There is a significant variability of salinity level in sensitive marine clays (SMC), which will produce an important impact on the development of mechanical characteristics in stabilized SMC. The influences of salt content (NaCl salt: 3, 10, and 20 g/L) on mechanical properties evolution of cement-stabilized SMC under different curing time (1, 7, 28, 60, and 90 days) have been experimentally investigated and modeled. The results indicate that the strength and modulus increase constantly with time but the time rates decrease. Meanwhile, the apparent improvement of strength and modulus at early age (up to 7 days) is observed. Higher NaCl content can bring a larger strength gain to stabilized SMC after same curing time and the enhancing effect of high salt contents (10 and 20 g/L) becomes more obvious with the extension of curing time. Whereas, the enhancing effect of high NaCl content on modulus is limited compared with strength. Further improvement provided by excessive NaCl salt (20 g/L) is not as effective. In addition, the predictive models have been established to quantitatively evaluate the evolution of mechanical properties in stabilized SMC with different NaCl contents. The capability of developed models has been demonstrated through the good agreement between simulated and experimental results.

Global warming accelerates permafrost degradation, compromising the reliability of critical infrastructure relied upon by over five million people daily. Additionally, permafrost thaw releases substantial methane emissions due to the thawing of swamps, further amplifying global warming and climate change and thus posing a significant threat to more than eight billion people worldwide. To mitigate this growing risk, policymakers and stakeholders need accurate predictions of permafrost thaw progression. Comprehensive physics-based permafrost models often require complex, location-specific fine-tuning, making them impractical for widespread use. Although simpler models with fewer input parameters offer convenience, they generally lack accuracy. Purely data-driven models also face limitations due to the spatial and temporal sparsity of observational data. This work develops a physics-informed machine learning framework to predict permafrost thaw rates. By integrating a physics-based model into machine learning, the framework significantly enhances the feature set, enabling models to train on higher-quality data. This approach improves permafrost thaw rate predictions, supporting more reliable decision-making for construction and infrastructure maintenance in permafrost-vulnerable regions, with a forecast horizon spanning several decades.

Foundation settlement is a common problem in civil engineering. In the case of un-even settlement, it can lead to structural deformation and damage, which seriously affects the safety and reliability of the project. Therefore, the influence of adjusting the stiffness of the foundation on un-even settlement was analyzed through finite element analysis to effectively solve un-even settlement. By simulating the settlement of soil under different foundation stiffness and load conditions, the influence of foundation stiffness adjustment on soil deformation and settlement distribution was analyzed, and its impact on structural safety was evaluated. These studies confirmed that thickened layers could effectively solve the un-even settlement. Within the range of 0.2 to 1.0 meters, the difference in thickness was the greatest. The adjustment of differential settlement by layer thickness was phased and decreased with increasing thickness. Adjusting the stiffness of the foundation could effectively solve un-even settlement, reduce differences in soil settlement, and improve the overall stability and safety of the structure. These results have important guiding significance for the design of foundation and the solution of un-even settlement problems in engineering practice and provide certain reference and basis for further research.

Vast amounts of soil organic matter (SOM) have been preserved in arctic soils over millennia time scales due to the limiting effects of cold and wet environments on decomposer activity. With the increase in high latitude warming due to climate change, the potential decomposability of this SOM needs to be assessed. In this study, we investigated the capability of mid infrared (MIR) spectroscopy to quickly predict soil carbon and nitrogen concentrations and carbon (C) mineralized during short-term incubations of tundra soils. Active layer and upper permafrost soils collected from four tundra sites on the North Slope of Alaska were incubated at 1, 4, 8 and 16 C for 60 days. All incubated soils were scanned to obtain the MIR spectra and analyzed for total organic carbon (TOC) and total nitrogen (TN) concentrations, and salt-extractable organic matter carbon (SEOM). Partial least square regression (PLSR) models, constructed using the MIR spectral data for all soils, were excellent predictors of soil TOC and TN concentrations and good predictors of mineralized C for these tundra soils. We explored whether we could improve the prediction of mineralized C by splitting the soils into the groups defined by the influential factors and thresholds identified in a principal components analysis: (1) TOC > 10%, (2) TOC 0.6%, (5) acidic tundra, and (6) non-acidic tundra. The best PLSR mineralization models were found for soils with TOC < 10% and TN < 0.6%. Analysis of the PLSR loadings and beta coefficients from these models indicated a small number of influential spectral bands. These bands were associated with clay content, phenolics, aliphatics, silicates, carboxylic acids, and amides. Our results suggest that MIR could serve as a useful tool for quickly and reasonably estimating the initial decomposability of tundra soils, particularly for mineral soils and the mixed organic-mineral horizons of cryoturbated soils.