Stochastic modeling and data analysis, which address uncertainty and complexity in business and industry, are important tools for optimizing decision-making processes. When it comes to providing access to healthy food, the importance of decision-making becomes even more prominent. In this work, the aim was to determine the criteria that healthy food businesses consider when choosing business-to-consumer (B2C) e-marketplaces and to evaluate the performance of alternative platforms. Nine main criteria were determined through a literature review and interviews with ten decision-makers (DMs), as sellers. To overcome the limitations of traditional fuzzy sets in handling uncertainty and inconsistent information, a hybrid decision-making method integrating interval-valued neutrosophic sets (IVNSs) with the stepwise weight assessment ratio analysis (SWARA) and evaluation based on distance from average solution (EDAS) methods was applied. Unlike existing approaches, this methodology better addresses expert uncertainty and inconsistency by simultaneously handling membership in truthiness, uncertainty, and falsity within an interval. Interval-valued neutrosophic (IVN)-SWARA was used to weight the selection criteria for e-marketplace platforms, and these weights were used to rank the platforms using IVN-EDAS. The findings show that integration capacity, site traffic density, and cost-effectiveness are the most important criteria in e-marketplace selection, while ease of membership and platform design usability were the least important. Theoretically, this research fills a gap in the literature by addressing B2C e-marketplace selection from the seller's perspective and better handles inconsistencies compared to standard fuzzy sets using IVN fuzzy sets. In addition, the proposed approach offers special solutions for healthy food businesses and sheds light on advanced research for other sectors and regions.

Workplace noise creates negative impacts on human health; therefore, preventive and control actions are compulsory, and sound absorption is one of the ways to prevent and control noise by minimizing noise exposure through the use of sound-absorbing materials. As such, an analysis dealing with the distribution of articles on sound absorption needs to be conducted as a basis for future researchers’ opportunities and interests. This study applied a bibliometric method, involving several stages: A Scopus data search stage using the keyword category “Sound and Absorption” (18,325), followed by a filtering stage for the year range 2015–2024 (8979), article type (5967), English language (5368), keyword “sound absorption” (1730), and journal type (document) (1720). The data analysis stage was conducted using the VOSviewer application (version 1.6.19) and Biblioshiny from Bibliometrix 3 with RStudio (version 4.3.1) to display bibliometric visualization maps. The results show sound absorption topics spreading in fields as material science (24.8%), physics (22.2%), engineering (22.2%), and environmental science (5.1%). China was the highest contributor, with more than 700 publications. There was an increasing trend in the number of article documents from 2015 (68) to 2024 (332). The journal with the most publications (219) and citations (4,606) was Applied Acoustics. The analysis found four clusters: Cluster 1 in red color (47 items), Cluster 2 in green color (46 items), Cluster 3 in blue color (15 items), and Cluster 4 in yellow color (10 items). Based on the analysis results, it can be concluded that sound absorption is still a current study trend that can be the basis for study opportunities in the topic of sound absorption for noise control.

Thermophilic anaerobic digestion often achieves higher biogas yields than mesophilic conditions, but its high energy demand limits long-term sustainability. As a result, many thermophilic digesters may transition toward mesophilic conditions. This study examined the effects of stepwise temperature shift from thermophilic to mesophilic ranges on biogas production and microbial community dynamics, while identifying factors that support or inhibit methane generation. At 55 °C and 50 °C, methane yields reached 342 and 311 mL CH₄/g COD, respectively, with a predominance of hydrogenotrophic Methanobacterium, alongside abundant Coprothermobacter, Defluviitoga, Acetomicrobium, Candidatus Bipolaricaulis, Brachyspira, and Dictyoglomus, suggesting potential syntrophic interactions between these microorganisms. At 45 °C and 40 °C, the methane yield declined to 234 and 219 mL CH₄/g COD but recovered to 289 and 251 mL CH₄/g COD at 35 °C and 30 °C, respectively. In the mesophilic range, frequent accumulation of acetate and butyrate coincided with a methanogenic shift from hydrogenotrophic Methanobacterium to acetoclastic Methanothrix, indicating a pathway transition. These results reveal that lowering the digestion temperature from thermophilic to mesophilic conditions enables recovery of acetogenic activity and shifts methanogenesis from hydrogenotrophic to acetoclastic pathways, with implications for optimizing biogas production during long-term operation.

In this review, we synthesized advancements in remote sensing and machine learning (ML) for wildfire risk assessment, early detection, and spread prediction across forests, grasslands, shrublands, and WUI ecosystems. We examined the capabilities and limitations of key satellite, UAV, and multisensor datasets, alongside the performance and transferability of ML and deep learning models. Major research gaps were identified in cross-ecosystem generalization, real-time data integration, model optimization, and uncertainty quantification. In the review, we contribute an ecosystem-inclusive comparison of RS–AI frameworks and outline future directions for operational, next-generation wildfire monitoring.

Water pollution represents a critical global challenge affecting human health and natural ecosystems. Among the most hazardous contaminants are heavy metals, particularly arsenic (As), which is frequently detected at elevated concentrations in water bodies as a consequence of anthropogenic activities such as agriculture, mining, and metallurgical processes. In this study, magnetic iron oxide nanoparticles synthesized via a green route using tangerine (Citrus reticulata) peel extract (FeNPs-CR) were developed and evaluated for As removal from aqueous solutions. The nanoparticles were synthesized through a coprecipitation method employing 75 g of pulverized peel extract and characterized by scanning electron microscopy (SEM), FT-IR spectroscopy, UV–Vis spectroscopy, and dynamic light scattering (DLS). SEM analysis revealed quasi-spherical particles with primary sizes ranging from 70 to 300 nm, while DLS measurements indicated larger hydrodynamic diameters due to particle agglomeration in aqueous media. Adsorption experiments were conducted using hydride generation atomic absorption spectroscopy, evaluating the effects of contact time, adsorbent dosage, pH, temperature, and initial As concentration. Removal efficiencies exceeding 99% were achieved at 20 mg/L As within 10–30 min, while concentrations of 50–120 mg/L showed removal efficiencies above 95%. Adsorption kinetics were best described by the pseudo-second-order model, indicating a physicochemical adsorption mechanism. Equilibrium data were better fitted by the Freundlich isotherm (R² = 0.97), suggesting heterogeneous multilayer adsorption, with a maximum adsorption capacity (qmax) of 227.27 mg/g. Thermodynamic analysis revealed negative Gibbs free energy (ΔG° = −7.66 to −11.14 kJ/mol), confirming the spontaneous and exothermic nature of the adsorption process. Optimal performance was observed under acidic conditions (pH 1–3) and temperatures below 35 °C. These findings demonstrated that FeNPs-CR synthesized from citrus waste constitute an efficient, sustainable, and low-cost adsorbent for As removal, with promising potential for large-scale water treatment applications.

Air pollution, specifically PM₁₀, is a critical challenge in Saudi Arabia, where levels often exceed World Health Organization (WHO) guidelines due to industrial activities and arid conditions, posing serious risks to human health. A key barrier is obtaining accurate PM₁₀ data, as estimations are limited by the few and unevenly distributed air quality stations. Notably, despite its severity, research on PM₁₀ estimation in the region remains scarce. Atmospheric reanalysis datasets like MERRA-2 offer complementary data, but their model-based nature, lacking actual measurements, introduces potential biases. To bridge this gap, this study developed a machine learning framework to estimate daily and monthly PM₁₀ concentrations in three climatically distinct Saudi cities. The framework integrates ground-based PM₁₀ data, meteorological parameters, and MERRA-2 reanalysis data. To our knowledge, this study represents the first application of MERRA-2 for PM₁₀ estimation in Saudi Arabia. The proposed AtmoStack is a stacked machine learning model, and we compared it against individual models (RF, HGB, CatBoost, and MLP) and state-of-the-art models, including LightGBM, ANN, and LSTM. Moreover, the framework incorporates feature-importance analysis to identify the most influential factors, helping to interpret the model. AtmoStack outperformed all baselines; in the dust-dominated environment of Buraidah, it achieved a daily R² of 0.73 and a monthly R² of 0.96. In Taif, it achieved a daily R² of 0.63 and a monthly R² of 0.94, indicating that AtmoStack effectively captures realistic distribution characteristics. These results support effective air-quality management and public health decisions.

Due to micro-deformation, wide spatial distribution, and high randomness of instability, stability assessment and early warning of slope rock mass remain challenging. To identify highly sensitive and robust indicators of rock mass damage, we investigated time–frequency dynamic differences between unstable rock mass and bedrock during failure, focusing on micro-vibration characteristics. We further examined their correlation with structural plane constraint strength. Our results showed that the degree of structural plane damage in unstable rock mass is positively correlated with the amplitude ratio between rock mass and bedrock, and negatively correlated with the frequency ratio. By integrating these dynamic indicators with classification algorithms, a dynamics-PSO-SVM-based stability analysis method for unstable rock mass was proposed. This model enabled rapid classification of rock mass stability states based on dynamic indicators. Laboratory-scale similarity experiments analyzed the evolution of amplitude and frequency ratios during rock mass instability, revealing stage-specific patterns in unstable rock mass. The method achieved 100% classification accuracy between stable and unstable states. Moreover, it effectively reduced interference from complex environmental excitation and equipment temperature drift in vibration monitoring data, demonstrating strong noise immunity. These findings enhance the practical applicability of dynamic evaluation methods for assessing the stability of unstable rock mass.

Flash floods are among the most critical natural hazards in arid regions, where limited infiltration capacity and intense episodic rainfall generate rapid runoff and severe flooding. Egypt’s Northwest Coast, particularly the Sallum–Alexandria corridor, faces growing flood risks under changing climatic conditions, posing challenges to sustainable water and land management. In this study, we present an integrated assessment of flood hazard severity and geomorphological susceptibility across seven major arid coastal basins (>500 km²) using an integrated framework of Geographic Information Systems (GIS), remote sensing (RS), and hydrological and hydraulic modeling. Basin morphometry was derived from SRTM-based DEMs and analyzed through WMS 11.0 to extract 45 parameters describing geometry, drainage, relief, and texture. Rainfall trend analysis (1979–2020) demonstrated a significant increase in precipitation intensity over the last decade, with HYFRAN-PLUS employed to estimate 100-year return periods. Hydrological simulations revealed Basin 2 (Wadi El-Harika) produced the highest peak discharge (800 m³/s), while Basins 6 and 7 generated minimal runoffs. Conversely, hydraulic modeling indicated that Basins 5, 6, and 7 experienced the greatest inundation depths (up to 30 m), highlighting a disconnect between runoff volume and flood severity due to local topographic and drainage conditions. The results emphasized the need for basin-specific analysis and demonstrated that flood hazard severity could not be inferred from runoff magnitude alone. The proposed integrated framework provided a robust physical basis for flood hazard and geomorphological susceptibility assessment in arid coastal basins and offers a foundation for future flood risk or vulnerability analyses through the incorporation of exposure and adaptive capacity indicators.