Agriculture is increasingly exposed to systemic risks driven by climate change, including extreme weather events, yield instability, and market fluctuations. Effective risk zoning is therefore crucial for actuarially fair and sustainable agricultural insurance. However, in multi-branch systems such as Türkiye’s Agricultural Insurance Pool (TARSIM), different insurance lines such as Crop, Greenhouse, Livestock, and Income Protection, employ distinct criteria and coding schemes, often resulting in inconsistent regional classifications. This study introduces a novel Machine Learning-Based Cross-Branch Consistency Framework to evaluate and harmonize agricultural risk zoning across these branches. Using official 2025 TARSIM datasets covering 71,902 localities, ordinal risk codes (A–Z) were encoded and analyzed through correlation statistics, clustering, and supervised learning. The results revealed strong interdependencies among high-intensity perils such as hail and storm across open-field and greenhouse insurance lines (r > 0.98), and high spatial synchrony among staple crop risks (r ≈ 0.99), confirming the existence of shared systemic exposure. K-Means clustering identified five statistically robust regional risk archetypes that transcend administrative boundaries and branch definitions. A novel Cross-Branch Consistency Index (CBCI) was developed to quantify inter-branch alignment, highlighting regions of both coherence and discrepancy in current zoning logic. Supervised learning validation using Random Forest achieved over 95% classification accuracy, demonstrating the reliability and reproducibility of the ML-derived zones. The findings support a transition from fragmented, expert-driven zoning to an integrated, data-driven risk management framework. The proposed methodology enhances actuarial fairness, transparency, and policy efficiency, providing a scalable foundation for modernizing state-supported agricultural insurance systems and advancing climate-resilient risk governance.

Hailstorms rank among the most destructive extreme weather events globally, causing substantial property damage. While limited case studies suggest that cities may exacerbate hailstorms, the underlying mechanisms remain uncertain because of the complex physical processes. Here, we examine a hailstorm formation pathway associated with convective merging process using long-term observational data and high-resolution numerical simulations. This pathway helps explain the rising frequency of hailstorms across two distinct climate regimes, North America and East Asia. We find that merger hailstorms (MHs) occur approximately twice as often and tend to be more intense than non-merging normal hailstorms (NHs), which have been traditionally considered as the primary hailstorm formation mode. Favorable environmental conditions support the initiation of multiple convective cells and their subsequent merging, a tendency that may be enhanced by anthropogenic heat in large cities. Projections from a machine-learning model indicate an increase in the MH frequency and a decrease in NH frequency in North America. Together, these findings highlight an underexplored hailstorm formation pathway and suggest that climate change and human activities may play a role in shaping future hailstorm characteristics and the associated risks.

Modeling heterogeneity in fault attribution of Pedestrian-Vehicle crashes using a Random parameter Binary Logit approach.

Unmanned aerial vehicles (UAVs) are a revolutionary new surveillance and transport technology with important implications for healthcare systems, particularly in the era of climate change. Rapid shifts in environmental systems are reshaping global climates. These changes have led to increasingly common extreme weather events that threaten population health. Mitigating the impacts of climate change on human health depends on our ability to predict, detect, and rapidly respond to changing ecosystem dynamics. The use of UAVs to tackle these new environmental health challenges is gaining momentum across multiple disciplines. This review identified four main areas where UAVs are being used or piloted to address climate change and health-related concerns: (1) Disease vector management, (2) environmental risk factors management, (3) environmental resource management, and (4) medical deliveries. Over the coming decades, UAVs are likely to play an increasing role in our efforts to keep pace with monitoring and mitigating the accelerating impacts of climate change on human health.

Climate change threatens maternal, and newborn health, particularly by exacerbating climate-sensitive diseases like malaria. Malaria in pregnancy (MiP) contributes to maternal anaemia, stillbirth, preterm delivery, and low birth weight. In Kenya's Lake Victoria basin, recurrent floods and droughts disrupt antenatal care (ANC), the main delivery platform for intermittent preventive treatment of MiP with sulfadoxine-pyrimethamine (IPTp-SP). The extent to which these extreme weather events affect IPTp-SP uptake through ANC attendance remains unexplored. Data are drawn from a cross-sectional household survey conducted under the revive IPTp-SP project among women who had given birth to a live baby in the last 24months preceding the interview in malaria-endemic counties of Kisumu and Migori. Exposure was self-reported extreme weather events in the past 12months; the outcome was completion of three or more IPTp-SP doses (IPTp₃ +); and the mediator was attendance of ≥ 4 ANC visits (ANC4+). Using a counterfactual mediation framework, we decomposed the total effect of climate shock on IPTp₃ + into natural direct and indirect effects via ANC4+, adjusting for covariates. Models accounted for the complex survey design, with bootstrapping to estimate 95% confidence intervals (CI) for indirect effects. IPTp₃ + coverage was lower among shock-exposed pregnancies (49.5%) versus unexposed (58.8%). Extreme weather events were associated with reduction in ANC4+ attendance (-0.23, 95% CI -0.41 to -0.05). ANC4+ completion strongly predicted IPTp₃ + uptake (1.71, 95% CI 1.42-2.00). The total effect of shocks on IPTp₃ + was -0.45 (95% CI -0.68 to -0.23), of which 76.6% was mediated through ANC. After adjustment, the total effect attenuated (-0.27, 95% CI -0.56 to 0.02) and was no longer significant. Effects were stronger in flood-prone Kisumu than in Migori. Extreme weather events reduced IPTp₃ + uptake primarily through ANC disruption in unadjusted models, with effects concentrated in Kisumu. However, these associations attenuated after adjustment, suggesting the role of underlying socioeconomic and contextual vulnerabilities.

ABSTRACT Existing studies on risk in adventure often use fixed lists of risk categories and fail to consider the layered nature of how tourists perceive these risks. Using the extreme setting of cycling on the Sichuan-Tibet Highway as a case, this study employs a mixed-methods design that integrates Reflexive Thematic Analysis with a questionnaire survey to construct a two-dimensional Risk Perception Matrix (RPM). The study first identifies four primary risk themes: environmental, physical, planning, and accidental risks. Quantitative analysis further reveals that Adverse Weather ranks highest in probability, severity, and overall risk perception scores. The RPM displays four risk quadrants, where the core risk zone reflects a strong coupling between the environment and the body, while High Difficulty of Rescue is identified as a critical, high-impact, low-probability risk. By shifting from a one-dimensional checklist to a multidimensional matrix-based analysis, this study offers a more accurate framework for managing risks at tourism destinations.

Natural image quality is often degraded by adverse weather conditions, significantly impairing the performance of downstream tasks. Image restoration has emerged as a core solution to this challenge and has been widely discussed in the literature. Although recent transformer-based approaches have made remarkable progress in image restoration, their increasing system complexity poses significant challenges for real-time processing, particularly in real-world deployment scenarios. To this end, most existing methods attempt to simplify the self-attention mechanism, such as by channel self-attention or state space model. However, these methods primarily focus on network architecture while neglecting the inherent characteristics of image restoration itself. In this context, we explore a pyramid Wavelet-Fourier iterative pipeline to demonstrate the potential of Wavelet-Fourier processing for image restoration. Inspired by the above findings, we propose a novel and efficient restoration baseline, named Pyramid Wavelet-Fourier Network (PW-FNet). Specifically, PW-FNet features two key design principles: 1) at the inter-block level, integrates a pyramid wavelet-based multi-input multi-output structure to achieve multi-scale and multi-frequency bands decomposition; and 2) at the intra-block level, incorporates Fourier transforms as an efficient alternative to self-attention mechanisms, effectively reducing computational complexity while preserving global modeling capability. Extensive experiments on tasks such as image deraining, raindrop removal, image super-resolution, motion deblurring, image dehazing, image desnowing and underwater/low-light enhancement demonstrate that PW-FNet not only surpasses state-of-the-art methods in restoration quality but also achieves superior efficiency, with significantly reduced parameter size, computational cost and inference time. The code is available at: https://github.com/deng-ai-lab/PW-FNet.

Climate change poses significant challenges to global agriculture, particularly to horticulture, which is highly sensitive to climatic variations. Rising temperatures, changed precipitation patterns, and an increase in the frequency of extreme weather events highlight how urgent it is to implement climate-smart techniques since they all represent serious risks to conventional horticulture systems. Horticultural crops—fruits, vegetables, spices, ornamentals, and plantation crops—are uniquely vulnerable to climate change because quality traits and reproductive success are exquisitely sensitive to heat waves, droughts, salinity, flooding, and compound extremes. This review synthesizes recent advances (primarily 2021–2025) in the physiology and molecular biology of stress perception, signalling, and acclimation in horticultural species, with emphasis on traits that safeguard yield and quality. We summarise progress in understanding combinatorial stress responses, integrative hormone crosstalk, ROS–Ca²⁺ signalling, osmotic regulation, and source–sink balance under heat and water deficits. We then examine emerging levers for adaptation: epigenetic and priming-based memory, microbiome mediation, and organ-specific mechanisms in flowers, fruits, and storage tissues. A second focus covers practical strategies—from climate-smart rootstocks and grafting, CRISPR-enabled breeding, seed/foliar priming (e.g., melatonin, seaweed biostimulants), silicon nutrition, to protected cultivation and digital phenotyping—highlighting trade-offs and translational pathways. We close by outlining evidence gaps, research priorities, and policy-relevant directions to accelerate climate resilience in high-value horticulture.

Extreme weather events are increasing in frequency and intensity, but their ecological impacts remain less well understood than those of gradual climate change, largely owing to the challenge of studying unpredictable, short-lived events. The 2021 western North American heatwave is among the most extreme on record globally, yet a broad assessment of its ecological consequences is lacking. Here we synthesize meteorological, ecological, hydrological and wildfire data, along with process-based modelling, to quantify the heatwave and its impacts across the region. Our meta-analysis of 32 terrestrial and marine taxa reveals that over 75% were negatively impacted, but species responses ranged widely, from 99% declines to 89% increases. This variability reflects differences in organisms' thermal sensitivities, response capacities and exposures, with the latter dependent on geography, microclimate and refugia. Impacts tended to be greater for sessile marine invertebrates, algae and plants than for birds and mammals. At the ecosystem scale, changes in gross primary productivity ranged from 30% increases in cooler, wetter areas to 75% decreases in warmer, arid ones. Streamflow from snow and ice melt increased 40% during the heatwave before dropping below average, whereas wildfire activity surged 37% during the heatwave and 395% the following week. Our results underscore the urgent need for enhanced coordinated approaches to predict, detect and manage increasing heatwaves.

The nowcasting of extreme rainfall poses significant daily challenges on a global scale, especially in vulnerable regions of the Global South. Conventional Numerical Weather Prediction models often fail to deliver accurate and timely forecasts for extreme weather events, exacerbating socioeconomic inequalities and increasing climate vulnerability. Deep learning approaches present a promising opportunity to uncover more precise predictive patterns; however, their application remains constrained by the high computational costs associated to their large parameter spaces. This study evaluates the effectiveness of the MS-RNN framework for improving computational efficiency and predictive accuracy in extreme precipitation nowcasting, using real weather radar data from the TAASRAD19 and Rio de Janeiro datasets. While the framework has been extensively validated both theoretically and experimentally in other scenarios, this work examines its application to real radar data. Metrics related to sustainability, such as energy consumption, [Formula: see text] emissions, and water usage, have not been calculated in this specific context and are rarely addressed in current literature. Our findings demonstrate the potential of the solution to enhance computational efficiency maintaining predictive performance when applied to real weather radar data, supporting sustainable and accessible AI solutions for climate resilience in resource-limited regions.

ABSTRACT In the wake of extreme weather shocks, do firms “walk the walk” or merely “talk the talk” in their climate strategies? Drawing on legitimacy theory, this study empirically examines firms' climate strategy choices in response to extreme weather events using panel data from Chinese A‐share listed firms from 2012 to 2022. The results show that extreme weather events significantly increase firms' climate‐related discourse but do not lead to corresponding adaptation actions, suggesting a tendency toward symbolic strategies—“talking the talk” without “walking the walk.” These strategies are primarily motivated by institutional pressures and securing credit resources, as firms seek to gain legitimacy through low‐cost disclosures while alleviating financial constraints. Furthermore, firms with older top management teams and lower risk tolerance are more inclined to adopt symbolic strategies. Crucially, such symbolic climate approaches are found to hinder firms' long‐term development. Overall, this study advances understanding of corporate climate strategy choices under extreme weather shocks and offers managerial and policy implications for fostering substantive climate strategies in emerging economies.

Substantial increases in the likelihood of extreme fire weather events for fire-prone ecosystems in Australia

Against a backdrop of increasingly extreme weather events, including heavy rainfall, prolonged droughts and wildfires, this study provides a comprehensive analysis of the vulnerability of riverside communities in the northern part of the Mono Basin (Togo). Its originality lies in the joint use of quantitative and qualitative approaches, based on a field survey conducted from January to March 2024 among 149 randomly selected individuals. The semi-structured interviews conducted via KoboCollect involved a variety of stakeholders (farmers, community leaders, local elected officials, NGOs/CSOs, municipal services and ANPC), allowing for a comparison of local perceptions, socio-economic realities and environmental risks. The data were analysed using IBM SPSS Statistics 2.0 and R version 4.4.3 softwares. The results confirm a high vulnerability to flooding, drought and vegetation fires, with marked variations between localities. 𝜒2 tests reveal highly significant (p 0.50) associations, highlighting territorial heterogeneity. The populations attribute these hazards to seasonal disruption and global warming. The impacts identified are multiple: animal losses, crop destruction, water and fodder shortages, disease and injury during periods of flooding, but also famine, poverty, inactivity and economic losses due to droughts and fires. Public infrastructure appears to be particularly vulnerable, especially during floods. Finally, the study showed a hierarchy of impacts, focusing on the most visible and immediate effects, whether socio-economic or environmental.

Introduction: Mosquito-borne diseases (MBD) represent a global public health concern. Many mosquito species have rapidly expanded globally due to climate change and are expected to continue spreading beyond their current range into temperate regions. To support policy action, an umbrella review was conducted to summarize the growing literature on the impact of climate change indicators on MBD patterns in temperate zones. Methods: Studies published until December 31st, 2023, were search in PubMed, EMBASE, Cochrane, Epistemonikos, and Web of Science Core Collection. The Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) and Problem/Population, Intervention, Comparison, and Outcome (PICO) guidelines were used. The quality of the methodology and of the evidence of the included reviews were assessed using ‘A MeaSurement Tool to Assess Systematic Reviews 2’ (AMSTAR 2). Results: The initial search yielded 6518 studies, with 78 undergoing full-text assessment. Ten studies met the inclusion criteria. Key findings include a significant association between climate factors (specifically temperature, rainfall and humidity) and MBD in temperate regions, mostly malaria and dengue, with temperature consistently showing a strong predictive value. Our findings are likely to be robust as we employed strict quality criteria to ensure the quality of included primary studies and systematic reviews. Discussion and Conclusions: This umbrella review identifies concerning impacts of climate change on MBD in temperate regions, highlighting significant correlations between climate variables and diseases such as dengue, malaria, and the Ross River virus. The review underscores the importance of targeted public health strategies that integrate climatic data for effective management of MBD in temperate regions and calls for further research on extreme weather events and less-studied diseases.

To address the significant operational disruptions caused by inclement weather in maritime logistics, this study investigates the integrated rescheduling optimization of vessels and tugboats within one-way channel ports. The research aims to minimize total operational costs, including dispatching and delay penalties, by synchronizing vessel movements with tugboat service capabilities under uncertain conditions. Methodologically, a rolling horizon decision-making mechanism is proposed to accommodate dynamic operational scenarios driven by fluctuating weather. On this basis, an integrated rescheduling model is developed to address the compounded challenges of navigation rule changes, channel closures, vessel delays, and additional shifting tasks. The model explicitly incorporates critical constraints such as channel navigation protocols, tugboat availability, power capacity limits, and tidal windows for deep-draft vessels. To achieve efficient solution generation, an improved Variable Neighborhood Search (VNS) algorithm is designed to effectively handle the problem’s complexity. Experimental results validate the effectiveness of the proposed approach and the robustness of the algorithm in diverse disruption scenarios. Furthermore, sensitivity analyses reveal how channel closure duration, vessel delay intensities, and the volume of shifting tasks quantitatively influence rescheduling outcomes. This study contributes a novel synergistic optimization framework that enhances the operational resilience and decision-making capabilities of port authorities.

In rainy and foggy conditions, the scattering of light and the occlusion effects of atmospheric particles distort the reflected light from object surfaces, leading to inconsistent depth information. As a result, depth estimation models trained under clear weather conditions fail to generalize effectively to adverse weather conditions. To address this challenge, we propose a novel CNN-Transformer architecture, WeatherMono, for self-supervised monocular depth estimation under rainy and foggy weather. Rainy and foggy images often contain large regions of low contrast and blurry features. By combining Convolutional Neural Networks (CNNs) with Transformers, WeatherMono effectively captures both local and global contextual information, thus improving depth estimation accuracy. Specifically, we introduce a Multi-Scale Deformable Convolution (MDC) module and a Global-Local Feature Interaction (GLFI) module. The MDC module extracts detailed local features in rainy and foggy environments, while the GLFI module incorporates an efficient multi-head attention mechanism into the Transformer encoder, enabling more effective capture of both local and global information. This enhances the model's ability to comprehend image features, strengthens its capability to handle low-contrast and blurry images, and ultimately improves the accuracy of depth estimation in adverse weather conditions. Experiments on WeatherKITTI show WeatherMono achieves AbsRel of 0.097, outperforming WeatherDepth (0.104) and RoboDepth (0.107). On DrivingStereo, it achieves AbsRel of 0.149 (rain) and 0.101 (fog). Extensive qualitative and quantitative experiments demonstrate that WeatherMono significantly outperforms existing methods in terms of both accuracy and robustness under rainy and foggy conditions.

The frequency of extreme weather events has become higher, and electricity consumption has also become more complex. These changes increase the risk of overload in distribution transformers (DTs), and this risk threatens the stability and reliability of the power grid. Existing methods have significant limitations. Traditional static threshold methods (based on DGA gas ratios and electrical signal thresholds) fail to consider temporal changes and complex links between factors, while modern machine learning models lack cause–effect relationships over time and clear ways to describe uncertainty. With such motivations, this paper proposes a causal-enhanced hybrid framework, which combines Long Short-Term Memory (LSTM) networks and Random Forest (RF) algorithms. The framework uses causal Seasonal Trend decomposition using Loess (STL) to reveal load patterns at different time scales. The mutual information index and spatiotemporal graph convolutional network (ST-GCN) are used to explore nonlinear relations and reveal how temperature affects load changes. The LSTM model captures time dependence in load series, and the Bayesian optimized Random Forest is used to solve the problem of data imbalance and quantify uncertainty. In addition, the framework constructs an early warning system that combines data from many sources in real time. Test results show that the proposed algorithm exhibits excellent performance in multi-source data environments.

UAVs are widely used for all-weather, round-the-clock security inspections in urban and industrial areas. However, pure visible-light systems fail at night or in adverse weather conditions, while pure infrared methods are limited by thermal noise, low spatial resolutions, and high false alarm rates. Multispectral images render the task of object detection highly reliable and robust by providing complementary target feature information. This study suggests a frequency-based cross-attention transformer (FCAT) for multispectral object detection as a solution to this issue. This approach collects cross-modal complementary characteristics, effectively learns and integrates global contextual information via the cross-attention mechanism, and greatly increases multispectral object detection accuracy. At the same time, spatial-domain features are mapped to the frequency domain via the Fourier transform, and the scaled dot product attention is estimated via element-wise product operations, which break through the limitation of traditional spatial-domain matrix multiplication and effectively reduce the computational cost of the model. Additionally, this study independently builds a multi-scene multi-time climate visible–infrared dataset (OPVM-VIRD), which contains 20,025 target instances, to address the issue of the lack of all-weather cross-spectral data in object detection tasks from the perspective of UAVs. Experimental findings from the OPVM-VIRD, M3FD, and FLIR datasets demonstrate that our proposed approach outperforms prevailing state-of-the-art multispectral object detection algorithms on public benchmarks, while the FCAT model achieves an mAP50 score of 94.7% on our custom-built dataset—10.8% higher than ICAF. At the same time, the number of FCAT parameters is 85.26 M, which is significantly lower than that of mainstream models, such as ICAF. Therefore, the FCAT is a change detection strategy with strong model generalization abilities, and it has important application value in the all-day and all-weather security patrol of cities and industrial parks carried out by UAVs.

The increasing risk to Vulnerable Road Users (VRUs) at urban intersections necessitates advanced safety mechanisms capable of operating effectively under diverse conditions, including adverse weather like heavy rain. While optical sensors such as cameras and LiDAR often degrade in poor visibility, Radio Frequency (RF)-based systems offer resilient, all-weather tracking. This paper presents a novel approach to enhancing VRU protection by fusing two RF modalities: radar sensors and Ultra-Wideband (UWB) technology, a strong candidate for Joint Communication and Sensing (JCS). The research, conducted as part of the VIDETEC-2 project, addresses the limitations of existing vehicle-based and infrastructure-based systems, particularly in scenarios involving occlusions and blind spots. By leveraging radar's environmental robustness alongside UWB's precise, cost-effective short-range communication and localization, the proposed system delivers the framework for continuous vehicle and VRU tracking. The fusion of these sensor modalities, managed through a hybrid Kalman filter approach integrating an Unscented Kalman Filter (UKF) and an Extended Kalman Filter (EKF), allows reliable VRU tracking even in challenging urban scenarios. The experimental results demonstrate a reduction in tracking uncertainty and highlight the system's potential to serve as a more accurate and responsive safety mechanism for VRUs at intersections. This work contributes to the development of intelligent road infrastructures, laying the foundation for future advancements in urban traffic safety.

4D millimeter-wave radar provides a promising solution for robust perception in adverse weather. Existing detectors still struggle with sparse and noisy point clouds, and maintaining real-time inference while achieving competitive accuracy remains challenging. We propose SGE-Flow, a streamlined PointPillars-based 4D radar 3D detector that embeds lightweight spatiotemporal geometric enhancements into the voxelization front-end. Velocity Displacement Compensation (VDC) leverages compensated radial velocity to align accumulated points in physical space and improve geometric consistency. Distribution-Aware Density (DAD) enables fast density feature extraction by estimating per-pillar density from simple statistical moments, which also restores vertical distribution cues lost during pillarization. To compensate for the absence of tangential velocity measurements, a Transformer-based Inter-frame Flow (IFF) module infers latent motion from frame-to-frame pillar occupancy changes. Evaluations on the View-of-Delft (VoD) dataset show that SGE-Flow achieves 53.23% 3D mean Average Precision (mAP) while running at 72 frames per second (FPS) on an NVIDIA RTX 3090. The proposed modules are plug-and-play and can also improve strong baselines such as MAFF-Net.