An Environmental, Social, and Governance (ESG) report is an essential information source for evaluating a company’s performance in sustainability practices. Organizations structure their environmental impacts, social responsibilities, and governance practices within a defined framework. This standardization is provided by the Global Reporting Initiative (GRI), which constitutes an internationally recognized guideline for sustainability reporting. Traditional reporting workflows are time-consuming for organizations and prone to data-entry errors, which limits the reliability of disclosed information. In this context, leveraging the capabilities of Large Language Models (LLMs) offers significant time and resource savings. This study uses the Llama-3.1-8B-Instruct model under two scenarios, Retrieval-Augmented Generation (RAG) and Low-Rank Adaptation (LoRA) fine-tuning, to analyze 30 food-sector ESG reports and produce ESG summaries, SWOT analyses, and GRI-aligned recommendations. The two approaches are evaluated on a stratified hold-out set of 6 unseen test reports (24 reports used for training) under a fair, matched-budget setup in which RAG retrieves the target report at inference. On four quality metrics, LoRA achieved higher mean scores than RAG; however, statistically significant differences were observed in only 4 of the 12 task–metric comparisons. Token usage was comparable, whereas RAG was substantially faster at inference. Rather than favoring one approach over the other, these findings reveal a trade-off between output quality and computational efficiency: LoRA yields quality gains on specific metrics, whereas RAG is substantially more efficient at inference. Given the limited size of the held-out test set, these results should be interpreted with caution.

Efficient mixing under laminar flow conditions remains a critical challenge in microfluidic systems because molecular diffusion alone is generally insufficient to achieve rapid and homogeneous species transport. In this study, the influence of obstacle orientation on mixing performance in passive micromixers was systematically investigated through numerical simulations. Inclined straight obstacles with orientation angles of 15°, 30°, 45°, and 60° were incorporated into microchannels under both leaky and leak-free configurations. Flow and concentration fields were solved using COMSOL Multiphysics, and the resulting mixing efficiencies and times were quantitatively evaluated. It was found that the introduction of inclined obstacles substantially enhanced mixing performance relative to a simple unobstructed microchannel. Superior mixing behavior was consistently achieved in the leak-free configuration, where stronger flow perturbations and more pronounced recirculation zones were generated within the central mixing region. For the leak-free configuration, mixing efficiency was observed to increase with decreasing obstacle angle. In contrast, no monotonic relationship between obstacle angle and mixing performance was identified for the leaky configuration. Among all investigated designs, the 15° obstacle configuration exhibited the highest overall performance, achieving mixing efficiencies of approximately 92% and nearly 100% in the leaky and leak-free configurations, respectively. To further evaluate the influence of geometric scale, the microchannel length was doubled for the 45° configuration. Enhanced concentration uniformity and reduced mixing time were achieved in the extended leaky microchannel, whereas no improvements were observed in the corresponding leak-free design. These findings demonstrate that obstacle orientation and channel configuration exert a strong influence on microscale transport phenomena and mixing enhancement. The proposed obstacle-based passive micromixer design provides an effective and energy-efficient strategy for improving mixing performance in microfluidic devices and offers valuable design guidelines for applications in biomedical analysis, chemical processing, and lab-on-a-chip systems.

Shoreline change strongly affects ecosystem conditions and livelihood security in deltaic coasts. However, long-term and spatially explicit baselines remain limited for many rapidly changing coastal systems. This study quantified multi-decadal shoreline dynamics in Ujung Pangkah, Indonesia, to identify persistent patterns of accretion and erosion and to assess their implications for sustainable coastal management. Multi-epoch satellite images from 1973 to 2021 were used to extract shoreline positions through water-index-based classification. The extracted shorelines were analyzed using standardized Digital Shoreline Analysis System (DSAS) metrics to estimate net shoreline movement (NSM) and end point rate (EPR) across segmented coastal areas. The results indicate a segment-structured shoreline mosaic rather than a uniform coast-wide trend. Most sectors were accretion-dominated, with the accretion component reaching approximately +12 to +15 m$\cdot$yr$^{-1}$, particularly in Area C. In contrast, Area D formed the main erosional hotspot, with an erosion component of -8.68 m$\cdot$yr$^{-1}$ and an NSM erosion value of -416.53 m, while its net EPR and net NSM were -0.66 m$\cdot$yr$^{-1}$ and -31.77 m, respectively. These findings show that shoreline change in Ujung Pangkah is spatially concentrated in localized reaches. Therefore, coast-wide averages may obscure areas where erosion risk persists and where accretion gains are sustained. This study provides a quantitative long-term baseline and a reproducible remote-sensing and GIS-based workflow to support hotspot identification, segment-scale monitoring, and the prioritization of coastal protection and rehabilitation measures in dynamic deltaic environments.

Coastlines host dense human activity that concentrates combustion and elevates carbon monoxide (CO) and nitrogen dioxide (NO₂) burdens. Yet complex coastal meteorology often limits ground monitoring. This study addresses this gap with a multi-year, dual-pollutant, jurisdiction-scale analysis using a transparent Sentinel-5P column-burden workflow. This work employs the workflow on Canada’s Nova Scotia (NS), a cool and relatively stable North Atlantic coast, and the US state of Louisiana (LA), a warm-humid Gulf coast with one of the densest refining hubs, providing two contrasting coastal domains. The present analysis evaluates 2019–2024 tropospheric column CO and NO₂, applies uniform quality-assured screening, generates time series composites at native resolution, classifies spatial fields with Jenks Natural Breaks, and examines temporal trends. Columns are compared with inventories and ground networks as consistency checks. Six-year means highlight persistent contrasts: NS’s column CO is slightly higher than LA’s (0.0338 vs. 0.0321 mol m⁻²), and NS’s NO₂ is ≈ 2.5× LA’s (6.09×10⁻⁵ vs. 2.39×10⁻⁵ mol m⁻²). In NS, NO₂ peaks in summer, while CO reaches its highest seasonal mean in spring; in LA, NO₂ peaks in winter and CO peaks in spring. Recurring hotspots appear over Halifax-Dartmouth and North Sydney, and along the Baton Rouge-New Orleans corridor and northern parishes. These patterns may reflect a combined influence of coastal setting, seasonal atmospheric structure, and local activity, although direct meteorological attribution was not performed. By integrating satellite archives with ground networks, the study provides a reproducible, auditable approach that translates seasonal column dynamics into jurisdiction-ready evidence for evaluation calendars and corridor-focused siting, improving the timing and targeting of coastal air-quality management, and supporting United Nations Sustainable Development Goals (SDGs) 3 and 11.

This study aims to evaluate and rank regional agricultural technology competitiveness in East Java, Indonesia, using a structured multi-criteria decision-making approach. Specifically, it addresses four key objectives: (1) to apply the Preference Ranking Organization Method for Enrichment Evaluation (PROMETHEE) method to assess and rank regional competitiveness across multiple technological dimensions; (2) to examine whether agricultural technology adoption levels differ significantly across regions using one-way Analysis of Variance (ANOVA); (3) to evaluate the sensitivity and robustness of the ranking results under alternative weighting scenarios through sensitivity analysis and rank correlation measures (Spearman’s ρ and Kendall’s τ); and (4) to derive policy-relevant and system-oriented implications for enhancing competitiveness and reducing regional disparities. The study employs a quantitative approach based on primary survey data collected from 210 farmers across seven regions in East Java. Four key dimensions are considered, namely environmental, irrigation, marketing, and production technologies. The PROMETHEE method is used to generate regional rankings, while ANOVA is applied to test for statistically significant differences in technology adoption. Robustness is further assessed through systematic weight variations and rank correlation analysis. The results reveal substantial regional disparities in relative technological competitiveness, with leading regions demonstrating more balanced, integrated adoption across multiple technological dimensions. ANOVA results confirm that differences in technology adoption across regions are statistically significant (p < 0.01), thereby providing complementary statistical evidence for inter-regional variation in the underlying technology adoption indicators used in the PROMETHEE analysis. The robustness analysis shows that the ranking results are highly stable across most weighting scenarios, with only minor variations observed when marketing-related criteria are emphasized. This study contributes methodologically by integrating multi-criteria decision-making with statistical validation and robustness testing in a unified framework. From a policy perspective, the findings highlight the importance of strengthening market access, improving technological integration, and implementing region-specific interventions to enhance agricultural competitiveness and reduce disparities.

The research analyzes whether monitoring system design, calibration management, timestamp consistency, data traceability, and verification procedures relate to the risk control of the financial aspects of methane-abatement engineering projects. An analytical case study based on a single project and involving a before-and-after comparison of the implementation of an monitoring, reporting, and verification (MRV) regime was conducted under fixed engineering and accounting conditions. This design allows the comparison to focus on differences in MRV evidence management conditions rather than on changes in physical mitigation technology. Conservative issuability was estimated using the low-confidence adjustment metric (LCAM). This analytical metric scales engineering emission reductions by evidence-related factors without supplanting registry rules or verifier judgment. With the enhanced MRV regime, the conservatively supportable fraction was 77.0% to 91.3%, while the realized price wedge declined from 0.30 to 0.12. The monitoring-to-issuance period was also shortened by 50 days.