The UK faces an urgent challenge to simultaneously accelerate housing delivery and reduce whole-life carbon emissions, yet robust empirical evidence on the carbon performance of modular steel housing remains limited. This study aims to quantify the carbon impacts of a modular light-gauge steel frame social housing dwelling in the UK and to benchmark its performance against contemporary low-carbon construction typologies. A cradle-to-grave life cycle assessment was conducted using primary project data from a real modular housing development, with embodied carbon modelled in One Click LCA and operational energy assessed through SAP 10.2-verified datasets. The results indicate a total whole-life carbon footprint of 91.3 tCO2e over a 50-year period, with embodied emissions (A1–A3) accounting for 38.2% and operational energy and water use contributing 48.1%. The normalised embodied carbon intensity of 366 kgCO2e/m2 (A1–A5) is comparable to recent high-performing cross-laminated timber buildings, demonstrating that optimised modular steel systems can allow for low-carbon outcomes typically associated with bio-based construction. Sensitivity analysis shows that low-carbon foundation concrete, bio-based insulation, and steel optimisation can reduce upfront emissions by approximately 8–10%. Dynamic energy simulations were also used to assess how different design choices influence operational carbon emissions. This study provides transparent, real-project evidence of the whole-life carbon performance of UK modular light-gauge steel frame housing and identifies practical design strategies for further decarbonisation. The findings support informed decision-making for policymakers, designers, and housing providers seeking scalable, low-carbon residential solutions.
The monitoring of turbidity in estuarine environments is a challenging essential task for managing water quality and ecosystem health. This study focuses on the lower reaches of the River Mersey, Liverpool. Harmonized Sentinel-2 MSI Level-2A imagery was integrated with in situ measurements from seven Environment Agency monitoring stations for two consecutive years (January 2023–January 2025). The workflow included image preprocessing, spectral index calculation, and the application of four machine learning algorithms: Gradient Boosting Regressor, XGBoost, Support Vector Regressor, and K-Nearest Neighbors. Among these, Gradient Boosting Regressor achieved the highest predictive accuracy (R2 = 0.84; RMSE = 15.0 FTU), demonstrating the suitability of ensemble tree-based methods for capturing non-linear interactions between spectral indices and water quality parameters. Residual analysis revealed systematic errors linked to tidal cycles, depth variation, and salinity-driven stratification, underscoring the limitations of purely data-driven approaches. The novelty of this study lies in demonstrating the feasibility and proof-of-concept of using machine learning to derive spatially explicit turbidity estimates under data-limited estuarine conditions. These results open opportunities for future integration with Computational Fluid Dynamics models to enhance temporal forecasting and physical realism in estuarine monitoring systems. The proposed methodology contributes to sustainable coastal management, pollution monitoring, and climate resilience, while offering a transferable framework for other estuaries worldwide.
Suspended sediments regulate turbidity and act as major vectors for metals, organic contaminants, and nutrient fluxes in estuarine systems. Effective assessment of sediment-mediated pollution therefore depends on coordinated monitoring of sediment and contaminant dynamics. This study evaluates whether a long-term regulatory monitoring network can resolve sediment-associated pollution dynamics in the River Mersey estuary (UK). A total of 44,516 water-column observations (2015-2025) across 42 monitoring stations and 187 determinants were analysed. Monitoring frequency classification revealed pronounced structural imbalance: only 6.9% of determinants were sampled routinely (≥1000 observations), whereas 67.4% were recorded fewer than 100 times. Physical and oxygen-related variables accounted for 52% of all observations, while sediment and turbidity indicators comprised only 4.2%. Temporal separation analysis revealed a bimodal sampling structure characterised by episodic co-sampling embedded within extended periods of asynchronous measurement. While pairwise co-sampling coherence (CCR) was often moderate to high across several process groups, strict multivariate coherence (CCRₛₜᵣᵢcₜ) remained low across much of the network, indicating that complete multivariate observation states suitable for integrated sediment-contaminant assessment were uncommon. Diagnostic machine-learning reconstruction yielded moderate performance under random validation (R2 = 0.588) but substantially lower spatial generalisation under grouped cross-validation (mean R2 ≈ 0.26), indicating limited transferability across stations. These findings suggest that sediment-mediated pollution assessment is constrained more strongly by monitoring architecture, temporal alignment, and environmental heterogeneity than by modelling approaches alone. The study introduces a diagnostic framework combining temporal separation analysis, coherence metrics, and hierarchical validation for evaluating monitoring-network suitability in estuarine systems.