Increased agricultural production could improve household income but often generates adverse environmental impacts, including soil degradation, rising temperatures, and drought, thereby contributing to climate change. This study aims to optimize income and carbon emissions in the trade of rice, corn, and cattle commodities in the Indonesia–Timor Leste border region and to assess the performance of integrated sustainable trade among farmers, traders, and processing industries. A Multi-Objective Linear Programming (MOLP) model and Partial Least Squares Structural Equation Modeling (PLS-SEM) were employed for analysis. The findings indicated that increased trade activities could improve economic outcomes while maintaining emissions within manageable limits. Farmer income is projected to increase by IDR 5.779 billion per production season, with improved cost efficiency at approximately IDR 64,000 per acre and maximum emissions of 356,561 tons CO₂e. Traders’ income is expected to increase by IDR 8.526 billion, with maximum emissions of 2,443.241 tons CO₂e and average transport costs of IDR 4,600 per kilometer. Carbon emissions at the farm level primarily stem from inefficient use of fertilizer and land burning, while emissions at the trader level are driven by transport capacity and travel distance. Although processing industries have established direct relationships with farmers, most farmers remain dependent on traders for market access. Strengthening the capacity of processing industries in the border region is therefore considered essential for maximizing farmers’ income.
With the rapid expansion of high-speed railway networks and the continuous growth in urban travel demand, the efficiency of first/last-mile connections at transport hubs has become a critical factor constraining the performance of integrated transportation systems. Demand-responsive customized bus services dynamically match passenger demand with available capacity, providing a feasible solution for improving travel flexibility. However, in practical applications, the rational design of customized bus networks remains challenging due to heterogeneous spatial demand distributions, vehicle capacity limitations, and various operational constraints. This study proposes an integrated methodological framework for customized bus network design that combines three key components: stop identification, route optimization, and simulation-based validation. First, a hybrid clustering approach that integrates density-based clustering with centroid partitioning is employed to extract potential stop locations from passenger origin–destination data. A capacity-constrained mechanism is further introduced to regulate clustering results, ensuring that stop sizes are compatible with vehicle carrying capacity. Based on the identified stops, the network design problem is formulated as a vehicle routing problem with time window constraints, where operational cost, passenger travel time cost, and environmental impact are jointly considered as optimization objectives. A genetic algorithm is adopted to solve the model. A case study involving a feeder service between a high-speed rail station and the urban core business district is conducted, and the proposed framework is validated through simulation using the AnyLogic platform. The results demonstrate that the proposed method improves vehicle utilization and route efficiency while maintaining service quality and system stability. This research provides a practical technical pathway and decision support for the intelligent design and operation of demand-responsive customized bus services.
The physical quality of seeds is a critical determinant of sorting efficiency and crop productivity, yet conventional assessment approaches are often labor-intensive, invasive, and time-consuming. To address these limitations, computer vision-based methods have been increasingly adopted; however, most existing techniques rely primarily on reflected visible light, thereby capturing only surface-level features and limiting the detection of internal defects. In this study, a low-cost imaging system integrating both reflection and transmission of visible light was developed to enhance the characterization of maize seed translucency. By enabling simultaneous acquisition of information from the two principal faces of white maize seeds, a more comprehensive representation of both external morphology and internal structural variations was achieved. A comparative analysis was conducted between the conventional reflection-based method and the proposed imaging approach, with correlation coefficients between seed faces determined as 0.62 and 0.84, respectively, indicating a substantial improvement in feature consistency and information richness. A dedicated dataset was subsequently constructed using both imaging techniques and employed to train a YOLOv5s-based detection model over 200 epochs. The classification performance demonstrated a marked enhancement, with the proposed method achieving an accuracy of 93.07%, compared to 81.5% obtained using the conventional approach. Furthermore, real-time detection capability was validated through the implementation of the optimized imaging system, in which improved inference stability and robustness were achieved under practical operating conditions. The results indicate that the integration of transmission with reflection imaging provides a cost-effective and reliable solution for non-destructive seed quality assessment, offering significant potential for scalable deployment in agricultural sorting systems.
The threat posed by volcanic eruptions necessitates ongoing monitoring to assess their status. Mount Semeru is one of the active volcanoes located on the island of Java. Observations are made using remote sensing, utilizing data from the Copernicus satellite Sentinel-1 Single Look Complex (SLC) to track changes in Differential Interferometric Synthetic Aperture Radar (DInSAR) deformation, and Sentinel-3 satellite sea and land surface temperature radiometer (SLTR) to observe ground surface temperature variations due to the eruption of Mount Semeru that occurred in 2022, before, during, and after the event. The DInSAR deformation recorded before the eruption ranged from -0.025 cm to -0.054 cm on the scale bar, while the land surface temperature (LST) before the eruption was at a minimum of 18.6 ℃ and a maximum of 27.8 ℃. during the eruption, DInSAR deformation changes showed inflation, with values reaching from 0.015 cm to 0.3 cm on the scale bar, and the LST also rose, peaking at 36.3 ℃. after the eruption, DInSAR deformation changes indicated deflation, with measurements between 0.049 cm and 0.099 cm on the scale bar, and the temperature trend also fell, with the highest temperature observed being 33.6 ℃.
Adaptive multi-scale representation learning has become a fundamental component of modern image processing systems. However, existing fusion strategies often treat features extracted from different scales equally, resulting in suboptimal performance under uncertain conditions such as noise, blur, and low contrast. To address this limitation, this paper proposes an uncertainty-aware deep feature fusion framework for adaptive multi-scale image processing. The proposed framework decomposes input images into multiple scales using wavelet-based or Laplacian pyramid representations to capture complementary spatial-frequency information. Discriminative features are extracted at each scale using lightweight Convolutional Neural Networks (CNNs) or Vision Transformer (ViT) encoders. To estimate feature reliability, Bayesian deep learning with Monte Carlo (MC) dropout is employed to model uncertainty at the feature level. A principled uncertainty-aware fusion mechanism is then introduced to dynamically combine multi-scale features according to their estimated reliability. As a result, reliable features contribute more significantly to the fused representation, while uncertain features are suppressed. The fused representation is subsequently utilized in task-specific heads for image restoration, classification, and segmentation. Extensive experiments conducted under multiple degradation conditions demonstrate that the proposed framework consistently outperforms traditional fusion and attention-based methods in terms of Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and Fréchet Inception Distance (FID). The results further confirm the robustness and generalization capability of the proposed uncertainty-aware multi-scale fusion strategy in adverse imaging environments.
This study examined the mediating role of policy in the relationship between improvement in construction waste management and effectiveness of construction and demolition waste (C&DW) management in Malang City, Indonesia. Using Partial Least Squares Structural Equation Modeling (PLS-SEM) with data from 48 construction companies, this research investigated how policy frameworks translated improvement initiatives into sustainable waste management outcomes. Improvement in construction waste management was measured by four phases (planning, design, construction, and operation), policy through four instruments (regulation, incentive, reward-sanction, and standard), and C&DW management through triple bottom line dimensions (economic, social, and environmental). Results revealed that improvement in construction waste management significantly influenced policy formulation (β = 0.761, p < .001, f² = 1.374), and that policy substantially affected the effectiveness of C&DW management (β = 0.692, p < .001, f² = 0.950). However, the direct effect of improvement in C&DW management was not significant (β = 0.240, p = .249), indicating full mediation through policy (β = 0.526, p < .001, υ = 0.077). The model explained 78.8% variance in C&DW management with strong predictive relevance (Q² = 0.711). These findings demonstrated that improvement efforts should be channeled through robust policy frameworks to achieve systemic waste management transformation, thus highlighting the critical role of integrated policy instruments in translating operational improvements into sustainable outcomes in developing urban contexts.
The banking sector is experiencing a substantial transformation driven by digitalization, evolving customer expectations, and increasing competitive pressure. In hybrid banking environments, where customers interact through both digital and in-branch channels, customer experience and trust have become critical factors shaping managerial and customer decision processes. Although prior research has extensively examined the relationship between customer experience and behavioral intention, trust has predominantly been conceptualized as a mediating mechanism, while its moderating role in hybrid banking contexts remains insufficiently explored. This study investigates the influence of customer experience on trust and purchase intention, with particular emphasis on the moderating effect of trust in hybrid banking decision environments. A quantitative online survey was conducted among 371 bank customers in Germany. The collected data were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The results showed that customer experience exerted a strong positive effect on trust ($\beta$ = 0.858) and a significant direct effect on purchase intention ($\beta$ = 0.369). Trust also demonstrated a significant positive influence on purchase intention ($\beta$ = 0.370) and significantly strengthened the relationship between customer experience and purchase intention through its moderating effect ($\beta$ = 0.097). The model explained a substantial proportion of variance in trust ($R^2$ = 0.737) and a moderate proportion in purchase intention ($R^2$ = 0.454). The findings indicate that trust functions not only as a direct relational mechanism but also as a contextual condition influencing how customer experience translates into behavioral intention in hybrid banking settings. This study provides a more differentiated understanding of customer decision behavior in digitally integrated banking environments and offers practical implications for customer experience management and trust-oriented decision strategies in the financial services sector.
Wetlands are fundamental habitats for migratory birds and species habiting shallow waters. In this study, we quantitatively analyze the surface water area of fluctuations in the Al-Hawizeh Marshes, a transboundary wetland shared by Iraq and Iran. Following a severe drought in the past decade, these marshes have shown ecological recovery, positioning them today as a sustainable ecosystem. The study examines whether these marshes are once again facing the risk of drought or will continue along a trajectory of ecological conservation. This study employs Landsat satellite imagery spanning nearly a decade to monitor hydrologic dynamics for the 2015, 2018, 2021 and 2024 calendar years by relying on the computational capabilities of Google Earth Engine (GEE) platform. In parallel, the normalised difference water index (NDWI) was applied to delineate water bodies and quantify the spatial extent of surface water. The year 2021 proved to be the most anomalous in terms of water area, presenting an average of 448.4 km$^2$, in sharp contrast with the severe desiccation monitored over the years, including 2018 (48.4 km$^2$) and 2024 (49.6 km$^2$). The results demonstrate the utility of remote sensing for monitoring these largely inaccessible wetlands and provide vital, data-driven evidence of the critically endangered status of Al-Hawizeh Marshes. This article attains particular importance not only through its spatial analysis and statistical evaluation, employing the correlation coefficient matrix (CCM) and heatmap analysis effectively illustrating the fluctuations revealed across monthly and annual classifications, but also through the results it presents, which indicate that the shallow water bodies are undergoing a gradual recession and are generally progressing toward desiccation. Accordingly, the findings call for rapid solutions in the form of watershed-based transboundary water management agreements, along with a deeper exploration of the drivers behind such extreme hydrological regime shifts, in order to support decision-maker planning in conserving this ecologically rich corner of the world. This approach aims to ensure the continuity of the ecological environment, safeguard the local community, and ultimately achieve sustainability.
This study examines the potential waste generation from medium-scale fishing vessels (30–100 GT) operating at the Nizam Zachman Ocean Fishing Port (PPSNZ) in Jakarta Bay and analyzes existing practices and regulatory gaps in marine waste management. The results indicate that provisioning activities are the primary source of plastic-based waste, including wrappers, bottles, and containers. The findings revealed that most vessels lacked onboard waste-handling systems and failed to return waste to port facilities, thereby contributing to unmonitored marine debris in coastal waters. Moreover, the regulatory framework for vessel waste management in Indonesia was fragmented and did not adequately address the operations of medium-scale vessels. Inadequate infrastructure, limited enforcement capacity, and low environmental awareness among crew members further hindered compliance. This study highlights the urgent need for vessel-specific waste return policies, the adoption of digital reporting systems, and the provision of adequate port reception facilities. It also emphasizes the importance of incentive-based compliance mechanisms, such as reduced port fees for vessels that return waste, and underscores the broader need to strengthen port governance in order to support a more inclusive marine waste management system aligned with Sustainable Development Goal (SDG 14).
This study provides a comprehensive assessment of air pollution levels in the industrial areas of the Samarkand region, one of the most economically developed territories of Uzbekistan. Using regional industrial statistics, emission inventories, and enterprise-level environmental data, the research identifies the spatial distribution, composition, and intensity of atmospheric pollutants across major industrial zones. The analysis demonstrates that the Samarkand region hosts more than 5,400 environmentally significant facilities, including 171 high-hazard (Category I) enterprises, which collectively shape the regional air quality profile. Emission data from key industrial enterprises—such as “Azia Metall Prof,” Henguan Cement LLC, Jomboy Green Lights LLC, and Urgut Textile Shifer LLC—reveal substantial releases of nitrogen oxides, carbon monoxide, sulfur dioxide, cement and inorganic dust, hydrocarbons, and carcinogenic compounds such as benz(a)pyrene. Among these, nitrogen oxide and carbon monoxide dominate emissions from metallurgical production, while cement plants contribute significantly to dust, sulfur oxides, and carbon dioxide. Temporal analysis shows persistently high emissions in Samarkand city and Kattakurgan district, with slight reductions in recent years linked to industrial relocation and expansion of green zones. The findings highlight considerable environmental risks, including deteriorating air quality, increased respiratory hazards, and potential long-term ecological impacts. The study underscores the need for strengthened emission control technologies, expansion of monitoring networks, and improved regulatory enforcement. These results contribute new empirical evidence for environmental policy, urban planning, and public health management in rapidly industrializing regions of Central Asia.
Effective management of freshwater resources in agriculture is essential for ensuring sustainable economic development and environmental resilience, particularly in transitional economies such as Azerbaijan. Over the period 2000–2021, agricultural land area in Azerbaijan exhibited a steady increase, while the sector’s contribution to GDP declined, indicating structural transformation and potential inefficiencies in resource utilization. This study investigates the nonlinear effects of agricultural land use and agricultural value added on freshwater withdrawals using an interpretable machine learning framework. Specifically, Extreme Gradient Boosting (XGBoost) is employed to model complex relationships, while Shapley Additive Explanations (SHAP) quantify feature importance and elucidate threshold and asymmetric effects. The analysis draws on annual country-level data integrating national and international statistics to ensure temporal consistency and comparability. Results indicate that agricultural land area constitutes the dominant driver of freshwater withdrawals, contributing 57% of the model’s predictive gain, whereas agricultural value added accounts for 43%. SHAP dependence plots reveal pronounced nonlinearities: moderate land expansion exacerbates freshwater stress, whereas allocations beyond a critical threshold mitigate pressure, reflecting potential efficiency gains at scale. Agricultural value added exhibits a U-shaped relationship, wherein both low and high productivity levels are associated with increased freshwater use, while intermediate productivity generates the greatest negative impact. The XGBoost model achieves substantial predictive performance (Coefficient of Determination (R²) = 0.78, Root Mean Squared Error (RMSE) = 0.806, Mean Absolute Percentage Error (MAPE) = 0.86%), demonstrating its capacity to capture heterogeneous, nonlinear dynamics that linear models fail to detect. The robustness of the model was further assessed using Leave-One-Out Cross-Validation (LOOCV) to evaluate its out-of-sample predictive performance and mitigate potential overfitting arising from the limited sample size. These findings underscore the necessity of adaptive water management strategies that incorporate scale-dependent effects and productivity heterogeneity. Policies optimizing land allocation and promoting efficient agricultural practices can enhance water-use efficiency while sustaining sectoral output. The study highlights the value of interpretable machine learning in advancing empirical understanding of the water–agriculture nexus under conditions of structural economic change.
This study develops an interpretable forecasting framework for container throughput with a specific focus on supporting integrated port operations and transport system coordination. Using monthly operational data from Mwani, Qatar (2017–2023), the proposed approach captures trend evolution, seasonal patterns, and calendar-related variations to generate short- and medium-term forecasts of container flows. Beyond predictive accuracy, the framework is designed to provide transparent insights into the operational drivers of throughput dynamics. The analysis identifies vessel call frequency as the dominant factor influencing throughput fluctuations, while trade-related indicators contribute consistent explanatory signals across time. The resulting forecasts show strong agreement with observed values, achieving a mean absolute error (MAE) of 3.84%, which demonstrates the reliability of the approach for operational planning. From a transport integration perspective, the forecasting outputs are directly linked to key decision-making processes within port systems, including quay crane deployment, yard allocation, automated vehicle scheduling, and truck gate coordination. Scenario-based analysis under simulated trade disruptions reveals temporary degradation in forecasting performance, followed by gradual recovery as system conditions stabilize, highlighting the sensitivity of port operations to external shocks. By combining predictive modelling with interpretable analysis, this study provides a practical tool for enhancing coordination between maritime flows and landside logistics. The findings contribute to the development of data-informed strategies for port operation management and offer a scalable approach for improving decision support in integrated transport systems.
For Authors
For Reviewers
For Editors
Open Access
