Quantifying jellyfish bloom dynamics using aerial surveys and image-based cohort analysis in a Mediterranean lagoon

Fire can profoundly affect ecosystem dynamics, species distribution and plant traits, especially in open biomes. Post-fire strategies, namely, resprouters and reseeders, offer a useful framework to examine eco-evolutionary relationships between plants and fire. However, whether resprouter and reseeder plants are consistently formed by distinct trait coordination (syndromes) and responses to fire at the intraspecific level and when considering the role of ontogeny, remain underexplored. This is a relevant lack as, within-species, plants can adjust their functioning and trait coordination can vary considerably along ontogeny. To address this gap, we analysed intraspecific trait coordination and post-fire responses, accounting for the effect of ontogeny in three widely distributed and locally abundant Mediterranean woody species: two resprouters (Erica arborea, Quercus ilex) and one reseeder (Cistus salviifolius). We collected 12 plant functional and architectural traits, including intraspecific variability, well related to fire and drought from three sites in Italy. We ran pairwise correlation and multivariate analyses to explore trait syndromes. We conducted linear regressions to examine relationships between fire regime (time since last fire) and trait responses. We then inspected whether fire regime affects key bivariate trait coordination and if ontogeny influences some trait-fire links. Findings are highly species-specific and generally do not align with a priori classification into post-fire strategies. In most instances, we reveal how either one of the resprouter species exhibits trait patterns more similar to those of the reseeder than to the other resprouter species. Fire can strongly affect trait coordination shaping plant functioning, whereas ontogeny influences a few trait-firelinks for the reseeder species while it has a weak effect on the two resprouter species. Our study, while limited to three species and three sites, emphasizes the importance of looking at plant life through a continuous and multidimensional lens which contemplates the inclusion of various sources of within-species variability. We acknowledge that a category-based or dichotomous view on plant functional strategies, including post-fire ones, remains valid and justified when working at coarse scales, whereas it can be much less so for trait-based analyses at fine scales.

Abstract Morphing skin panels capable of undergoing large deformations are being investigated as a means of reducing fuel burn in next-generation aircraft through drag reduction. One approach to creating these panels is to 3D print Zero Poisson’s ratio cellular cores onto elastomeric facesheets, as proposed by the Geometrically Anisotropic ThermOplastic Rubber (GATOR) skins concept. Modelling these panels remains challenging due to multiple factors, including small feature sizes, hyperelastic material properties, geometric nonlinearities under large deformations, and complex interactions between the core and facesheets. While high-fidelity finite element analysis works well for unit cells and small components, it becomes prohibitively expensive for large-scale aircraft components. Hence, this work examines various approaches for mitigating the computational cost of analysing these sandwich panels, focusing on a morphing joint fairing for a folding wingtip. The analysis of these panels is simplified through scale separation, where the mechanical responses of the sandwich panel are homogenised into equivalent stiffness properties at the fine scale, allowing the structure to be modelled as a shell surface with equivalent stiffness properties at the coarse scale. This multiscale modelling approach can use homogenisation methods based on analytical formulations and finite element analysis, as well as shell formulations such as Kirchhoff-Love and Reisner-Mindlin plate theory, each with simplifying assumptions that affect the accuracy of the results. This paper examines the effects of these approaches and their underlying assumptions on solution accuracy and computational cost, providing a comprehensive understanding of the modelling approaches available. The results indicate that the analytical homogenisation underpredicts the stiffness of the panel, the effects of transverse shear deformation become negligible for long panels, and the geometric nonlinearity in the panel deformation becomes significant for large rotations of the wingtip. The paper further presents an approach based on data-driven, second-order, multiscale modelling that captures the geometric nonlinearities in GATOR panels’ responses accurately at a reduced computational cost.

Frequent and objective assessment of ataxia severity is essential for tracking disease progression and evaluating the effectiveness of potential treatments. Wearable-based assessments have emerged as a promising solution. However, existing methods rely on inertial data features directly correlated with subjective and coarse clinician-evaluated rating scales, which serve as imperfect gold standards. This approach may introduce biases and restrict flexibility in feature design. To address these limitations, this study introduces a novel contrastive learning-based model that leverages motor severity differences in wearable inertial data to learn relevant features. The model was trained on inertial data collected from 87 individuals with diagnostically heterogeneous ataxias and 44 healthy participants performing the finger-to-nose task. A pairwise contrastive loss function was proposed to learn representations capturing relative differences in ataxia severity, which were evaluated through downstream regression and classification tasks. The learned features demonstrated strong cross-sectional (r = 0.84) and longitudinal (r = 0.68) associations with clinical scores and robust measurement reliability (intraclass correlation coefficient = 0.96). Additionally, the model exhibited strong known-group validity, distinguishing between ataxia and healthy phenotypes with an area under the receiver operating characteristic curve of 0.95. The proposed contrastive model captures robust representations of disease severity with reduced reliance on clinical scales, outperforming state-of-the-art methods that derive features directly from clinical scores. Combining wearable sensors with contrastive learning enables a more objective, scalable, and frequent method for assessing ataxia severity, with the potential to enhance patient monitoring and improve clinical trial efficiency.

Existing research fails to balance flow detail capture with data stationarity and suffers from feature homogenisation in spatiotemporal modelling. To address these challenges, an innovative three-stage forecasting framework is proposed. Firstly, an optimal temporal granularity selection model is constructed by integrating morphological similarity, distributional consistency, and stationarity tests. Increasing similarity or stationarity weights shifts the optimal granularity toward finer or coarser scales, respectively. Second, a Clustering-Guided Adaptive Multi-Graph Fusion Network resolves feature homogenisation after multi-graph convolution from four complementary perspectives. Finally, the LSTM-KanFormer framework is designed, achieving high-precision passenger flow prediction via a novel multi-head attention mechanism that effectively fuses LSTM, KAN, and causal convolution. Experiments demonstrate that at the optimal 15 min granularity, all models achieve optimal forecasting performance, with LSTM-KanFormer significantly outperforming other baseline models in forecasting fitting. Moreover, CGA-AMGCN’s clustering outperforms traditional methods in global structure, information relevance, and intra-cluster consistency.

Multiscale entropy (MSE) as a measure of brain complexity provides substantial insights into the adaptability of the brain. However, it is often applied to resting-state electroencephalography (EEG) or in static tasks. The current study assessed the reliability, validity and classification accuracy of MSE computed on mobile EEG data for linking brain complexity to motor performance within a kicking task. Eleven novice participants underwent repeated measurements to assess test-retest reliability, while the data from 15 novices and 15 football players were used to evaluate known-groups validity, convergent validity and classification accuracy. EEG data were recorded using 65 active electrodes and MSE estimates were computed for 64 time scales on preprocessed data. Results showed poor to excellent reliability for MSE estimates, exhibiting channel- and scale-specific variations, with reliability generally higher at fine-to-mid scales. Experts exhibited significantly lower entropy at coarse scales in left frontal and at fine scales in centroparietal regions compared to novices. Negative correlations were found between entropy estimates and kicking accuracy. Receiver operating characteristic curves of entropy estimates and their principal components demonstrated moderate to good classification accuracy between expertise levels. These findings suggest MSE as a promising metric for investigating brain complexity in movement contexts, revealing distinct patterns of complexity associated with motor performance. Future research across diverse tasks and populations is crucial to further elucidate this relationship and explore the applied potential of MSE.

Scale-dependent evidence for the abundant-center hypothesis and underlying mechanisms

Coral reefs are important both economically and culturally to over 1 billion people. However, reefs continue to be threatened by climate change, with some areas now experiencing mass coral bleaching and mortality events due to heat stress on an annual basis. Satellite-derived sea surface temperature data (SSST) are often used as a proxy for in situ temperatures on reefs, and are relied on to identify heat accumulation and assign bleaching risk on reefs worldwide. However, SSST has limitations – readings are only taken at night and on a relatively coarse spatial scale, and multiple studies have exposed discrepancies between SSST and in situ temperatures. In this study, we compare satellite-derived sea surface temperature in Palau to in situ temperature records at 87 reef locations in order to assess how well SSST captures physiologically important thermal trends experienced by corals. We find that while SSST captures average nightly temperatures relatively well, it fails to accurately capture thermal maxima, diurnal range in temperature and heat accumulation measurements like degree heating weeks (DHW) that are relevant in determining coral bleaching risk levels. Though SSST data remain key indicators of temperature stress over global scales, local management of coral reefs, coral restoration, and reef replenishment require more fine scale data in order to accurately understand thermal trends and their implications for coral resilience.

IntroductionAge-related changes in brain signal complexity are associated with cognitive decline and reduced neural adaptivity in older adults. Exergaming offers a promising prophylactic intervention combining physical and cognitive training. The aim of the present study was to assess how exergaming alters the temporal trajectory of brain signal complexity at rest and during gameplay in older adults.MethodsTwenty-eight healthy older adults participated in a 4-week exergaming intervention. Electroencephalography was recorded using 64 electrodes at rest (pre- and post-intervention) and during exergaming (pre-, mid-, and post-intervention). Brain signal complexity was quantified using multiscale entropy across 64 time scales on preprocessed signals.ResultsPost-intervention resting-state analysis revealed significant reductions at fine and increases at coarse scales in frontal, central, and posterior entropy. During gameplay, entropy declined widespread by mid-intervention, particularly at coarse scales over frontal, central and temporal regions. From mid- to post-intervention, the decline narrowed leaving a net pre-to-post reduction concentrated at coarse scales in these regions.DiscussionResting-state changes indicated a shift toward a younger brain profile, characterized by a transition from age-related increases in local processing to enhanced distributed processing, which may potentially mitigate the rise in neural modularity associated with aging. During gameplay, brain signal complexity decreased in week 2, followed by a modest change by week 4, consistent with the framework in which complexity initially streamlines and then adjusts toward a task-specific optimum. These findings suggest that exergaming can beneficially modulate brain complexity in older adults, offering the potential to reduce age-related neural decline and support healthy brain aging.

Entropy analysis of electroencephalographic (EEG) data provides insight into the complexity of neural activity. This study primarily examined whether cannabis use frequency is associated with alterations in multiscale entropy (MSE) in the prefrontal cortex (PFC) and mapped baseline MSE differences across cortical regions. Resting-state EEG was collected from 57 adults: non-users (n=18), low-frequency users (≤ 1x/week; n=24), and frequent users (≥ 2x/week; n=15). MSE values were binned into fine, medium, coarse, and very-coarse scale ranges. Linear mixed-effects models assessed group×scale bin interactions in the PFC and regional differences among lobes, irrespective of cannabis use. Entropy increased with coarser scales in all groups, but the slope was significantly flatter in frequent users. From the medium bin onward, their PFC scale-entropy slope was∼0.12 bits lower than in non-users, widening to∼0.16 bits at very-coarse scales (FDR-corrected q<0.012). Across all participants, the PFC exhibited lower MSE than parietal, occipital, and temporal lobes, with regional gaps expanding at coarser scales. The PFC exhibits intrinsically reduced signal complexity compared with other cortical regions, with further attenuation in frequent cannabis users. MSE captures regional cortical dynamics in resting EEG and detects cannabis-related reductions in prefrontal signal complexity at longer temporal scales.

Deep learning–based artificial intelligence approach has provided an extremely efficient means for evaluating aerodynamic pressure over aircraft surfaces. Yet current approaches face two major challenges: (1) the capacity to represent features in geometrically complex regions is insufficient, leading to low prediction accuracy in those areas, and (2) they usually adopt a shallow, fixed fusion strategy for flight conditions, ignoring the fact that different flight conditions influence different geometric regions to varying degrees, which limits accuracy improvement under multiple flight conditions. To tackle these two challenges, we propose a multi-scale, condition-adaptive deep-learning framework. First, to cope with complex-geometry representation, we design the Laplacian Geometric Feature Extractor. It employs discrete Laplacian operators to compute additional local features, markedly enhancing the model's perception of intricate geometries. Second, to achieve deep and adaptive fusion between flight conditions and geometric shapes, we build a multi-scale conditional-fusion pipeline: the Point-level Conditional Fusion module dynamically adjusts the influence of flight conditions on every single point, enabling fine-grained interaction, while the Global Conditional Fusion module optimizes the interaction between global geometry and flight conditions at a coarse scale, allowing condition effects to be adaptively tuned for different shapes. Experiments on four pressure-prediction datasets show that, under various aircraft geometries and flight conditions, the proposed framework achieves relative errors of 6.45%, 5.49%, 7.14%, and 2.88%, outperforming existing deep-learning methods. Ablation studies and fine-tuning experiments further verify the effectiveness, generalization capability, and transferability of our approach.

ABSTRACTWe sought to assess bridle shiner (Notropis bifrenatus) habitat associations at local and regional scales across southern Maine and New Hampshire. We used local habitat data at 95 Maine sites to predict occupancy with classification and regression trees (CART). We then used ensemble species distribution models (SDMs) to model the historical (1898–2008) and current (2009–2022) ranges of the species. We used the BIOMOD platform to model the association between 35 environmental variables and bridle shiner presence during both time periods and at fine (pseudo‐HUC14) and coarse (HUC12) spatial scales. We then calculated the change in predicted occupied drainages to estimate the change in the species' distribution at both scales. Within a site, bridle shiners were associated with submerged aquatic vegetation, organic substrate, and watermilfoil (Myriophyllum spp.). SDMs revealed an association with Appalachian (Hemlock‐)Northern Hardwood Forest, sand substrate, and low‐elevation terrain (at both spatial scales). Ensemble fine‐scale SDMs suggest a substantial loss of historical bridle shiner habitat in both Maine (36% of drainages) and New Hampshire (16%), with comparable described losses (of 21% and 14%) at a coarse scale. Our local and regional models may be used to focus surveys on areas with high predicted habitat suitability or to inform habitat restoration efforts.

Abstract Future population growth is expected to concentrate in urban agglomerations that overlap with various natural hazard zones. However, quantifying the resulting risks remains challenging, as hazard areas tend to be bounded locally while population forecasts are produced at much coarser scales. Addressing this gap, the high‐resolution 2UP model disaggregates national‐level, scenario‐based population projections to a 30 arc‐seconds grid, simultaneously simulating urban expansion and the distribution of urban and rural populations through 2100. By overlaying these projections with comparably detailed fluvial flood and landslide hazard data, this study demonstrates that, at a global scale, rapid urbanization will disproportionately increase population growth in hazard‐prone zones compared to safer areas. This trend is particularly pronounced in sub‐Saharan Africa and South Asia, where both the extent of exposed urban land and the magnitude of exposed populations are projected to rise sharply. In contrast, slower growth in North America and Europe leads to more moderate increases in hazard exposure, with smaller differences between hazardous and non‐hazardous sites. Notably, while urban areas in many countries continue expanding into high‐risk regions, the fraction of the total population exposed to these hazards may stabilize or even decline after 2050. The 2UP model's fine‐grained outputs are especially valuable in regions with fragmented urban landscapes, large rural populations, and rapid demographic shifts, providing decision‐makers and researchers with critical insights for integrated risk management and sustainable development planning.

• Four different snow melt-out data sources at high to very high spatial resolution • Comparison of predictive power of these melt-out datasets in SDMs • Fine-scale and in-situ melt-out did not improve SDMs more than coarse-scale melt-out • Coarse-scale melt-out from satellite imagery and a process-based snow model performed surprisingly well in SDMs • Collection of fine-scale melt-out data might be worthwhile for individual species, but not in general Snow cover is a crucial driver for plant species distributions in cold environments. The primary source of snow cover data used in distribution models is remotely sensed satellite imagery, which is characterized by coarser spatial resolutions than plot-scale observations of plant distributions. This scale-mismatch was hypothesized to limit model accuracy. Here, we used a common modeling framework to assess the contribution of snow melt-out dates derived from four data sources (satellite imagery, numerical snowpack modeling, webcam imagery and in-situ soil temperature measurements) at 1 m and 20 m spatial resolution to the predictive power of distribution models of 74 plant species in an alpine landscape of the Austrian Alps. We found that >80 % of the distribution models of all species were significantly improved by at least one snow melt-out data set when considering Area Under the Curve (AUC). Satellite-based melt-out led to significantly improved models for the highest number of species (>50 % for AUC) and increased True-Skill-Statistic and AUC on average by 16 % and 5 %, respectively. Surprisingly, fine-scale and in-situ measured melt-out data did not improve models more than the coarser scale (20 m) satellite-based melt-out data. Moreover, numerical snowpack modeling delivered results comparable to the other sources, which supports its use for projecting future species distributions. We conclude that the additional effort needed for producing high resolution, in-situ datasets as compared to commonly used satellite imagery might hence be worthwhile for some species but not for plant distribution modeling in cold ecosystems in general.

Coral reefs are important both economically and culturally to over 1 billion people. However, reefs continue to be threatened by climate change, with some areas now experiencing mass coral bleaching and mortality events due to heat stress on an annual basis. Satellite-derived sea surface temperature data (SSST) are often used as a proxy for in situ temperatures on reefs, and are relied on to identify heat accumulation and assign bleaching risk on reefs worldwide. However, SSST has limitations - readings are only taken at night and on a relatively coarse spatial scale, and multiple studies have exposed discrepancies between SSST and in situ temperatures. In this study, we compare satellite-derived sea surface temperature in Palau to in situ temperature records at 87 reef locations in order to assess how well SSST captures physiologically important thermal trends experienced by corals. We find that while SSST captures average nightly temperatures relatively well, it fails to accurately capture thermal maxima, diurnal range in temperature and heat accumulation measurements like degree heating weeks (DHW) that are relevant in determining coral bleaching risk levels. Though SSST data remain key indicators of temperature stress over global scales, local management of coral reefs, coral restoration, and reef replenishment require more fine scale data in order to accurately understand thermal trends and their implications for coral resilience.

The SDG 15.3.1 framework provides a standardized approach using land use/land cover (LULC) change, land productivity, and soil organic carbon (SOC) dynamics to assess land degradation. However, SDG 15.3.1. faces limitations like coarse resolutions of Landsat-8 and Sentinel-2, particularly for fine-scale studies. Accordingly, this paper integrates Very Deep Super-Resolution (VDSR) for downscaling Landsat-8 imagery to 1 m resolution and the Vegetation Health Index (VHI) into SDG 15.3.1 to enhance detection in the heterogeneous Loiret region, France—a temperate agricultural hub featuring mixed croplands and peri-urban interfaces—using 2017 as baseline and 2024 as target. Results demonstrated that 1 m resolution detected more degraded LULC areas than coarser scales. SOC degradation was minimal (0.15%), concentrated in transitioned zones. VHI reduced overestimation of productivity declines compared to the Normalized Difference Vegetation Index by identifying more stable areas and 2.69 times less degradation in integrated assessments. The “One Out, All Out” rule classified 2.6% (using VHI) and 7.1% (using NDVI) of the region as degraded, mainly in peri-urban and cropland hotspots. This approach enables metre-scale land degradation mapping that remains effective in heterogeneous landscapes where fine-scale LULC changes drive degradation and would be missed at lower resolutions. However, future ground validation and longer timelines are essential to enhance the presented methodology.

The spatial resolution of current microwave remote sensing soil moisture (SM) data is about 25 km in global scale. The coarse scale hinders the application of SM product at regional scale. The global 9 km SM can be released by radar observations of Soil moisture Active and Passive (SMAP) satellite since 2015. For the failed radar sensor, SMAP 9 km SM is less than three months. Therefore, European Space Agency Climate Change Initiative (CCI) SM data is downscaled to 9 km using spatial temporal fusion model in the study. And the 43-year 9 km SM is downscaled by CCI data from 1978 to 2020. Results display that downscaled 9 km SM gets more detailed spatial information than CCI data. Moreover, temporal variation of CCI data in anomaly can be well captured by downscaled data. The evaluations against in-situ data indicate that temporal accuracies of downscaled data (r = 0.676, μbRMSE = 0.069 m3/m3) are comparable with CCI data (r = 0.670, μbRMSE = 0.070 m3/m3). Overall, downscaled data improves the spatial resolution of CCI data and inherits the temporal accuracy with slight improvement. Higher spatial resolution SM offers greater application potential. Additionally, the model herein enriches SM downscaling techniques.

Urban neighborhoods in India face an uneven distribution and limited accessibility to parks and playgrounds, particularly in dense mixed-use areas where rapid urbanization constrains green infrastructure planning. To address these challenges, the Sustainable Transportation Assessment Index (SusTAIN) framework was developed to evaluate sustainable transportation in Indian urban neighborhoods, with ‘Accessibility’ identified as a crucial subtheme. Recent advancements in Geographic Information Systems (GISs) and urban data analysis tools have enabled accessibility assessments of parks and playgrounds at a neighborhood scale, yet the OSMnx approach has been only marginally explored and compared in the literature. This study addresses this gap by comparing three tools—the Quantum Geographic Information System (QGIS), OSMnx, and Space Syntax—for accessibility assessments of parks and playgrounds in Ward 60 of Kalyan Dombivli city, based on the 15-Minute City concept. Accessibility was evaluated using 25 m and 100 m grid resolutions under peak and non-peak conditions across public and private transportation modes. The findings show that QGIS offers highly consistent results at micro-scale (25 m), while OSMnx provides better accuracy at coarser scales (100 m+). The results were validated with space syntax through integration and choice values. The comparison highlights spatial disparities in accessibility across different tools and transportation modes, including Intermediate Public Transport (IPT), which remains underexplored despite its crucial role in last-mile connectivity. The presented approach can support municipal authorities in optimizing neighborhood mobility and is transferable for applying the SusTAIN framework in other urban contexts.

Hierarchical analysis of spreading dynamics in complex systems

• We identified hotspots and coldspots at four scales for three confidence levels. • We compared rigorously collected data to data collected by highway patrol personnel. • We found considerably more animals and identified more species than highway patrol. • Rigorous surveys provide a better understanding of impacts of roads on biodiversity. • Installation of fences combined with designated wildlife passages is needed. We conducted 66 road mortality surveys along a 31.5 km stretch of a high-traffic 4-lane highway between Montreal and Sherbrooke in Québec, Canada. Surveys by vehicle between May and August 2019 recorded 212 animal carcasses from 48 species. Hotspots and coldspots for ground-dwelling vertebrates (mammals, amphibians, and reptiles) and birds were identified at four scales. The number of hotspots was higher at finer scales, while the combined length of hotspots was greater at coarser scales. We found significantly more animals and identified more species than the highway patrol, suggesting that rigorous road mortality surveys are beneficial for better understanding road impacts on biodiversity. We estimated that the highway patrol’s reporting probability for medium-sized mammals (more than 0.65 kg, less than 30 kg) was between 21 % and 54 % that of our systematic surveys. We recommend priority locations for mitigation to reduce road mortality and re-establish connectivity between wildlife populations separated by the highway.