Climate change and extreme weather events are significantly impacting geotechnical systems, necessitating urgent research into unsaturated soil mechanics and advanced computational methods. As global temperatures rise and precipitation patterns become more erratic, soils are experiencing frequent cycles of wetting and drying, leading to altered shear strength, swelling, and settlement behaviours. Unsaturated soil mechanics plays a critical role in understanding these changes, particularly in expansive clays and partially saturated slopes prone to failure during heavy rainfall or prolonged droughts. However, traditional soil models often fail to account for the dynamic interactions between climate variables and soil moisture, highlighting the need for more robust, physics-based frameworks.
To address these challenges, researchers are increasingly turning to artificial intelligence (AI)–driven and computational geotechnics to improve predictive modelling and risk assessment. Machine learning algorithms can analyse vast datasets from field and laboratory tests, enabling better predictions of soil behaviour under varying climatic conditions. In addition, digital twin technology allows for real-time monitoring of geo-structures, simulating how they respond to environmental stressors such as extreme rainfall or seismic activity. These advancements help mitigate uncertainties in geotechnical design, particularly when dealing with the nonlinear behaviour of unsaturated soils under climate-induced loading.
The integration of climate science, unsaturated soil mechanics, and computational modelling is paving the way for more resilient infrastructure. Future research must focus on developing adaptive geotechnical systems that can withstand long-term climatic shifts while leveraging AI for real-time decision making. By combining advanced soil characterisation techniques with predictive analytics, engineers can design foundations, slopes, and embankments that are not only stable today but also resilient to the changing conditions of tomorrow. This interdisciplinary approach will be crucial in ensuring sustainable and safe infrastructure in an era of climate uncertainty.
The latest contributions to Environmental Geotechnics reflect the field’s growing advancement in addressing complex subsurface challenges. The four papers featured in this issue – by Akhtar and Li (2025), Toualbia et al. (2025), He et al. (2025), and Niu et al. (2025) – span a spectrum of geotechnical challenges, from freeze–thaw (F–T) durability and capillary deterioration in stabilised soils to advanced modelling of tunnel convergence and its environmental impacts. Together, they showcase the integration of experimental, statistical, and numerical methods to improve infrastructure resilience in diverse geotechnical settings.
Akhtar and Li’s experimental study on lime-treated Quebec silty clay under F–T cycling offers a detailed characterisation of both compressive and tensile strength evolution. Their use of a custom-designed double-punch test to capture brittle tensile behaviour is particularly innovative. The study identifies a critical threshold – around 10–12 F–T cycles – beyond which strength degradation stabilises or reverses, suggesting secondary pozzolanic reactions or structural adaptation. This has direct implications for infrastructure design in permafrost and seasonally frozen regions.
Complementing this, Toualbia et al. present a robust statistical model (UCS–FT) for predicting the unconfined compressive strength of lime-stabilised clayey soils under F–T conditions. Utilising a comprehensive dataset of 290 experimental points, the model integrates key soil parameters – lime content, curing time, plasticity index, and moisture content – into a predictive framework validated against independent datasets. The model’s high R2 value (0.80) and low relative mean error (6.25%) underscore its practical utility for early-stage design and risk assessment.
He et al. shift the focus to coastal geotechnics, examining the deterioration of soft soils solidified with soda residue (SR), ground granulated blast furnace slag (GGBS), and carbide slag (CS) under capillary seawater absorption. Their findings challenge the dominance of cement as a stabiliser, demonstrating that SR–GGBS–CS blends not only reduce water absorption and swelling but also exhibit superior resistance to salt crystallisation and strength loss. The study’s multi-scale approach –combining unconfined compressive strength testing, strain monitoring, and microstructural analysis by way of NMR and XRD – provides a holistic view of deterioration mechanisms. Notably, the optimal blend (25% SR, 10% GGBS, 4% CS) maintained over 75% of its strength after 7 days of crystallisation, outperforming traditional cement-based treatments.
Finally, Niu et al. introduce a novel function fitting method for tunnel convergence evaluation, addressing the limitations of traditional circular or elliptical fitting techniques. By using the tunnel central angle as an independent variable, their method enables the mathematical representation of non-uniform convergence patterns and integrates seamlessly with displacement-controlled finite element simulations. The approach is validated through multiple case studies – including the Heathrow Express and Milan Metro tunnels – demonstrating its ability to predict ground surface settlement (GSS) with high fidelity. Importantly, the method supports probabilistic GSS estimation even in the early stages of tunnelling, using only gap parameters or volume loss data.
Collectively, these studies reflect a paradigm shift in geotechnical engineering – from empirical and idealised models to data-driven, multi-parameter, and context-sensitive approaches. They also highlight the importance of tailoring solutions to specific environmental stressors, whether thermal cycling in cold regions, saline intrusion in coastal zones, or deformation in urban tunnelling. These contributions not only advance the scientific understanding of soil and tunnel behaviour under environmental pressure but also offer practical tools for building more resilient and sustainable infrastructure in vulnerable regions.
