Hyperspectral anomaly detection faces fundamental challenges in balancing spatial context, statistical rigor, and interpretability without ground truth supervision. This article presents spatially informed theoretical model (STIM), a novel unsupervised framework that addresses these challenges through a principled two-stage reference computation architecture. STIM systematically aggregates local spectral statistics into globally informed spatial references, enabling the derivation of three complementary features: energy (photometric deviation), entropy (local spectral coherence), and divergence (global statistical rarity). We establish theoretical foundations including noise robustness, Lipschitz continuity, and information-theoretic optimality with convergence guarantees. Comprehensive validation on five Airborne Visible/Infrared Imaging Spectrometer—Next Generation benchmark datasets demonstrates STIM's substantial superiority over traditional statistical and deep learning methods, achieving 14.6× to 585× improvements in mean anomaly scores with a reliability index of 0.933. Feature dynamics analysis confirms multimodal orthogonality and consistent interpretability across diverse hyperspectral environments. STIM enables robust, interpretable, and generalizable anomaly detection for operational hyperspectral imaging without requiring labeled supervision or scene-specific calibration, advancing the state-of-the-art in unsupervised hyperspectral analysis.
Dulce Sonia Oreano Hernández, Manuel Mendoza, Juan Alfredo Hernández Guerrero
et al.
Understanding the hazards associated with water requires the use of a hydrogeomorphic approach, which considers both hydrologic and geomorphic aspects. In recent decades, research on flash floods has increasingly adopted this approach. However, the overall number of studies and their trends have not been thoroughly documented. Our study aims to analyze the evolution of scientific publications examining flash floods from a hydrogeomorphic perspective on a global scale to understand the trends and research gaps based on a bibliometric analysis. A comprehensive search for relevant publications was conducted in the Web of Science and Scopus databases, covering the period from 1973 to 2024. The resulting data were processed using R software, whereas the spatial distribution network of the publications was analyzed using VOSviewer software. Our analysis identified 212 articles focusing on flash floods as a hydrogeomorphic process. The number of publications has increased since 2012, peaking in 2023 with 21 new articles. Twenty-eight percent of the publications originated from the United States, Spain, and Italy, whereas the most extensive global collaboration network involved researchers from France, the United States, and Canada. A total of 87% of the publications on temperate zones, whereas 13% addressed intertropical environments, where hydrogeomorphic hazards can be particularly devastating. Our study underscores the importance of future research on flash floods hazards in intertropical zones, highlighting the need to incorporate hydrogeomorphic characteristics and processes into studies of flash floods and related phenomena such as floods, debris flows, landslides, and erosion.
A. S. Joglekar, A. G. R. Thomas, A. L. Milder
et al.
Differentiable programming, enabled by automatic differentiation (AD), provides a robust framework for gradient-based optimization in computational plasma physics. While optimization is often only used towards design, we demonstrate that it can also be used for discovery and bridging the gap towards multi-scale modeling. We discuss four applications: (1) discovering novel nonlinear plasma phenomena, including a previously unknown superadditive wavepacket interaction regime, by optimizing differentiable kinetic simulations; (2) learning hidden variables that capture spatiotemporally non-local kinetic effects in fluid simulations, enabling hydrodynamic models to reproduce large Knudsen number physics typically requiring kinetic solvers; (3) accelerating Thomson scattering analysis by over $140\times$ while enabling extraction of velocity distribution functions with $\mathcal{O}(10^3)$ parameters; and (4) inverse design of spatiotemporal laser pulses that achieve target far-field behavior where full space-time coupling improves performance by $15\times$ over spatial or temporal optimization alone. These examples illustrate that differentiable programming not only accelerates existing design and inference workflows but enables qualitatively new capabilities, from algorithmic physics discovery to high-dimensional inference and design previously considered intractable.
We compute the dominant QED correction to the neutrino-electron interaction rate in the vicinity of neutrino decoupling in the early universe, and estimate its impact on the effective number of neutrino species N eff in cosmic microwave background anisotropy observations. We find that the correction to the interaction rate is at the sub-percent level, consistent with a recent estimate by Jackson and Laine. Relative to that work we include the electron mass in our computations, but restrict our analysis to the enhanced t-channel contributions. The fractional change in N eff SM due to the rate correction is of order 10-5 or below, i.e., about a factor of 30 smaller than that recently claimed by Cielo et al., and below the nominal computational uncertainties of the current benchmark value of N eff SM = 3.0440 ± 0.0002. We therefore conclude that aforementioned number remains to be the state-of-the-art benchmark for N eff SM in the standard model of particle physics.
Promptly after high-resolution experiments harbinger the field of precision cosmology low- and high-redshift observations abruptly gave rise to a tension in the measurement of the present-day expansion rate of the Universe ($H_0$) and the clustering of matter ($S_8$). The statistically significant discrepancies between the locally measured values of $H_0$ and $S_8$ and the ones inferred from observations of the cosmic microwave background assuming the canonical $\Lambda$ cold dark matter (CDM) cosmological model have become a new cornerstone of theoretical physics. $\Lambda_s$CDM is one of the many beyond Standard Model setups that have been proposed to simultaneously resolve the cosmological tensions. This setup relies on an empirical conjecture, which postulates that $\Lambda$ switched sign (from negative to positive) at a critical redshift $z_c \sim 2$. We reexamine a stringy model that can describe the transition in the vacuum energy hypothesized in $\Lambda_s$CDM. The model makes use of the Casimir forces driven by fields inhabiting the incredible bulk of the dark dimension scenario. Unlike the $\Lambda_s$CDM setup the model deviates from $\Lambda$CDM in the early universe due to the existence of relativistic neutrino-like species. Using the Boltzmann solver CLASS in combination with MontePython we confront predictions of the stringy model to experimental data (from the Planck mission, Pantheon+ supernova type Ia, BAO, and KiDS-1000). We show that the string-inspired model provides a satisfactory fit to the data and can resolve the cosmological tensions.
Dark Matter Particle Explorer (short for DAMPE) is one of the space missions innovated by the Chinese Academy of Sciences. Its scientific goals are to discover dark matter particles, study the characteristics of dark matter particles and their laws of space deployment, detect gamma rays, and find out the source’s cosmic rays. The attainment of such goals will lead to great leaps in the frontiers of modern physics and astronomy. Based on precision orbit elements of DAMPE newly available, some main orbit elements were simulated. Further, with the help of SOCRATES Plus software, the close approach of DAMPE to the space objects was computed. The objects are singled out that may endanger the wholesomeness of the DAMPE.
The origin of ultra-high-energy cosmic rays (UHECRs) remains a major open question in astro-physics. Observational data suggest that starburst galaxies and active galactic nuclei (AGN) are the most promising sources. However, accelerating particles to energies above 1 EeV in these environments is complex due to the demanding requirements on energy, density, and metallicity imposed by observations. In this work, we explore the theoretical challenge of explaining the presence of intermediate and heavy nuclei within the context of AGNs. The interaction of AGN jets with the winds of embedded stars leads to turbulent mixing and jet mass loading. The winds of Wolf-Rayet stars are rich in nuclei of Carbon-Nitrogen-Oxygen, and in this contribution we focus on the role of Wolf-Rayet stars in enhancing the metallicity of AGN jets. We estimate the flux of UHECRs from a Centaurus A-like radiogalaxy and find it to be comparable to the fluxes observed by the Pierre Auger Observatory under certain assumptions. We conclude that the proposed scenario is potentially one that could shed light on the sources of UHECRs.
Data from multiple experiments suggest that the current interaction models used in Monte Carlo simulations do not correctly reproduce the hadronic interactions in air showers produced by ultra-high-energy cosmic rays (UHECR). We have created a large library of UHECR simulations where the interactions at the highest energies are slightly modified in various ways - but always within the constraints of the accelerator data, without any abrupt changes with energy and without assuming any specific mechanism or dramatically new physics at the ultra-high energies. Recent results of the Pierre Auger Observatory indicate a need for a change in the prediction of the models for both the muon content at ground and the depth of the maximum of longitudinal development of the shower. In our parameter space, we find combinations of modifications that are in agreement with this analysis, however a consistent description of UHECR showers remains elusive. Our library however provides a realistic representation of the freedom in the modeling of the hadronic interactions and offers an opportunity to quantify uncertainties of various predictions. This can be particularly valuable for the design of future observatories where hadronic models are often used as input for the prediction of the performance. We demonstrate this powerful capability on several selected examples.
The existence of axion-like particles (ALPs) can be probed from their signatures in the Cosmic Microwave Background (CMB) due to the photon-ALP resonant conversion over the mass range of ALPs that matches with the effective mass of photons in the plasma in the astrophysical systems. Such a conversion can also occur in the Milky Way halo and disk and can cause a unique spatial and spectral distortion. The signal is highly non-Gaussian and cannot be measured precisely by the usual power-spectrum approach. We devise a new technique to search for this signal from the upcoming full-sky CMB experiment LiteBIRD using its multi-frequency band using a template-based spatial profile of the ALP distortion signal. This technique captures the large-scale non-Gaussian aspects of the ALP distortion signal in terms of a spatial template and makes it possible to search for any non-zero ALP signal. We show that the inference of the ALP coupling using the template-based technique from LiteBIRD can provide constraints on the coupling constant approximately gaγ < 6.5 × 10-12 GeV-1 for ALP masses below 10-14 eV at 95% confidence interval which is an order of magnitude better than the current bounds from CERN Axion Solar Telescope (CAST) at gaγ < 6.6 × 10-11 GeV-1, This shows the capability of future multi-band CMB experiment LiteBIRD in opening the discovery space towards physics beyond the standard model.
Brown and Korringa (1975) extended Gassmann’s framework to account for multi-mineral porous media, providing a theoretical basis for calculating elastic moduli in heterogeneous rocks. Despite its importance in geophysics and rock physics, a comprehensive 3D numerical validation of these equations for generic pore geometries and multi-mineral compositions has remained lacking. In this study, we address this gap by presenting a detailed numerical examination of the Brown and Korringa and Gassmann equations for a particular 3D model of multi-mineral porous media. Using a high-resolution 3D finite-element approach, we solve the coupled equations of linear momentum conservation and stressstrain relationships in the solid phases, combined with the quasi-static, linearized NavierStokes equations in the fluid phase. The analysis shows excellent agreement between the Brown and Korringa predictions and the numerical results under quasi-static conditions. We examine the predictions of Gassmanns equations using a solid bulk modulus that accounts for the multi-mineral composition of the matrix. While Gassmanns predictions exhibit a small but systematic divergence from the numerical results, the differences are negligible for practical applications. These results support the current practice of using Gassmanns equation for multi-mineral rocks, despite its formal derivation assuming a single isotropic mineral phase. However, the use of the VoigtReussHill (VRH) estimate for the solid bulk modulus is not supported by our analysis.
Abstract The El Niño‐Southern Oscillation (ENSO) influences ocean wave activity across the Pacific, but its effects on island shores are modulated by local weather and selective sheltering of multi‐modal seas. Utilizing 41 years of high‐resolution wave hindcasts, we decipher the season‐ and locality‐dependent connections between ENSO and wave patterns around the Hawaiian Islands. The north and west‐facing shores, exposed to energetic northwest swells during boreal winters, experience the most pronounced ENSO‐related variability, with increased high‐surf activity during El Niño years. While the year‐round trade wind waves exhibit moderate correlation with ENSO, the basin‐wide climate influence is masked by locally accelerated trade winds in channels and around large headlands. The remarkable global‐to‐local pathway through the high‐resolution hindcast enables development of an ENSO‐based semi‐empirical wave model to statistically describe and predict severe wave conditions on vulnerable shores with potential application in coastal risk management and hazard mitigation for Pacific Islands and beyond.
Abstract Knowledge of the original dimension of Greater India in the Jurassic Period is critical for reconstructing the paleogeographic framework of Gondwana. However, two end‐member reconstructions for Greater India (≤1,000 or ≥2,000 km) exist in the Jurassic. Here, we conduct comprehensive petrographic, rock magnetic, and paleomagnetic studies on Lower to Middle Jurassic Lure Formation marlstones from the eastern Tethyan Himalaya. The new paleomagnetic data fulfill stringent quality criteria and provide a reliable Early Jurassic paleopole of 16.6°N/308.8°E, A95 = 1.2°, n = 396 (specimens), which constrains the Tethyan Himalaya to a paleolatitude of 31.8° ± 1.2°S at ca. 177.3 Ma. The overlap of this paleolatitude with the expected paleolatitude of India is consistent with a small Greater India (∼900 km) in the Jurassic. The improved estimate of the dimension of Greater India contributes to an updated reconstruction of the paleogeography of Gondwana during the Early Jurassic.
Priyanka Pandit, Prathibha Chandrashekhar, Sparsh Sharma
et al.
Abstract Seismic studies in cold subduction zones indicate several discontinuity structures near the 660‐km boundary. Studies indicate that the akimotoite to bridgmanite transition may play a significant role in unraveling the complexity of this region. In this study, we used first‐principles methods to explore the stability field of iron‐rich analogs of akimotoite and bridgmanite (Mg1−xFex2+SiO3 under high‐pressure‐temperature conditions. The Fe2+ inclusion significantly reduces the phase transition pressure. Overall, our calculated phase boundary and thermoelastic properties compare well with the available results from previous studies. The onset transition pressure and the width of the two‐phase field exhibit a clear dependence on iron concentration, with the width of the two‐phase field increasing as iron concentration increases. Our results indicate that the relatively high Fe2+ (∼x = 0.5) found in natural Fe analogs of akimotoite and bridgmanite would not be possible under mantle transition conditions. However, Fe2+ incorporation relevant for mantle composition (<10 mol.% FeO) may explain the slab stagnation above 660 km depth as well as seismically observed trends of velocity perturbations in the slabs of the northwest Pacific region around ∼500–600 km depth.
Abstract Given the pressures on water resources caused by global climate change and human activities, the assessment and management of groundwater resources in mountainous region have become increasingly important. The central mountainous region of Taiwan, as one of the significant sources of groundwater recharge, plays a critical role in overall water resource management due to its groundwater storage capacity and recharge capability. Addressing the challenges of limited survey and observational data in mountainous groundwater assessments, this study uses the lumped parameter groundwater model AquiMod to analyze long-term groundwater level changes at 23 monitoring stations in mountainous areas of central Taiwan. This study is based on long-term groundwater level monitoring data (2010–2021) analyzing the relationship between groundwater levels and precipitation, and performs model calibration and prediction. The results indicate a strong correlation between groundwater levels in mountainous areas and precipitation. While the model predictions were satisfactory for most monitoring stations, obtaining Nash Sutcliffe efficiency scores of between 0.5 and 0.9 at 14 of the 23 monitoring stations. However, poorer performance at several stations reflects limitations arising from data gaps, complex local geology, and the inability of the lumped model to represent lateral recharge or anthropogenic influences. Model sensitivity analysis further highlights the critical role of unsaturated zone parameters, such as rooting depth, soil storage and upper-layer saturated hydraulic conductivity, in shaping groundwater responses. In summary, the lumped parameter groundwater model has proven practical for evaluating groundwater in Taiwan’s mountainous regions and can serve as a reference for the sustainable management of future water resources.
In this review we motivate ultrahigh energy neutrino searches and their connection to ultrahigh energy cosmic rays. We give an overview of neutrino production mechanisms and their potential sources. Several model-independent benchmarks of the ultrahigh energy neutrino flux are discussed. Finally, a brief discussion of approaches for model-dependent neutrino flux predictions are given, highlighting a few examples from the literature.
The rise of foundation models -- large, pretrained machine learning models that can be finetuned to a variety of tasks -- has revolutionized the fields of natural language processing and computer vision. In high-energy physics, the question of whether these models can be implemented directly in physics research, or even built from scratch, tailored for particle physics data, has generated an increasing amount of attention. This review, which is the first on the topic of foundation models in high-energy physics, summarizes and discusses the research that has been published in the field so far.