Fant 2634 publikasjoner. Viser side 2 av 264:
Road traffic externalities are an important consequence of land-use and transport interactions and may be especially induced by their inefficient combinations. In this study, we integrate land-use, transport and emission modelling tools (the LUTEm framework) to assess how suburban expansion vs. inward densification scenarios influence journey parameters, road network performance and traffic emissions. Case-study simulations for Warsaw (Poland) underscore the negative consequences of suburban sprawl development, which are hardly mitigated by additional land-use or transport interventions, such as rebalancing of population-workplace distribution or road capacity reductions. On the other side, compact city development lowers global traffic congestion and emissions, but can also raise the risks of traffic externalities in central city area unless complemented with further interventions such as improved public transport attractiveness. This study aims to enrich the understanding of how integrating the land-use development and transport interventions can ultimately influence travel parameters and reduce urban road traffic externalities.
Elsevier
2025
Potato plant disease detection: leveraging hybrid deep learning models
Agriculture, a crucial sector for global economic development and sustainable food production, faces significant challenges in detecting and managing crop diseases. These diseases can greatly impact yield and productivity, making early and accurate detection vital, especially in staple crops like potatoes. Traditional manual methods, as well as some existing machine learning and deep learning techniques, often lack accuracy and generalizability due to factors such as variability in real-world conditions. This study proposes a novel approach to improve potato plant disease detection and identification using a hybrid deep-learning model, EfficientNetV2B3+ViT. This model combines the strengths of a Convolutional Neural Network - EfficientNetV2B3 and a Vision Transformer (ViT). It has been trained on a diverse potato leaf image dataset, the “Potato Leaf Disease Dataset”, which reflects real-world agricultural conditions. The proposed model achieved an accuracy of 85.06, representing an 11.43 improvement over the results of the previous study. These results highlight the effectiveness of the hybrid model in complex agricultural settings and its potential to improve potato plant disease detection and identification.
BioMed Central (BMC)
2025
Introduction
Per- and polyfluoroalkyl substances (PFAS) have been linked to effects on human lipid profiles, with several epidemiological studies reporting associations between specific PFAS and blood lipid concentrations. However, these associations have been inconsistent, and most studies have focused on cross-sectional analyses with limited repeated measurements.
Objective
In this study, we investigated associations between serum PFAS concentrations and major blood lipid classes over a 30-year period (1986–2016) and up to five time points. Lipids analyzed included total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TG).
Methods
This study included 145 participants from The Tromsø Study, Norway, who donated plasma samples three to five times over the study period. Linear mixed-effects (LME) models assessed longitudinal associations between PFAS and lipid classes, while multiple linear regression (MLR) models were used for cross-sectional associations.
Results
LME models demonstrated positive longitudinal associations between perfluorooctanoic acid (PFOA), perfluorononanoic acid (PFNA), perfluorodecanoic acid (PFDA), perfluoroundecanoic acid (PFUnDA), perfluorododecanoic acid (PFDoDA), and perfluorotridecanoic acid (PFTrDA) with TC. Additionally, PFOA, PFDA, PFUnDA, PFDoDA, and PFTrDA were associated with LDL-C, and PFUnDA and summed perfluorooctane sulfonate isomers (∑PFOS) with HDL-C. Cross-sectional analyses corroborated positive associations between the six PFAS compounds and TC at least three times, but the LDL-C and HDL-C associations were not confirmed. Summed perfluorooctane sulfonamide isomers (∑PFOSA) showed a negative association with LDL-C longitudinally, but this was not confirmed cross-sectionally. No associations were observed between PFAS and TG, longitudinally or cross-sectionally.
Conclusion
Concentrations of multiple PFAS were positively associated with blood lipids in longitudinal analyses, with the most consistent associations observed between six PFCA compounds and TC. These findings highlight the need for further investigation into these complex associations.
Elsevier
2025
Supervised Anomaly Detection in Univariate Time-Series Using 1D Convolutional Siamese Networks
In time-series data analysis, identifying anomalies is crucial for maintaining data integrity and ensuring accurate analyses and decision-making. Anomalies can compromise data quality and operational efficiency. The complexity of time-series data, with its temporal dependencies and potential non-stationarity, makes anomaly detection challenging but essential. Our research introduces ADSiamNet, a 1D Convolutional Neural Network-based Siamese network model for anomaly detection and rectification. ADSiamNet effectively identifies localized patterns in time-series data and smooths detected anomalies using a quantile-based technique. In tests with physical activity data from Actigraph watches and MOX2-5 sensors, ADSiamNet achieved accuracies of 98.65% and 85.0%, respectively, outperforming other supervised anomaly detection methods. The model uses a contrastive loss function to compare input sequences and adjusts network weights iteratively during training to recognize intricate patterns. Additionally, we evaluated various univariate time-series forecasting algorithms on datasets with and without anomalies. Results show that anomaly-smoothed data reduces forecasting errors, highlighting our approach’s effectiveness in enhancing time-series data analysis’s integrity and reliability. Future research will focus on multivariate time-series datasets.
IEEE (Institute of Electrical and Electronics Engineers)
2025
2025
Australia has significant sources of atmospheric methane (CH₄), driven by extensive coal and natural gas production, livestock, and large-scale fires. Accurate quantification and characterization of CH₄ emissions are critical for effective climate mitigation strategies in Australia. In this study, we employed an inverse analysis of atmospheric CH₄ observations from the GOSAT satellite and surface measurements from 2016 to 2021 to assess CH₄ emissions in Australia. The inversion process integrates anthropogenic and natural emissions as prior estimates, optimizing them with the NIES-TM-FLEXPART-variational model (NTFVAR) at a resolution of up to 0.1° × 0.1°. We validated the performance of our inverse model using data obtained from the United Nations Environment Program Methane Science (UNEP), Airborne Research Australia 2018 aircraft-based atmospheric CH₄ measurement campaigns. Compared to prior emission estimates, optimized emissions dramatically enhanced the accuracy of modeled concentrations, aligning them much better with observations. Our results indicate that the estimated inland CH4 emissions in Australia amount to 6.84 ± 0.51 Tg CH4 yr−1 and anthropogenic emissions amount to 4.20 ± 0.08 Tg CH4 yr−1, both slightly lower than the values reported in existing inventories. Moreover, our results unveil noteworthy spatiotemporal characteristics, such as upward corrections during the warm season, particularly in Southeastern Australia. During the three most severe months of the 2019–2020 bushfire season, emissions from biomass burning surged by 0.68 Tg, constituting over 71% of the total emission increase. These results highlight the importance of continuous observation and analysis of sectoral emissions, particularly near major sources, to guide targeted emission reduction strategies. The spatiotemporal characteristics identified in this study underscore the need for adaptive and region-specific approaches to CH₄ emission management in Australia.
2025
Sb-PiPLU: A Novel Parametric Activation Function for Deep Learning
The choice of activation function—particularly non-linear ones—plays a vital role in enhancing the classification performance of deep neural networks. In recent years, a variety of non-linear activation functions have been proposed. However, many of these suffer from drawbacks that limit the effectiveness of deep learning models. Common issues include the dying neuron problem, bias shift, gradient explosion, and vanishing gradients. To address these challenges, we introduce a new activation function: Softsign-based Piecewise Parametric Linear Unit (Sb-PiPLU). This function offers improved non-linear approximation capabilities for neural networks. Its piecewise, parametric design allows for greater adaptability and flexibility, which in turn enhances overall model performance. We evaluated Sb-PiPLU through a series of image classification experiments across various Convolutional Neural Network (CNN) architectures. Additionally, we assessed its memory usage and computational cost, demonstrating that Sb-PiPLU is both stable and efficient in practical applications. Our experimental results show that Sb-PiPLU consistently outperforms conventional activation functions in both classification accuracy and computational efficiency. It achieved higher accuracy on multiple benchmark datasets, including CIFAR-10, CINIC-10, MWD, Brain Tumor, and SVHN, surpassing widely-used functions such as ReLU and Tanh. Due to its flexibility and robustness, Sb-PiPLU is particularly well-suited for complex image classification tasks.
IEEE (Institute of Electrical and Electronics Engineers)
2025
More than 70 years of industrial production of per- and polyfluoroalkyl substances (PFAS) have resulted in their ubiquitous presence in the environment on a global scale, although differences in sources, transport and fate lead to variability of occurrence in the environment. Gull eggs are excellent bioindicators of environmental pollution, especially for persistent organic pollutants such as PFAS, known to bioaccumulate in organisms and to be deposited in bird eggs by maternal transfer. Using yellow-legged gull (Larus michahellis) eggs, we investigated the occurrence of more than 30 PFAS, including the most common chemicals (i.e., legacy PFAS) as well as their alternatives (i.e., emerging PFAS) in the Bay of Marseille, the second largest city in France. Compared to eggs from other colonies along the Mediterranean coast, those from Marseille had PFAS concentrations ranging from slightly higher to up to four times lower, suggesting that this area cannot be specifically identified as a hotspot for these compounds. We also found several emerging PFAS including 8:2 and 10:2 FTS, 7:3 FTCA or PFECHS in all collected eggs. Although the scarcity in toxicity thresholds for seabirds, especially during embryogenesis, does not enable any precise statement about the risks faced by this population, this study contributes to the effort in documenting legacy PFAS contamination on Mediterranean coasts while providing valuable novel inputs on PFAS of emerging concern. Identifying exposure in free-ranging species also participate to determine the main target for toxicity testing in wildlife.
Elsevier
2025
Genotoxicity assessment is essential for ensuring chemical safety and mitigating risks to human health and the environment. Traditional methods, reliant on animal models, are time-consuming, costly, and raise ethical concerns. New Approach Methods (NAMs) offer innovative, cost-effective, and ethical alternatives, playing a pivotal role in both traditional and next-generation risk assessment (NGRA) by minimizing the need for animal testing, particularly in genotoxicity evaluations. However, the development of NAMs often overlooks the particular physicochemical properties of nanomaterials (NMs), which significantly influence their toxicological behaviour and can interfere with genotoxicity evaluation. This underscores an urgent need for the standardization and adaptation of NAMs to address nano- and advanced material-specific genotoxicity challenges. In this review, we summarize the challenges associated with genotoxicity testing of NMs and highlight the suitability of existing in vitro and in silico NAMs for NMs and advanced materials, enabling genotoxicity testing across various exposure routes and organ systems. Despite considerable progress, regulatory validation remains constrained by the absence of approved test guidelines and standardized protocols. To achieve regulatory acceptance, it is crucial to adapt NAMs to NM-specific exposure scenarios, refine test systems to better mimic human biology, develop tailored in vitro protocols, and ensure thorough characterisation of NMs both in pristine form and dispersed in culture medium. Collaborative efforts among scientists, regulators, industry, and advocacy groups are vital to improving the reliability and regulatory acceptance of NAMs. By addressing these challenges, NAMs have the potential to revolutionize genotoxicity risk assessment, advancing it towards a more sustainable, efficient and ethical framework.
2025
At the same time Arctic ecosystems experiences rapid climate change, at a rate four times faster than the global average, they remain burdened by long-range transported pollution, notably with legacy polychlorinated biphenyls (PCBs). The present study investigates the potential impact of climate change on seabird exposure to PCB-153 using the established Nested Exposure Model (NEM), here expanded with three seabird species, i.e. common eider (Somateria mollissima), black-legged kittiwake (Rissa tridactyla) and glaucous gull (Larus hyperboreus), as well as the filter feeder blue mussel (Mytulis edulis). The model's performance was evaluated using empirical time trends of the seabird species in Kongsfjorden, Svalbard, and using tissue concentrations from filter feeders along the northern Norwegian coast. NEM successfully replicated empirical PCB-153 concentrations, confirming its ability to simulate PCB-153 bioaccumulation in the studied seabird species within an order of magnitude. Based on global PCB-153 emission estimates, simulations run until the year 2100 predicted seabird blood concentrations 99% lower than in year 2000. Model scenarios with climate change-induced altered dietary composition and lipid dynamics showed to have minimal impact on future PCB-153 exposure, compared to temporal changes in primary emissions of PCB-153. The present study suggests the potential of mechanistic modelling in assessing POP exposure in Arctic seabirds within a multiple stressor context.
Royal Society of Chemistry (RSC)
2025