Fine-scale Features Research Articles

Local differences in network organization in auditory and parietal cortex, revealed with single neuron activation.

The basic structure of local circuits is highly conserved across the cortex, yet the spatial and temporal properties of population activity differ fundamentally in sensory-level and association-level areas. In sensory cortex, population activity has a shorter timescale and decays sharply over distance, supporting a population code for the fine-scale features of sensory stimuli. In association cortex, population activity has a longer timescale and spreads over wider distances, a code that is suited to holding information in memory and driving behavior. We tested whether these differences in activity dynamics could be explained by differences in network structure. We targeted photostimulations to single excitatory neurons of layer 2/3, while monitoring surrounding population activity using two-photon calcium imaging. Experiments were performed in auditory (AC) and posterior parietal cortex (PPC) within the same mice of both sexes, which also expressed a red fluorophore in somatostatin-expressing interneurons (SOM). In both cortical regions, photostimulations resulted in a spatially restricted zone of positive influence on neurons closely neighboring the targeted neuron, and a more spatially diffuse zone (akin to a network-level 'suppressive surround') of negative influence affecting more distant neurons. However, the relative spatial extents of positive and negative influence were different in AC and PPC. In PPC, the central zone of positive influence was wider, but the negative suppressive surround was more narrow than in AC, which could account for the larger-scale network dynamics in PPC. The more narrow central positive influence zone and wider suppressive surround in AC could serve to sharpen sensory representations.Significance Statement The basic structure of local circuits is highly conserved across the cortex, and yet local processing goals vary across the sensorimotor hierarchy, from sensory perception to the control of behavior. It has been unclear whether these differences in function require different network organization. To probe the spatial structure of networks in sensory and association-level cortex, we photostimulated single excitatory neurons and measured the effect on local population activity in mice. Stimulations triggered a centered, positive activity change in neighboring neurons, and a surrounding, suppressive change in more distant neurons. The relative sizes of the center and surround differed across areas, suggesting that network structure is tailored for sharper, more restricted activity in sensory cortex, and more dense network activity in association cortex.

Seabed Depth Prediction Using Multi-Scale Gravity Anomalies and Fully Connected Deep Neural Networks: A Novel Approach Applied to the South China Sea

Accurate seabed topography is crucial for marine research, resource exploration, and engineering applications. While deep learning techniques have been widely applied in seabed inversion, existing methods often overlook the multi-scale influence of gravity anomalies, particularly the critical role of short-wavelength gravity anomalies in resolving fine-scale bathymetric features. In this study, we propose a novel Fully Connected Deep Neural Network (FCDNN) approach that systematically integrates long-wavelength, short-wavelength, and residual gravity anomaly components for seabed topography inversion. Using multi-satellite altimetry-derived gravity anomaly data (SIO V32.1) and shipborne bathymetric data (NCEI), we constructed a high-resolution (1′ × 1′) seabed topography model for the South China Sea (108°E–121°E, 6°N–23°N), termed FCD_Depth_SCS. The workflow included multi-scale decomposition of gravity anomalies, linear regression-based residual calculation, and FCDNN-based nonlinear training to capture the complex relationships between gravity anomalies and water depth. The FCD_Depth_SCS model achieved a difference standard deviation (STD) of 44.755 m and a mean absolute percentage error (MAPE) of 2.903% when validated against 160,476 shipborne control points. This performance significantly outperformed existing models, including GEBCO_2024, SIOv25.1, DTU18, and GGM_Depth (derived from the Gravity–Geologic Method), whose STDs were 82.234 m, 108.241 m, 186.967 m, and 58.874 m, respectively. Notably, the inclusion of short-wavelength gravity anomalies enabled the model to capture fine-scale bathymetric variations, particularly in open-sea regions. However, challenges remain near coastlines and complex terrains, highlighting the need for further model partitioning to address localized nonlinearity. This study highlights the benefits of integrating multi-scale gravity anomaly data with a fully connected deep neural network. Employing this innovative and robust approach enables high-resolution inversion of seabed topography with enhanced precision. The proposed method provides significant advancements in accuracy and resolution, contributing valuable insights for marine environmental research, resource management, and oceanographic studies.

Leveraging overfitting for good - Two-Step Deep Image Prior Model For Seismic Denoising

Accurate separation of signal and noise constitutes a fundamental prerequisite for achieving high-resolution seismic imaging. A notable recent advancement in this domain is the Deep Image Prior (DIP), an unsupervised deep learning method leveraging deep neural networks (DNNs). The success of this approach lies in the adoption of autoencoders that enables the adaptive extraction of high-fidelity data features. However, establishing an optimal balance between noise suppression and signal preservation remains a non-trivial challenge for DIP-based seismic denoising methods, which is affected by the potential issue of overfitting. This arises from the inappropriate selection of a network architecture and the corresponding hyperparameters, especially the number of training epochs, which strongly influence the learning capacity and feature extraction capabilities of the model. In response to this challenge, we introduce Two-Step DIP (TSDIP), a novel denoising method that exploits overfitting to enhance seismic data quality. In the initial stage, the proposed DNNs are intentionally trained to overfit by effectively attenuating high-frequency noise from the input data. Subsequently, the proposed DNNs are employed iteratively to suppress any residual noise in the newly processed data without damaging useful signal. The overfitting in the first step helps precondition the data to be at a lower noise level while preserving as much as fine-scale features in the signal. To employ an optimal balance, we carefully determine an ideal number of epochs, which is consistently applied in both denoising steps. To assess the effectiveness of the TSDIP method, we present the test results derived from 3D synthetic and field seismic datasets. Our analysis indicates that TSDIP effectively reduces strong noise while preserving key seismic details through the use of overfitting.

Constructions of multi-scale 3D digital rocks by associated image segmentation method

Digital rocks constructed from micro-CT image at a single-resolution face limitations in accurately identifying the entire pore space and mineral components of tight sandstones, due to their high content of nanoscale pores and clay. Consequently, the porosity values derived from such digital rocks are significantly lower compared to those obtained through laboratory measurements, resulting in discrepancies between the measured and calculated petrophysical properties. This study introduces a multi-scale digital rock modeling method by integrating three-dimensional micro-CT images acquired at two distinct resolutions and two-dimensional SEM images. Plunger-shaped core samples and their corresponding sub-samples were scanned at resolutions of 13.99 μm/voxel and 2.99 μm/voxel, respectively. The scale-invariant feature transform (SIFT) image registration technique was employed to accurately align the two sets of grayscale CT images. Correlation curves between the grayscale value in low-resolution CT images and various mineral contents were established based on the aligned regions, and utilized to construct multi-scale digital rock models. Intragranular pores, unresolvable by the micro-CT images, were identified using SEM imaging, enabling the incorporation of fine-scale features into the models. The resulting multi-scale digital rock models exhibited bulk porosity values that closely matched laboratory helium porosity measurements. Additionally, the elastic moduli calculated by the differential effective medium (DEM) model and the finite element method (FEM) demonstrated good correspondence with experimental results. These results validate the proposed multi-scale digital rock modeling method as an effective approach for accurately characterizing the porosity and mineral components of tight sandstone reservoirs.

Sweating the small stuff: microclimatic exposure and species habitat associations inform climate vulnerability in a grassland songbird community.

Assessment of species' vulnerability to climate change has been limited by mismatch between coarse macroclimate data and the fine scales at which species select habitat. Habitat mediates climate conditions, and fine-scale habitat features may permit species to exploit favourable microclimates, but habitat preferences can also constrain their ability to do so. We leveraged fine-resolution models of near-surface temperature and humidity in grasslands to understand how microclimates affect climatic exposure and demographics in a grassland bird community. We asked: (i) Do species select favourable nest-site microclimates? (ii) Do habitat preferences limit the ability of species to access microclimates? (iii) What are the demographic consequences of microclimatic exposure? We found limited evidence that grassland birds select cooler microclimates, which may buffer eggs and nestlings from extreme heat. Instead, many species appeared constrained by vegetation preferences. While facultative generalists displayed flexibility to use denser vegetation that provided buffering from high temperatures (>39°C), most obligate species nested in more exposed microclimates. Nesting success in facultative species was not well explained by microclimate, but success in specialized grassland obligates declined with increasing microclimate temperatures. These findings illustrate how microclimate and habitat use can interact to influence the potential vulnerability of species to future climate change.

Atmospheric fronts detection and cold/warm type determination based on wind field retrieval from SAR imagery

ABSTRACT Atmospheric fronts are one of the important atmospheric phenomena in the marine atmospheric boundary layer. Currently, the detection of atmospheric fronts relies on low-resolution global Numerical Weather Prediction (NWP) gridded datasets, which cannot analyse the fine-scale structural features of atmospheric fronts. High-resolution Synthetic Aperture Radar (SAR) imagery can be used to study the structure and characteristics of atmospheric fronts, but the potential of SAR-retrieved high-resolution wind field for atmospheric fronts detection has not yet been fully exploited. In this paper, we have presented for the first time a method for atmospheric fronts detection and cold/warm type determination based on sea surface wind field retrieval from SAR imagery. The proposed method was experimentally validated using Sentinel-1 SAR images, and the detection and determination results were compared with those from classical methods and the surface maps analyses datasets. The results showed good consistency, demonstrating the effectiveness of the proposed method.

Refining Heterogeneities near the Core–Mantle Boundary Beneath East Pacific Regions: Enhanced Differential Travel-Time Analysis Using USArray

Recent advancements in seismic data analysis have enhanced our grasp of the seismic heterogeneities near the core–mantle boundary (CMB). Through seismic tomography, persistent lower-mantle structures like the large low shear velocity provinces (LLSVPs) beneath the Pacific and South Africa have been identified. However, variations in the finer-scale features across different models raise questions about their origins. This study utilizes differential travel-time measurements from the USArray, operational across the contiguous United States from 2007 to 2014, to examine the impact of upper-mantle heterogeneities on tomographic models. By averaging the P-wave travel times and calibrating them with diffracted P-waves at the same stations, we mitigate the effects of shallow heterogeneities. The findings confirm that this method accurately maps the seismic anomalies beneath the USArray, consistent with other North American studies. Calibrated Pdiff travel-time data indicate significant anomalies in the mid-Pacific and Bering Sea and lesser anomalies in the northern Pacific, aligning with the global tomographic images. Moreover, the study highlights sharp travel-time variations over short distances, such as those across the northern boundary of the mid-Pacific anomaly, suggesting a chemically heterogeneous Pacific LLSVP.

Next-Generation Remote Sensing Data at Multiple Spatial Scales Improves Understanding of Habitat Selection by a Small Mammal.

Recent advances in optical remote sensing (RS) technology in combination with lightweight Global Positioning System (GPS) tracking devices now make analyzing the multi-scale habitat selection (HS) of small mammals < 2 kg possible. However, there have been relatively few multi-scale HS studies integrating fine-scale RS data with data-rich, GPS-derived movement data from small mammals. This is critical because small mammals commonly select habitat features across multiple scales. To address this gap, we investigated the HS of a small mammal, fox squirrels (Sciurus niger), which are known to cover relatively large areas and select fine-scale environmental features. We specifically asked the following questions: (1) Do next-generation RS variables improve HS models at single spatial scales? (2) Do multi-scale HS models improve upon those at single spatial scales? Using data from 45 individuals, we constructed HS models at three spatial scales: 4 ha (210 m × 210 m), 0.09 ha (30 m × 30 m), and 0.01 ha (10 m × 10 m) using traditional and next-generation RS data. The 4-ha model, using traditional and next-generation RS data, produced the best single-scale model, explaining 58% of the variations in HS. However, the multi-scale model provided the most informative model, explaining 68% of the variations in HS. Our models provide evidence for the value of next-generation RS data when quantifying HS and additional support for the idea of studying HS at multiple spatial scales.

Duets convey information about pair and individual identities in a Neotropical bird

Abstract Vocal individuality is essential for social discrimination but has been poorly studied in animals that produce communal signals (duets or choruses). Song overlapping and temporal coordination make the assessment of individuality in communal signals more complex. Additionally, selection may favor the accurate identification of pairs over individuals by receivers in year-round territorial species with duetting and long-term pair bonding. Here, we studied pair and individual vocal signatures in the polyphonal duets of rufous horneros Furnarius rufus, a Neotropical bird known for its long-term pair bonds. Hornero partners engage in duets to deter territorial intruders and protect their partnership year-round, and can discern duets from neighbors versus strangers. Using a dataset of 471 duets from 43 pairs in two populations, we measured fine-scale acoustic features across different duet levels (e.g., complete duets to non-overlapping syllable parts) and analysis levels (pair or individual). Permuted linear discriminant function analyses classified pairs and individuals more accurately than expected by chance (means: 45% and 47% versus 4 and 2%). Pair identity explained more variance in the multivariate acoustic features of duets than individual or population identities. The initial frequency of the duet showed strong potential for encoding pair identity. The acoustic traits contributing most to individual vocal signatures varied between sexes, which might facilitate the simultaneous assessment of duetters’ identities by receivers. Our study indicates that vocal individuality may exist even in species with intricate and innate communal signals and elucidates the mechanisms employed by horneros in their social discrimination ability.

Predicting unsaturated soil strength of coarse-grained soils for mobility assessments

Accurately estimating surficial soil moisture and strength is integral to determining vehicle mobility and is especially important over large spatial extents at fine resolutions. While methods exist to estimate soil strength across landscapes, they are empirical and rely on class average soil behavior. The Strength of Surficial Soils (STRESS) model was developed to estimate moisture-variable soil strength with a physics-based approach. However, there is a lack of field data from a diverse landscape to evaluate the STRESS model. To test the STRESS model, a field study was conducted in northern Colorado. Soil moisture and strength were measured on 10 dates. Data from the surficial layer (0–5 cm) were used to test the STRESS model and determine if soil strength trends could be estimated from topographical and soil textural differences. High variability was observed in soil strength measurements stemming from fine-scale terrain features and user variability. Observations show lower soil strengths for greater soil moistures, steeper slopes, higher vegetation, and lower soil fines content. STRESS estimated rating cone index values with a mean relative error of 37.5 %, while a pre-existing model had a mean relative error of 47.4 %. The STRESS model outperforms the current RCI prediction method, but uncertainty remains large.

Enhancing Surface Wind Speed and Temperature Prediction Using Surface-Layer Emulator and Transfer Learning

Abstract Accurate prediction of surface wind speed and temperature is crucial for many sectors. Physical schemes in numerical weather prediction (NWP) and data-driven correction approaches have limitations due to uncertainties in parameterization and lack of robustness, respectively. This study introduces a physics-informed data model, PhyCorNet, which combines a deep learning–based physics emulator (PhyNet) and a subsequent correction network (CorNet). PhyNet imitates the revised surface-layer parameterization scheme [default option of the Weather Research and Forecasting (WRF) Model]. CorNet refines predictions by mitigating the difference between PhyNet and observations. PhyCorNet enables gridded forecasts despite the training data being point based. Therefore, PhyCorNet can be regarded as a rediagnostic scheme for surface wind speed at 10 m (WS10) and temperature at 2 m (T2). Compared to WRF, it reduced diagnostic root-mean-square errors in the 24-h forecasts for WS10 and T2 by 40% and 36%, respectively, across China, with unbiased forecasts at almost all sites. PhyCorNet addresses the oversmoothing prediction issue of other deep learning models by providing the ability to represent fine-scale features and perform well in statistically extreme samples. In grid cells without observations for training, PhyCorNet performed much better than WRF, demonstrating the zero-shot learning capability. This study implies that the emulator plus bias correction provided by PhyCorNet could be used as a simple but effective approach to improve the performance of other diagnostic quantities in NWP. Significance Statement We developed a new model called PhyCorNet to improve the surface wind speed and temperature forecasts. Existing weather forecast models have limitations in accurately predicting these variables. PhyCorNet first uses a neural network (PhyNet) to mimic an existing physics-based wind and temperature calculation method. Another neural network [correction network (CorNet)] improves the PhyNet forecasts by comparing them with observations. Across China, PhyCorNet reduced 24-h forecast errors for wind speed by 40% and temperature by 36% compared to a conventional model. It performed well under diverse conditions and overcame the oversmoothing issues faced by other machine learning models. Importantly, PhyCorNet outperformed traditional models in areas without historical observations. We suggest that this new framework can also be used to enhance the forecasts of other atmospheric variables.

Mesoscale cyclonic eddies born in an eastern boundary upwelling system enhance microbial eukaryote diversity in oligotrophic offshore waters

Mesoscale eddies which originate in Eastern Boundary Upwelling Systems (EBUS) such as the Canary Current System, entrap nutrient-rich coastal water and travel offshore while aging. We have analyzed the protistan plankton community structures in the deep chlorophyll maximum (DCM), sub-DCM, and oxygen minimum zone (OMZ) of three differently aged cyclonic EBUS eddies off Northwest Africa, as well as of non-eddy affected reference sites, using DNA metabarcoding. Throughout all water depths, we found that the investigated eddies generated local dispersal-driven hotspots of protistan plankton diversity in the naturally oligotrophic subtropical offshore waters off Northwest Africa. Based on the taxonomic composition of protistan plankton communities, these diversity hotspots are likely to play an important role in carbon sequestration and for regional food webs up to top predatory levels. Thereby, the life-span of an eddy emerged as an important criterion, how local offshore protistan plankton diversity is transformed quantitatively and qualitatively: each of the three eddies was characterized by notably distinct protistan plankton communities. This could be linked to the physicochemical water properties (predominantly macronutrients, temperature, and salinity) of the eddies' cores and rings, which experience pronounced changes during the eddies’ westward trajectories. Furthermore, we found evidence that eddy-specific deep-water protistan communities are relatively short-lived compared to the ones in the sunlit DCM. However, our results do not only witness from the importance of fine-scale physical ocean features for regional ecosystem processes, but they also show the complexity of these ocean features and that we are still far from understanding the biological processes and their driving forces in such features.

Determination of Microtopography of Low-Relief Tidal Freshwater Forested Wetlands Using LiDAR

The microtopography of tidal freshwater forested wetlands (TFFWs) impacts biogeochemical processes affecting the carbon and nitrogen dynamics, ecological parameters, and habitat diversity. However, it is challenging to quantify low-relief microtopographic features that might only vary by a few tens of centimeters. We assess the high-resolution fine-scale microtopographic features of a TFFW with terrestrial LiDAR and aerial LiDAR to test a method appropriate to quantify microtopography in low-relief forested wetlands. Our method uses a combination of water-level and elevation thresholding (WALET) to delineate hollows in terrestrial and aerial LiDAR data. Close-range remote sensing technologies can be used for microtopography in forested regions. However, the aerial and terrestrial LiDAR technologies have not been used to analyze or compare microtopographic features in TFFW ecosystems. Therefore, the objectives of this study were (1) to characterize and assess the microtopography of low-relief tidal freshwater forested wetlands and (2) to identify optimal elevation thresholds for widely available aerial LiDAR data to characterize low-relief microtopography. Our results suggest that the WALET method can correctly characterize the microtopography in this area of low-relief topography. The microtopography characterization method described here provides a basis for advanced applications and scaling mechanistic models.

A new high-resolution Coastal Ice-Ocean Prediction System for the East Coast of Canada

The Coastal Ice Ocean Prediction System for the East Coast of Canada (CIOPS-E) was developed and implemented operationally at Environment and Climate Change Canada (ECCC) to support a variety of critical marine applications. These include support for ice services, search and rescue, environmental emergency response and maritime safety. CIOPS-E uses a 1/36° horizontal grid (~ 2 km) to simulate sea ice and ocean conditions over the northwest Atlantic Ocean and the Gulf of St. Lawrence (GSL). Forcing at lateral open boundaries is taken from ECCC’s data assimilative Regional Ice-Ocean Prediction System (RIOPS). A spectral nudging method is applied offshore to keep mesoscale features consistent with RIOPS. Over the continental shelf and GSL, the CIOPS-E solution is free to evolve according to the model dynamics. Overall, CIOPS-E significantly improves the representation of tidal and sub-tidal water levels compared to ECCC’s lower resolution systems: RIOPS (~ 6 km) and the Regional Marine Prediction System – GSL (RMPS-GSL, 5 km). Improvements in the GSL are due to the higher resolution and a better representation of bathymetry, boundary forcing and dynamics in the upper St. Lawrence Estuary. Sea surface temperatures show persistent summertime cold bias, larger in CIOPS-E than in RIOPS, as the latter is constrained by observations. The seasonal cycle of sea ice extent and volume, unconstrained in CIOPS-E, compares well with observational estimates, RIOPS and RMPS-GSL. A greater number of fine-scale features are found in CIOPS-E with narrow leads and more intense ice convergence zones, compared to both RIOPS and RMPS-GSL.

Simulating cardiac fluid dynamics in the human heart.

Cardiac fluid dynamics fundamentally involves interactions between complex blood flows and the structural deformations of the muscular heart walls and the thin valve leaflets. There has been longstanding scientific, engineering, and medical interest in creating mathematical models of the heart that capture, explain, and predict these fluid-structure interactions (FSIs). However, existing computational models that account for interactions among the blood, the actively contracting myocardium, and the valves are limited in their abilities to predict valve performance, capture fine-scale flow features, or use realistic descriptions of tissue biomechanics. Here we introduce and benchmark a comprehensive mathematical model of cardiac FSI in the human heart. A unique feature of our model is that it incorporates biomechanically detailed descriptions of all major cardiac structures that are calibrated using tensile tests of human tissue specimens to reflect the heart's microstructure. Further, it is the first FSI model of the heart that provides anatomically and physiologically detailed representations of all four cardiac valves. We demonstrate that this integrative model generates physiologic dynamics, including realistic pressure-volume loops that automatically capture isovolumetric contraction and relaxation, and that its responses to changes in loading conditions are consistent with the Frank-Starling mechanism. These complex relationships emerge intrinsically from interactions within our comprehensive description of cardiac physiology. Such models can serve as tools for predicting the impacts of medical interventions. They also can provide platforms for mechanistic studies of cardiac pathophysiology and dysfunction, including congenital defects, cardiomyopathies, and heart failure, that are difficult or impossible to perform in patients.

Moho depth model of South America from a machine learning approach

Crustal structure models play an important role in the characterization of seismogenic zones, in the regional delimitation of geological provinces and particularly, in understanding the genesis and evolution of sedimentary basins. Despite the increasing number of new seismic surveys and seismographic networks in South America, crustal thickness measurements are still scarce and irregularly sampled, reducing the resolution of crustal model maps. To overcome these challenges, a novel approach based on machine learning techniques is proposed, in order to explore higher resolution gravity datasets in the interpolation of crustal thickness measurement points obtained from previous seismic/seismological compilations. The algorithm used in this study is based on Gaussian processes prediction methods, which allowed inferring the depth of Moho to South America. The prediction error of the model obtained from the training and testing database was 3.48 km, which is compatible with the uncertainties derived from the H-k stacking analysis. The depth range varied from 69.8 km beneath the Andes to 4.3 km in oceanic regions. The average Moho depth for the South American Platform is 39.1 km, allowing a spatial correlation of deeper and shallower Moho regions with different types of continental basins. Compared to other models, the model resulting from this study presents fine-scale features highlighting the limits of the main tectonic domains and a good agreement with the suture zones. Overall, this study demonstrates the potential of applying machine learning tools in crustal-scale imaging using sparse datasets, providing new advances in Moho modeling of the South America, as well as new perspectives on its the history and tectonic evolution.

Identifying fine-scale archaeological features using KH-9 HEXAGON mapping and panoramic camera images: evidence from Liangzhu Ancient City

ABSTRACT Historical fine-scale information of archaeological landscapes is crucial in archaeological investigations. However, documenting such information using satellite sensor data prior to 2000 remains a daunting challenge. Images from the declassified archives of KH-9 HEXAGON (KH-9) cameras, including the panoramic camera system (PCS) and mapping camera system (MCS), offer fine-scale information about archaeological sites. However, noise, contrast distortion and the availability of only a single panchromatic band can limit their potential, particularly for identifying features in subtropical climates within heterogeneous landscape types. This paper focuses on developing a novel multifaceted analytical framework with two components: image pre-processing and feature identification. The image pre-processing component is divided into two steps. First, a trained stationary wavelet transform (SWT) based on the normalized sill (NS) is developed to not only de-noises the image, but also preserve its original image characteristics. Then, the contrast of the de-noised images is optimized by the multi-resolution Top-hat (MTH) using multi-scale information. In the feature identification component, the MCS image is analysed using spatial colour composite write function memory (SCCWFM) and spatial novelty detection (SND). An ultra-fine spatial three-dimensional colour composite (UFSTCC) image and ultra-fine spatial digital surface model (UFSDSM) are produced to aid interpretation of the KH-9 PCS images. The proposed processing pipelines were tested on KH-9 MCS and PCS images of the World Heritage site at Liangzhu Ancient City (LAC) in China, which is characterized by a subtropical climate and a heterogeneous landscape types. The proposed pre-processing pipeline improved considerably the appearance of these images across the LAC landscape while maintaining the original image information. The developed digital analytical approaches for KH-9 PCS and MCS images facilitated straightforward identification of archaeological features in the LAC. The proposed framework has the potential to increase exploitation of the available KH-9 images in archaeological applications.

Fine-Scale Lithogeochemical Features Influence Plant Distribution Patterns in Alpine Grasslands in the Western Alps of Italy.

Bedrock geology is crucial in structuring alpine plant communities. Old studies mainly focused on the compositional differences between alpine plant communities on carbonate rocks and crystalline rocks, i.e., calcareous vs. siliceous vegetation. Increasing attention is being paid to bedrock types other than calcareous or siliceous ones, viz. those which have intermediate geochemical characteristics between pure calcareous and pure siliceous ones. Among these types of 'intermediate' bedrocks, calc-schists and serpentines are generally characterized by vegetation comprised of a mixture of basiphilous and acidophilous species. We selected several sites in alpine grasslands in the Western Italian Alps, on calc-schist and serpentine bedrocks, located at 2500 ± 100 m above sea level. X-ray fluorescence quantification of major and trace elements, combined with stereomicroscopic examination of bedrock samples with a petrographic approach, revealed a much broader range of bedrock types than recognized by inspection of geological maps. The vegetation investigated in our study was mostly composed of a set of species found more or less frequently in alpine silicicolous or calcicolous plant communities of the Alps and other European mountains. The carbonate content in the bedrock was one of the main drivers of variation in grassland vegetation, not necessarily related to soil pH. There were no distinctive species uniquely characterizing grassland vegetation on serpentines or calc-schists.

Enhanced detection of Aspergillus flavus in peanut kernels using a multi-scale attention transformer (MSAT): Advancements in food safety and contamination analysis

In this study, a multi-scale attention transformer (MSAT) was coupled with hyperspectral imaging for classifying peanut kernels contaminated with diverse Aspergillus flavus fungi. The results underscored that the MSAT significantly outperformed classic deep learning models, due to its sophisticated multi-scale attention mechanism which enhanced its classification capabilities. The multi-scale attention mechanism was utilized by employing several multi-head attention layers to focus on both fine-scale and broad-scale features. It also integrated a series of scale processing layers to capture features at different resolutions and incorporated a self-attention mechanism to integrate information across different levels. The MSAT model achieved outstanding performance in different classification tasks, particularly in distinguishing healthy peanut kernels from those contaminated with aflatoxigenic fungi, with test accuracy achieving 98.42±0.22%. However, it faced challenges in differentiating peanut kernels contaminated with aflatoxigenic fungi from those with non-aflatoxigenic contamination. Visualization of attention weights explicitly revealed that the MSAT model's multi-scale attention mechanism progressively refined its focus from broad spatial-spectral features to more specialized signatures. Overall, the MSAT model's advanced processing capabilities marked a notable advancement in the field of food quality safety, offering a robust and reliable tool for the rapid and accurate detection of Aspergillus flavus contaminations in food.

How does street environment affect pedestrian crash risks? A link-level analysis using street view image-based pedestrian exposure measurement

Street space plays a critical role in pedestrian safety, but the influence of fine-scale street environment features has not been sufficiently understood. To analyze the effect of the street environment at the link level, it is essential to account for the spatial variation of pedestrian exposure across street links, which is challenging due to the lack of detailed pedestrian flow data. To address these issues, this study proposes to extract link-level pedestrian exposure from spatially ubiquitous street view images (SVIs) and investigate the impact of fine-scale street environment on pedestrian crash risks, with a particular focus on pedestrian facilities (e.g., crossing and sidewalk design). Both crash frequency and severity are analyzed at the link level, with the latter incorporating two distinct aggregation metrics: maximum severity and medium severity. Using Hong Kong as a case study, the results show that the link-level pedestrian exposure extracted from SVIs can lead to better model fit than alternative zone-level measurements. Specifically, higher pedestrian exposure is found to increase the total pedestrian crash frequency, while reducing the risk of serious injuries or fatalities, confirming the “safety in numbers” effect for pedestrians. Pedestrian facilities are also shown to influence pedestrian crash frequency and severity in different ways. The presence of crosswalks can increase crash frequency, but denser crosswalk design mitigates this effect. In addition, two-side sidewalks can increase crash frequency, while the absence of sidewalks leads to higher risks of crash severity. These findings highlight the importance of fine-scale street environment and pedestrian facility design for pedestrian safety.