV and PyTorch. It can operate on both synthetic and experimental data and supports the same workflow for offline benchmarking and real-time deployment. Flow Gym is designed to improve reproducibility, reduce barriers to method development, and facilitate the translation of flow-field quantification algorithms from research to experimental settings.

Prior to pioneer surface adhesion, bacteria have to navigate in flows, often in confined environments. While much is known about their individual swimming dynamics, our understanding of their transport properties at the population level remains limited. This is primarily due to the experimental challenges associated with tracking, in three dimensions and under flow, a large sample of these microorganisms. Here we investigate, through fast digital holographic microscopy (DHM), a suspension of Shewanella oneidensis MR-1 bacteria in a confined Poiseuille flow. Based on the analysis of several thousand Lagrangian trajectories, we first demonstrate the ability of DHM to discriminate between the fraction of bacteria that adhere to walls, those that are only convected by the flow, and those for which motility helps them migrate through the channel and cross streamlines. Focusing on the motile part of the population, we report new experimental results concerning the spatial and orientational distributions across the confinement direction. In the central part of the flow, migration and shear trapping are observed and well described by existing kinetic models. Close to the walls, averaging over the entire motile population clearly shows an accumulation layer but with no net orientational order, which contradicts models that neglect hydrodynamic interactions with solid surfaces. We think that DHM coupled with the analysis of large-scale bacterial populations will help to understand their transfer mechanisms from the bulk to surfaces.

The concept of correlation appears straightforward: measurement outcomes coincide, and patterns emerge. For any record of events, the coefficients are uniquely determined. Thus, if correlations change spontaneously, as seen in quantum monogamy, then individual behavior must have changed first. Surprisingly, this is not always true. When two observables are mutually exclusive, they cannot coincide objectively and need to be grouped across time. Yet, sectioning the flow of events into "iterations" is not trivial in this case. Even with blind windows of coincidence, the same order of outcomes can produce different coefficients of correlation, depending on the number of joint measurements. Therefore, quantum monogamy can happen with fixed pre-determined events. A new concept ("subjective correlation") is required to explain this phenomenon.

Cosmological scenarios wherein the cumulative number of spontaneously formed, cognitively impaired, disembodied transient observers is vastly larger than the corresponding number of atypical `ordinary observers' (OOs) formed in the conventional way – essentially via cosmic evolution and gravitational instability – are disqualified in modern cosmology on the grounds of Cognitive Instability – the untrustworsiness of one own's reasoning – let alone the atypicality of OOs like us. According to the concordance $\Lambda$CDM cosmological model – when described in the (expanding) `cosmic frame' – the cosmological expansion is future-eternal. In this frame we are atypical OOs, which are vastly outnumbered by typical Boltzmann Brains (BBs) that spontaneously form via sheer thermal fluctuations in the future-eternal asymptotic de Sitter spacetime. In the case that dark energy (DE) ultimately decays, the cumulative number of transient `Freak Observers' (FOs) formed and destroyed spontaneously by virtue of the quantum uncertainty principle ultimately overwhelms that of OOs. Either possibility is unacceptable. We argue that these unsettling conclusions are artifacts of employing the (default) cosmic frame description in which space expands. When analyzed in the comoving frame, OOs overwhelmingly outnumber both BBs and FOs. This suggests that the dual comoving description is the cognitively stable preferred framework for describing our evolving Universe. In this frame, space is globally static, masses monotonically increase, and the space describing gravitationally bounded objects monotonically contracts.

The exact solution of the two-dimensional (2D) Ising model at an external magnetic field is derived by a modified Clifford algebraic approach. At first, the transfer matrices are analyzed in three representations, i.e., Clifford algebraic representation, transfer tensor representation and schematic representation, to inspect nonlocal effects in this many-body interacting system. It is ensured that nontrivial topological structures exist in this system, which is analogous to (but different with) those in the three-dimensional (3D) Ising model at zero magnetic field. Therefore, the approaches developed for the 3D Ising models are modified to be appropriable for solving analytically the solution of the 2D Ising model at a magnetic field. An additional rotation, serving as a topological Lorentz transformation, is applied for dealing with the topological problems in the present system. The rotation angle for the transformation is determined by Yang-Baxter relations and a subsequent average of rotation angles treating the linear change of the topological actions. Application of a magnetic field increases the magnetization, shifting the critical point to higher temperatures. At the temperature above the critical point, the magnetization keeps zero until a critical field at which it jumps rapidly as a first-order magnetization process. The partition function and the magnetization obtained are helpful for understanding the physical properties, in particular, the magnetization processes of the 2D magnetic materials.

Thrombosis in rotary blood pumps arises from complex flow conditions that remain difficult to translate into reliable and interpretable risk predictions using existing computational models. This limitation reflects an incomplete understanding of how specific flow features contribute to thrombus initiation and growth. This study introduces an interpretable machine learning framework for spatial thrombosis assessment based directly on computational fluid dynamics-derived flow features. A logistic regression (LR) model combined with a structured feature-selection pipeline is used to derive a compact and physically interpretable feature set, including nonlinear feature combinations. The framework is trained using spatial risk patterns from a validated, macro-scale thrombosis model for two representative scenarios. The model reproduces the labeled risk distributions and identifies distinct sets of flow features associated with increased thrombosis risk. When applied to a centrifugal pump, despite training on a single axial pump operating point, the model predicts plausible thrombosis-prone regions. These results show that interpretable machine learning can link local flow features to thrombosis risk while remaining computationally efficient and mechanistically transparent. The low computational cost enables rapid thrombogenicity screening without repeated or costly simulations. The proposed framework complements physics-based thrombosis modeling and provides a methodological basis for integrating interpretable machine learning into CFD-driven thrombosis analysis and device design workflows.

Interpreting complex datasets remains a major challenge for scientists, particularly due to high dimensionality and collinearity among variables. We introduce a novel application of Kolmogorov-Arnold Networks (KANs) to enhance interpretability and parsimony beyond what traditional correlation analyses offer. We present two interpretable, color-coded visualization tools: the Pairwise KAN Matrix (PKAN) and the Multivariate KAN Contribution Matrix (MKAN). PKAN characterizes nonlinear associations between pairs of variables, while MKAN serves as a nonlinear feature-ranking tool that quantifies the relative contributions of inputs in predicting a target variable. These tools support pre-processing (e.g., feature selection, redundancy analysis) and post-processing (e.g., model explanation, physical insights) in model development workflows. Through experimental comparisons, we demonstrate that PKAN and MKAN yield more robust and informative results than Pearson Correlation and Mutual Information. By capturing the strength and functional forms of relationships, these matrices facilitate the discovery of hidden physical patterns and promote domain-informed model development.

The deformation and rupture of a lipid vesicle due to the forced normal approach of an inclusion are essential for optimizing the design of magnetic giant unilamellar vesicles [magGUVs, Malik et al., Nanoscale 17, 13720 (2025)], with implications for active colloid-membrane interactions and cellular-scale chemical delivery. Here, we investigate vesicles propelled by a force-driven rigid inclusion and reveal a robust elastohydrodynamic mechanism: the inclusion outpaces the vesicle, sustaining a thinning film that drains symmetrically and self-similarly, largely independent of initial shape. For soft membranes and small inclusions, coupling drives a monotonic tension increase that can exceed the lysis tension. Evaluating the maximal tension over a delivery distance, we map an operating window in vesicle reduced area and size relative to the inclusion.

The shock-boundary layer interaction over the convex wall of a supersonic turbine vane is studied with a focus on extreme separation bubble events and the interplay between the bubble, streaks, and streamwise vortices. The present analysis is performed on a dataset computed by a LES of a supersonic turbine at Ma = 2.0 and Re = 395000. Building on findings related to near-wall streaks and vortices that drive the bubble breathing motion, here, we employ conditional analysis to study extreme bubble events. In events where the recirculation region is small, near-wall high-speed streaks penetrate the bubble and lead to higher tangential Reynolds stresses upstream of the incident shock compared to those with large recirculation. The streaks are accompanied by streamwise vortices that meander and induce intense fluid mixing, leading to higher wall-normal and spanwise Reynolds stresses, and consequently, higher turbulent kinetic energy. This turbulent activity also causes intense fluctuations in wall pressure and skin-friction coefficient along the separation region. In contrast, during events with bubble expansion, high-speed streaks are advected over the separation region, with streamwise vortices appearing only downstream of the shock, resulting in minimal fluid mixing inside the bubble. The analysis of mass flux along the bubble surface reveals that during its contraction phase, mass flux out of the bubble occurs predominantly upstream of the incident shock because of high-speed streaks dragged towards the wall by the streamwise vortices. In the expansion phase, pronounced mass flux into the bubble is observed downstream of the shock, near reattachment. This increased fluid injection is associated with the presence of streamwise vortices and low-momentum flow structures near reattachment, suggesting that fluid entrainment by vortices plays a key role in the mass flux into the bubble.

Complex systems are made up of many interacting components. Network science provides the tools to analyze and understand these interactions. Community detection is a key technique in network science for uncovering the structures that shape the behavior of these networks. This thesis introduces the Adaptive Cut, a novel method that improves clustering methods by employing multi-level cuts in hierarchical dendrograms. Overcoming the limitations of traditional single-level cuts-especially in unbalanced dendrograms-the Adaptive Cut provides a multi-level cut by optimizing a Markov chain Monte Carlo with simulated annealing. In addition, we propose the Balanceness score, an information-theoretic metric that quantifies dendrogram balance and predicts the benefits of multilevel cuts. Evaluations on over 200 real and synthetic networks show significant improvements in partition density and modularity. In the second part, our analysis shows that incorporating network geometry allows redefining administrative boundaries into non-contiguous regions that better reflect social and spatial dynamics. We also discuss the representation of hierarchical data in hyperbolic space through Poincare maps, which can represent tree-like structures in low dimension. In addition, we examine how geography constrains human mobility, an aspect often overlooked in scale-free characterizations of mobility. By incorporating geography via the pair distribution function from condensed matter physics, we separate geographic constraints from mobility choices. Analyzing datasets containing millions of individual movements, we identify a universal power law that spans five orders of magnitude, thereby bridging the divide between distance-based and opportunity-driven models of human mobility.

A variety of photon-mediated operations are critical to the realization of scalable quantum information processing platforms and their accurate characterization is essential for the identification of optimal regimes and their experimental realizations. Such light-matter interactions are often studied with a broad variety of analytical and computational methods that are constrained by approximation techniques or by computational scaling. Quantum processors present a new avenue to address these challenges. We consider the case of cavity mediated two-qubit gates. To investigate quantum state transfer between the qubits, we implement simulations with quantum circuits that are able to reliably track the dynamics of the system. Our quantum algorithm, compatible with NISQ (Noisy Intermediate Scale Quantum) era systems, allows us to map out the fidelity of the state transfer operation between qubits as a function of a broad range of system parameters including the respective detunings between the qubits and the cavity, the damping factor of the cavity, and the respective couplings between the qubits and the cavity. The algorithm provides a robust and intuitive solution, alongside a satisfactory agreement with analytical solutions or classical simulation algorithms in their respective regimes of validity. It allows us to identify under-explored regimes of optimal performance, relevant for heterogeneous quantum platforms, where the two-qubit gate can be rather effective between far-detuned qubits that are neither resonant with each other nor with the cavity. Besides its present application, the method introduced in the current paper can be efficiently used in otherwise untractable variations of the model and in various efforts to simulate and optimize photon-mediated two-qubit gates and other relevant operations in quantum information processing.

We consider nonlinear partial differential equations (PDEs) for advection-diffusion processes which are augmented by an auxiliary parameter $\delta$ such that $\delta=0$ corresponds to linear advection-diffusion. We derive potentially non-perturbative series expansions in $\delta$ that provide a process to obtain the solution of the nonlinear PDE through solving a hierarchical system of linear, forced PDEs with the forcing terms dependent on solutions at lower orders in the hierarchy. We rigorously detail our approach for a particular deformation that interpolates between linear advection-diffusion and the canonical Burgers' equation modeling nonlinear advection. In this case, we prove that the series has infinite radius of convergence for arbitrary integrable initial data, analyze the cases of a Dirac-delta initial condition (IC) (i.e., the fundamental solution) in an infinite domain and arbitrary IC in a periodic domain, and demonstrate the approach to turbulent behavior in a scenario with periodic forcing. We then treat models of nonlinear diffusion involving the $p$-Laplacian operator, including generalizations of the Poisson equation in $1$ and $2$ dimensions, and the heat equation in $1+1$ dimensions. We detail series expansions for two different deformations of these equations about their linear (ordinary Laplacian) counterparts, providing numerical evidence for the convergence of the series outside of a perturbative regime and demonstrating that the rate and radius of convergence are affected by choice of deformation. Our results provide a rigorous foundation for using series expansion techniques to study nonlinear advection-diffusion PDEs, opening new pathways for analysis and potential applications for quantum-assisted computational fluid dynamics.

We propose a randomized algorithm to compute the log-partition function of weakly interacting fermions with polynomial runtime in both the system size and precision. Although weakly interacting fermionic systems are considered tractable for many computational methods such as the diagrammatic quantum Monte Carlo, a mathematically rigorous proof of polynomial runtime has been lacking. In this work we first extend the proof techniques developed in previous works for proving the convergence of the cumulant expansion in periodic systems to the non-periodic case. A key equation used to analyze the sum of connected Feynman diagrams, which we call the tree-determinant expansion, reveals an underlying tree structure in the summation. This enables us to design a new randomized algorithm to compute the log-partition function through importance sampling augmented by belief propagation. This approach differs from the traditional method based on Markov chain Monte Carlo, whose efficiency is hard to guarantee, and enables us to obtain a algorithm with provable polynomial runtime.

Physics-Informed Neural Networks (PINNs) and more recently Physics-Informed Kolmogorov-Arnold Networks (PIKANs) have emerged as promising approaches for solving partial differential equations (PDEs) without reliance on extensive labeled data. In this work, we propose a novel multi-resolution Hybrid Wavelet-Fourier-Enhanced Physics-Informed Kolmogorov-Arnold Network (HWF-PIKAN) for solving advection problems based on collisionless Boltzmann equation (CBE) with both continuous and discontinuous initial conditions. To validate the effectiveness of the proposed model, we conduct systematic benchmarks on classical advection equations in one and two dimensions. These tests demonstrate the model's ability to accurately capture smooth and abrupt features. We then extend the application of HWF-PIKAN to the high-dimensional phase-space setting by solving the CBE in a continuous-velocity manner. This leverages the Hamiltonian concept of phase-space dynamics to model the statistical behavior of particles in a collisionless system, where advection governs the evolution of a probability distribution function or number density. Comparative analysis against Vanilla PINN, Vanilla PIKAN, as well as Fourier-enhanced and Wavelet-enhanced PIKAN variants, shows that the proposed hybrid model significantly improves solution accuracy and convergence speed. This study highlights the power of multi-resolution spectral feature embeddings in advancing physics-informed deep learning frameworks for complex kinetic equations in both space-time and phase-space.

The problem of representing laminated structures by an equivalent volume and determining the elastic constants of this equivalent volume from the layer properties is a fundamental issue in the analysis of composite and multilayered systems. In the literature, the most widely used approach for this purpose is the Voigt-type volume-weighted averaging method. Although this method is widely accepted in practice, the uniqueness (as a material characteristic) of the equivalent elastic constants, their independence from the applied loading, and the mathematical conditions under which they can be defined have not been made explicit. This issue is particularly unclear for structures with asymmetric stacking sequences and general anisotropic behavior, including triclinic cases. In this study, a theoretical framework referred to as the Zor model is presented, and it is shown that the elastic constants of the equivalent volume can be obtained only under the action of all in-plane force components (Fx + Fy + Fxy), directly from static equilibrium, linear elasticity, and perfect bonding conditions between the layers. No additional assumptions are introduced in the formulation; reciprocity is not assumed at the system level, but instead emerges naturally as a result of the solution process. The resulting equivalent elastic constants are independent of the applied loading and represent the intrinsic mechanical characteristics of the laminated structure.

Detecting broken time-reversibility at micro- and nanoscale is often difficult when experiments offer limited state resolution. We introduce a lumping method that builds an effective semi-Markov model able to reproduce exactly the full entropy-production statistics of the microscopic dynamics. The mean entropy production stays accurate even when hidden current-carrying cycles are merged, though higher-order information can be unavoidably lost. In these cases, we capture violations of fluctuation theorems consistent with experiments, opening a path to novel inference strategies out of equilibrium.

The energy distribution is a fundamental property of magnetohydrodynamic (MHD) turbulence. In strongly magnetized turbulence energy imbalances arise and are quantified by the residual energy: $E_r~=~(E_{kin}~ - ~E_{mag})$; $E_{kin}$ and $E_{mag}$ stand for the volume-averaged kinetic and magnetic energy, respectively. We explore the properties of $E_r$ in weakly compressible MHD turbulence in the presence of an initially strong (guide) magnetic field, investigating how the driving mechanism and the magnetic field strength affect the cascade of $E_r$. We run a suite of direct numerical simulations with the PENCIL code. The sonic Mach number is approximately equal to 0.1 in all simulations, whereas the plasma beta varies. We drive turbulence by either injecting velocity or magnetic fluctuations at large scales and study the power spectra of kinetic, magnetic, density, and $E_r$. Magnetically driven simulations show locally imbalanced Alfvénic fluctuations and a $\propto k^{-3/2}$ cascade, consistent with the dynamic alignment theory. In the inertial range, $E_r \approx$ 0. Kinetically driven simulations give rise to a $\propto k^{-1}$ scaling, consistent with weakly interacting modes that preserve a high level of coherence throughout the inertial range. Residual energy is positive at all scales of the inertial range. The spectral slope of the $E_r$ cascade steepens systematically with increasing magnetization, varying from approximately -1 at $\beta = 0.3$ to between -2.0 and -5/3 at $\beta = 4.0$. The energy partition in weakly compressible turbulence is strongly influenced by the forcing mechanism, even when the global sonic and Alfvénic Mach numbers are comparable across simulations.

A data-driven methodology is proposed to model the distribution of multivariate stochastic trajectories from an observed sample. As a first step, each trajectory in the sample is reduced to a vector of features by means of Functional Principal Component Analysis. Next, the joint distribution of features is modeled using (i) a non-parametric vine copula approach for the bulk of the distribution, and (ii) the conditional modeling framework of Heffernan and Tawn (2004) for the multivariate tail. The method is applied to the modeling of water waves. The dataset used is the DeRisk database, which consists of numerical simulations of water waves. The analysis is restricted to the portion of the wave period between the free-surface zero-upcrossing and the wave crest. The kinematic variables considered are the free-surface slope, the normal component of the fluid velocity at the free surface, and the vertical Lagrangian acceleration of the fluid at the free surface. The stochastic trajectories of these three variables are modeled jointly. The vertical Lagrangian acceleration of the fluid is employed to enforce a wave-breaking filter in the stochastic model. The number of hyperparameters in the stochastic framework is reduced to three, and a stepwise calibration strategy is proposed for their adjustment. The capabilities of the model are illustrated by predicting the distributions of selected response variables and by generating synthetic trajectories.

We investigate the perturbations induced by a periodic bathymetry on traveling Boussinesq solitons in a two-dimensional configuration. We present two perturbation approaches to solve the nonlinear, dispersive and non-autonomous differential equations of the model and compare the solutions with numerical simulations of the original system of equations. In the approximation for small periodic corrugations we built the solutions as modulated traveling waves using Fourier series. The coefficients of the series are solved using the Green function method and the pathordered exponential method. At the second order in the relative height of bed corrugations, we obtain the perturbation as the fourth-order linear dispersive waves generated by the modulated traveling soliton in its wake. In the second approach, we rewrite the Boussinesq system into a perturbed Korteweg-de Vries (KdV) nonlinear equation, and obtain the corresponding perturbed solitons. These analytic solutions are compared with the results of numerical simulations, for various parameters that characterize the effects of nonlinearity, dispersion, and bottom bathymetry. We also discuss the stability of the perturbed solitons in time. The perturbation approaches developed in this study are valid for any type of periodic bathymetries, and the method can readily be extended to non-periodic ones.

Quantum Krylov subspace methods can extract ground and excited states by diagonalizing the Hamiltonian in a compact variational space. In practice, these spaces are almost always generated by real or imaginary time evolution, forcing a timestep trade-off between dynamical accuracy and basis collapse and often producing ill-conditioned overlap matrices that stall convergence. Here we introduce Quantum Krylov using Unitary Decomposition (QKUD), a time-evolution-free construction that maps Hamiltonian powers to implementable unitaries via the Hermitian transform $\sin(\epsilon H)/\epsilon$. QKUD reduces to the exact Hamiltonian-power Krylov recursion as $\epsilon\rightarrow0$, while finite $\epsilon$ provides a controllable deformation that tunes subspace geometry and improves conditioning. Across molecular active-space benchmarks and a frustrated 2D J1-J2 Heisenberg model, QKUD reproduces exact-Krylov convergence in well-conditioned regimes and systematically restores variational improvement when both exact Krylov and time-evolution Krylov stagnate. These results identify overlap conditioning, instead of time-evolution fidelity, is the key resource for robust quantum Krylov simulation and provide a resilient way forward for accurate quantum simulation of challenging quantum many-body problems.

Volatile organic compounds (VOCs) are valuable health indicators, with synthetic breath biomarkers offering rapid and disease specific diagnostics. However, their <100 ppb level exhalation requires mass spectrometry, limiting clinical integration. Surface-enhanced Raman spectroscopy (SERS) offers a portable, cost-effective alternative. Yet, detecting synthetic breath biomarkers, with inherently low Raman cross-sections, at <100 ppb remains challenging. We demonstrate SERS detection down to clinically relevant 10 ppb via water-mediated trapping in hydroxylated nanoporous silica-coated plasmonic nanogaps, using pentafluoropropylamine (PFP) as a representative synthetic breath biomarker. Uniform nanogaps, with >1000 times electric field enhancement, were generated between a gold film and gold-silica core-shell nanoparticle assemblies using electric field-driven evaporation. Oxygen plasma treatment hydroxylated the silica, enabling water-mediated hydrogen bonding that strengthened PFP adsorption, confirmed by density functional theory. This mechanism improved SERS sensitivity by 10000 fold, enabling ppb level PFP detection in mouse bronchial fluid and establishing a VOC capturing SERS platform for breath-based diagnostics.

Surface ablation measurements of glaciers are critical for understanding mass change over time. Mass-balance stakes are commonly used for localized measurements, with the exposed length typically measured manually at infrequent intervals. This paper presents the design and validation of new instrumentation that automates mass-balance stake readings, thus enabling continuous measurements with high temporal resolution. The instrumentation comprises readout electronics that are mounted on mass-balance stakes to measure wind-induced vibrations. The stake vibrational frequency depends sensitively on the exposed length, and changes in the measured frequency therefore probe glacier surface melt and accumulation. Initial instrumentation field tests conducted at Color Lake on Umingmat Nunaat (Axel Heiberg Island), Nunavut, demonstrate centimeter-level precision on length measurements. The instrumentation can be attached to existing mass-balance stakes and is low-cost (~ $50 USD) in comparison to many other systems that perform automated surface ablation measurements. The accessibility of this instrumentation opens new possibilities for localized, high temporal resolution measurements of glacier surface activity at any locations where mass balance stakes are deployed.

Water behaves very differently at surfaces and under extreme confinement, but the boundary between these two regimes has remained unclear. Despite evidence that interfacial effects persist under sub-nanometre confinement, the molecular-scale behaviour and its evolution with slit width remain unclear. Here, we use machine-learning molecular dynamics with first-principles accuracy to probe water at graphene surfaces across slit widths ranging from the open-interface limit to angstrom-scale confinement. We find that water undergoes a sharp structural transition: when three or more water layers fit between the walls, the structure of the graphene-water interface is effectively indistinguishable from that in an open system, with density layering, hydrogen bonding, and orientational ordering retaining interfacial character. Below this threshold, however, angstrom-scale confinement strongly reorganises the liquid, producing enhanced ordering, a restructured hydrogen-bond network, and modified orientational motifs. These results establish a molecular-level picture that clearly separates interfacial behaviour from genuine nanoconfinement and provide guidance for predicting and controlling the structure of water in nanoscale solid-liquid environments.

Cities are increasingly challenged by extreme heat events, which pose serious risks to public health and urban livability. Micromobility users, whose numbers have increased rapidly in recent years, are particularly vulnerable to outdoor heat exposure. Yet, their exposure patterns and the effectiveness of mitigation measures remain poorly understood. Here, we couple a high-resolution urban microclimate model (WRF–BEP–SOLWEIG) with 4.76 million Citi Bike trips in New York City to quantify cyclists' thermal exposure during the June 2024 heatwave and to evaluate targeted cooling strategies. Results show that a small fraction of the street network concentrates the majority of rider heat exposure, and that localized interventions along these segments yield the greatest benefits. Targeted tree planting along just 1.5% of the city's street network reduces total heat-exposed kilometers ridden by 19%, equivalent to a thermal stress reduction of about 4{\deg}C, with its impact maximized during midday hours. In contrast, randomized citywide tree planting produces diffuse, resource-intensive cooling, highlighting the superior efficiency of spatially prioritized interventions. Baseline results further indicate that daytime heat stress is higher in lower-income neighborhoods, adding an important social dimension of urban heat exposure. Together, these findings provide a quantitative basis for designing heat-resilient and equitable cycling networks in a warming climate.

In this work, we present a proof-of-concept investigation of non-equilibrium chemical reaction dynamics at a molecule-electrode interface, driven out of equilibrium by an applied votage bias and mediated by a confined, enhanced vacuum electromagnetic field inside an optical cavity. The coupled electron-vibration-photon system, together with the electrodes and a dissipative environment, is described within an open quantum system framework and solved using a numerically exact quantum dynamical approach. The reaction coordinate is modeled with a Morse potential, enabling explicit treatment of molecular anharmonicity and bond-breaking behavior. By varying the cavity frequency across the infrared regime to cover typical nuclear vibrational energies, we observe multiple resonant rate suppression features that emerge whenever the cavity mode is brought into resonance with a dipole-allowed vibrational transition along the anharmonic ladder up to the dissociation threshold. These findings open the door to extending polaritonic chemistry into genuinely nonequilibrium scenarios relevant to molecule-electrode interfaces. Moreover, building on these results, we further propose a multi-mode vibrational strong coupling strategy in which several cavity modes are individually matched to distinct vibrational transitions. This engineered multi-resonant cavity induces a stepwise vibrational ladder descending process that efficiently drains vibrational excited energy. The resulting cavity-assisted cooling suggests a potential route toward mitigating voltage-induced bond rupture and the long-standing instability issues of molecular junctions operating under high bias.

We propose an interaction-free ghost imaging scheme based on a thermal light source. By utilizing the quantum Zeno-like effect, our approach significantly reduces the light dose absorbed by the sample, thereby effectively preventing sample damage induced by light-matter interactions. Combined with the elimination of entangled photon sources and single-photon detectors, our approach enables significantly more photons to be utilized for image reconstruction, thereby markedly enhancing image quality compared to conventional ghost imaging. We further demonstrate active suppression of background noise via controllable photon loss. Our work offers a practical and cost-effective route to non-destructive, high-quality imaging for light-sensitive samples in fields such as life sciences.

This article investigates the sensemaking demonstrated in 6th grade students' written responses to a single question free-response final exam for physics asking them to recount everything they learned over the course of the academic year. International exams such as the PISA and TIMSS show that students continue to have persistent difficulties with their comprehension, and appreciation, of science. Sensemaking and development of a deeper understanding of concepts is of fundamental importance when teaching science, but little progress has been made over the ensuing decades. The research questions are: What sensemaking of physics is communicated through a single question free-response exam? In particular, what topics are preferred by students and how are the semiotic resources of written and visual representations utilized to express sensemaking? With specific attention given to the topic of sound we observe two levels of comprehension, defined as Basic and Advanced, related to Bloom's Taxonomy, and we see evidence that previously low-performing students are capable of sensemaking at the Advanced level, thereby lending support to recent research calling for an increase in the level of complexity employed in primary and lower secondary science curricula. In addition to the sensemaking analysis, we discuss how these results are facilitated by the single-question free-response format which has been completely unexplored in the research literature and has the potential to be a valuable asset for research on or using assessment, as well as for teacher self-assessment.