Water inrush hazard seriously threatens construction safety of subsea tunnels in unfavorable geological areas. In recent years, a large number of subsea tunnels have been built worldwide, some of which have experienced many water inrush disasters, especially in Japan and Norway. In this paper, a systematic methodology is proposed to rigorously review the current literature about water inrush in subsea tunnels. Emphasis is placed on recorded causes and evolution processes of water inrush, as well as relevant mitigation measures. In particular, the geological conditions that generate such water inrush hazards are initially discussed by counting cases of tunnel water inrush in the past decades (43 cases of water inrush hazards in tunnels (including mountain tunnels)). The process of formation of failure modes of water inrush, and the corresponding research methods (including theoretical, numerical and experimental) are reviewed, and can be used to pave the ways for hazard prevention and future research. This is followed by a summary of the prevention methods and mitigation measures used in practice, and a short discussion of the achievements and limitations of each method. Then combined with the evolution characteristics of the failure area, the water inrush process of different modes is divided into three stages, with a proposed a grouting scheme for each stage. Finally, concluding remarks, current research gaps and future research directions on subsea tunnel water inrush are provided and discussed.

The soil‐water characteristic curve (SWCC) plays an essential role in the analysis of the shear strength, deformation, hydraulic conductivity, and aqueous diffusion of unsaturated soils. A review of methods proposed for fitting the main SWCCs is presented. The methods can be categorized into three groups, namely, (i) empirical methods, (ii) domain methods, and (iii) theoretical methods. A hysteretic model considering the contact angle hysteresis is proposed for fitting the main SWCCs based on a theoretical method. The model links the physical quantities at the micro‐and macro‐scale. Specifically, a modified expression describing the relationship between the volume of the liquid bridge and the water content is derived, and the contact angle hysteresis behavior is modeled based on the study of total thermodynamic potential energy. The model has a clear physical meaning, and the parameters can be obtained easily. Comparisons of the model simulation and experimental data show that the proposed model accurately captures the main features of the SWCC and describes the nonlinear relationship between the matric suction and the water content during wetting and drying processes.

UNSODA, a free international soil database, is very popular and has been used in many fields. However, missing soil property data have limited the utility of this dataset, especially for data-driven models. Here, three machine learning-based methods, i.e., random forest (RF) regression, support vector (SVR) regression, and artificial neural network (ANN) regression, and two statistics-based methods, i.e., mean and multiple imputation (MI), were used to impute the missing soil property data, including pH, saturated hydraulic conductivity (SHC), organic matter content (OMC), porosity (PO), and particle density (PD). The missing upper depths (DU) and lower depths (DL) for the sampling locations were also imputed. Before imputing the missing values in UNSODA, a missing value simulation was performed and evaluated quantitatively. Next, nonparametric tests and multiple linear regression were performed to qualitatively evaluate the reliability of these five imputation methods. Results showed that RMSEs and MAEs of all features fluctuated within acceptable ranges. RF imputation and MI presented the lowest RMSEs and MAEs; both methods are good at explaining the variability of data. The standard error, coefficient of variance, and standard deviation decreased significantly after imputation, and there were no significant differences before and after imputation. Together, DU, pH, SHC, OMC, PO, and PD explained 91.0%, 63.9%, 88.5%, 59.4%, and 90.2% of the variation in BD using RF, SVR, ANN, mean, and MI, respectively; and this value was 99.8% when missing values were discarded. This study suggests that the RF and MI methods may be better for imputing the missing data in UNSODA.