The soil in the northwest loess irrigation area has been affected by the dual effects of dry-wet cycling and salt pollution, leading to significant deterioration of the loess properties in this region. Taking Q₃ loess contaminated by Na₂SO₄ as the research object, direct shear tests and scanning electron microscopy (SEM) are used to analyse the changes in mechanical properties and microstructure characteristics after different numbers of dry-wet cycles. The results show that: Under the dual effects of dry-wet cycling and salt pollution, the strength of the loess deteriorates significantly, and the microstructure changes significantly. Among them, the cohesion shows obvious deterioration, with a decline rate of up to 38.24% to 51.4%. With the increase of salt pollution and dry-wet cycling times, the proportion of individual particles in the loess body increases significantly, and the aggregates begin to disintegrate; Na₂SO₄ repeatedly crystallizes and dissolves during dry-wet cycling, thereby exerting a combined effect of salt swelling and salt erosion on the soil; among the two effects, the destructive effect of dry-wet cycling on the soil structure is greater than that of salt pollution. The research results provide an effective reference for engineering activities and pollution control in loess areas.

With the rapid development of infrastructure such as railways, highways, and water conservancy and hydropower projects in the western region of China, a number of extremely complex geological conditions and highly challenging ultra-deep buried tunnels have emerged. These developments place higher demands on the collection, analysis, and application of tunnel geological information. Geological information transparency technology plays a crucial role in ensuring safety, controlling costs, and optimizing construction progress, serving as key technical support for the smooth execution of ultra-deep buried tunnel projects. This paper systematically reviews the current development status of geological information detection technologies, covering remote sensing survey technology, geophysical exploration technology, and directional drilling technology in the pre-construction phase, as well as advanced geological forecasting technology during construction. The application scope and development directions of various technologies are analyzed. The paper also reviews the research status of multi-source geological information fusion technology and 3D geological modeling technology, detailing their progress and challenges in predicting geological conditions, model building, and applications. Finally, this paper looks ahead to the overall development of geological information transparency technology for tunnels: First, establishing a five-dimensional integrated detection system encompassing "space-air-ground-borehole-tunnel" to improve detection accuracy; second, developing intelligent geological monitoring technologies and equipment; and third, based on multi-source data fusion, constructing 3D geological models and applying virtual reality technology to achieve dynamic geological information visualization, thus enhancing the safety and efficiency of tunnel construction.

At present, the construction of large-span and super-large-span subway tunnels is increasing. Temporary support is often designed to reduce the span in the construction of those tunnels. However, temporary support needs to be removed before secondary lining is constructed due to the spatial overlap between them. The removal of temporary support is the weak situation, because the original stress balance of the structure will be broken. Improper construction can easily cause safety accidents such as tunnel collapse. In view of the lack of systematic research on the structural mechanical response and safety of the super-large-span tunnel constructed by the double-layer initial support arch-cover method at home and abroad, this paper takes the four-line parallel tunnel of Huahua section of Guangzhou Metro Line 11 as the basis project. Through the combination of theoretical analysis and numerical calculation, the mechanical response and safety of the structure during the dismantling of the tunnel are analyzed from both horizontal and vertical aspects. Finally, a scientific, reasonable, safe, efficient and rapid dismantling scheme was determined, and successfully passed the inspection of on-site construction. The results show that it is feasible to dismantle the super-large-span tunnel in urban soft stratum by using the scheme of 'first edge and then middle, and symmetrical demolition in cross direction, first three demolition and one demolition, and then one demolition and one demolition in longitudinal direction', which can provide reference and guidance for subsequent projects.

It is quite important for the refined design and construction of geotechnical engineering to obtain a certain amount of borehole data and then determine the stratum profile information through geotechnical investigation. However, due to the particularity of the area where some engineering sites are located (existing old urban areas, cultural Relic Protection Building, etc.), there is often a problem that the borehole data is difficult to obtain, which makes the corresponding stratum information determination challenging. To this end, an artificial intelligence (AI) method for stratum information reconstruction is developed based on the fully convolutional network (FCN). The core idea of this method is to use the existing borehole data in the region as a learning sample, analyze and extract the multi-dimensional information features of the sample (vertical stratification, horizontal extension), and then use this information feature as a template to perform probability-based stratum profile information interpolation reconstruction for engineering sites with only a small amount of borehole data. Through the study and reconstruction of the geological survey data of a tunnel project and foundation pit project in the ancient city of Suzhou, it is found that the accuracy of stratum prediction gradually tends to be stable after the number of simulations increases to more than 30 times, and can reach about 90%. This verifies the applicability of the developed AI reconstruction method of stratum information, which will provide an effective choice for the prediction of complex stratum information in related projects.

The sandy clayey purple soil is susceptible to alterations in particle size distribution due to hydraulic erosion, such as rainfall and fluctuations in reservoir water levels. These changes can affect the soil's strength. To elucidate the strength characteristics of sandy clayey purple soil in the influence of the absence of different particle groups, consolidated drained triaxial shear tests of saturated sandy clayey purple soil with five distinct particle groups were conducted. The fractal dimension was employed to quantify the physical properties of varying particle sizes. The relationship between the strength characteristics of purple soil with different particle groups and the fractal dimension was obtained. The results show that: (1) There is no significant change in the internal friction angle of the soil shear strength index under the missing effect of different grain groups, but the cohesion and fractal dimension show a hyperbolic relationship; the initial tangent modulus of different grain groups is affected by confining pressure and fractal dimension, and there is a quadratic function relationship between parameter n and fractal dimension. Under low confining pressure, the stress-strain relationship of different grain groups is obviously different. With the increase of confining pressure, the stress-strain characteristics tend to be consistent. (2) Based on the Duncan-Chang E-ν model, the nonlinear elastic hyperbolic constitutive model of saturated sandy clayey purple soil under the effect of different particle groups is established by fitting the cohesion-fractal dimension hyperbolic equation, and the validity of the model is verified.

In order to study the dynamic changes and amplitude values of groundwater level in karst mountainous areas, taking the karst mountainous city of Guiyang as an example, selecting daily monitoring data and precipitation data from nine groundwater level dynamic observation points from 2022 to 2023, autocorrelation and cross-correlation analysis are used to analyze the response of groundwater level to precipitation and explore the influence of runoff and drainage conditions and terrain slope on the dynamic changes of groundwater level. The results show that: (1) The groundwater level in the study area is buried at a depth of 1.21 ~ 27.68 m, with an annual variation range of 1.54 ~ 11.99 m, and there are significant differences in the spatiotemporal distribution of groundwater level dynamics; (2) The relevant analysis results indicate that there is a significant lag in the response of groundwater level to precipitation signals in the study area, with an average lag time of 0~4 days, and it gradually increases from the supply area to the discharge area; (3) The terrain slope is an important factor affecting the amplitude of groundwater level variation in karst mountainous areas. The terrain slope in the study area is positively correlated with the amplitude and variation of groundwater level, with linear goodness of fit (R²) of 0.65 and 0.78, respectively; (4) The depth of groundwater level gradually decreases from the recharge area to the discharge area, and the range of water level changes from the runoff area to the recharge area to the discharge area.