Introduction

Ion-adsorbed rare earth ores, as globally scarce resources, are of strategic importance to high-tech industries, including aerospace, renewable energy, semiconductor materials, and medicine1,2,3,4,5. In recent years, in situ leaching has become the predominant mining technique in this region. This process involves the injection of leaching solutions directly into the ore body, thereby avoiding disturbance to the surface vegetation and minimizing soil excavation. Through ion-exchange reactions, cations in the leaching solution replace the rare earth ions adsorbed on the clay mineral surfaces, producing a rare earth-bearing pregnant solution that is subsequently recovered and extracted26,7. This method substantially enhances the mining efficiency while reducing the environmental disturbance. Nevertheless, in situ leaching also introduces significant geotechnical risks, most notably landslides8. The interplay between fluid injection, pore pressure dynamics, and slope stability is a critical geoenvironmental concern, as demonstrated in other contexts such as hydrocarbon reservoirs9. Slope instability induced by solution injection represents a critical geotechnical challenge widely recognized across major engineering projects, as evidenced by research conducted in transportation corridors and hydropower development zones10. Between 2001 and 2016, more than 890 landslides were recorded in the Gannan mining area, with 80.8% occurring during solution injection11. Research on the mechanisms of slope instability indicates that fluid migration patterns are a key factor controlling the instability process. As revealed by the spatial kinematic analysis of the Seymareh giant rockslide, even in tectonically driven landslides, changes in hydrological conditions often constitute an important triggering mechanism12. During the in-situ leaching of ionic rare earth minerals, the migration and accumulation of leachate can similarly induce catastrophic landslides by altering the mechanical properties of the soil, necessitating real-time monitoring and early warning. Excessive injection or blocked flow path can lead to a sharp increase in soil pore water pressure, resulting in a decrease in shear strength, and ultimately leading to slope failure13. Consequently, elucidating the seepage pathways and accumulation zones during the injection process is essential for effective landslide identification and early warning in rare earth mining regions.

Extensive research has examined the hydrogeological and geotechnical aspects of in situ leaching. Guo et al.14 conducted saturated permeability tests under varying leaching conditions using the variable head method, highlighting the roles of the leachate type, concentration, and injection path. Gui et al.15 derived a steady-state saturation line equation for exposed linear ore bodies based on groundwater dynamics and Dupuit’s assumption and subsequently verified it through engineering practice and error analysis. Liu et al.16 investigated the influence of particle size distribution, particle migration, and Atterberg limits on permeability during the leaching of weathered crust eluvial ores. And then they further revealed how particle size distribution controls the evolution of pore structure and seepage path ways by the computed tomography scanning17. Ju et al.18 conducted physical modeling experiments on the preferential flow in ore piles under horizontal injection, and, together with numerical simulations, clarified its impact on the seepage characteristics. Wang et al. examined the evolution of pore structures in ion-adsorbed rare earth ores during the leaching process by conducting column leaching experiments and nuclear magnetic resonance (NMR) tests on remolded, water-saturated samples. Their findings revealed that leaching alters the ore-body microstructure through a coupled mechanism of solution seepage and ion exchange19,20. Wu et al. utilized nuclear magnetic resonance imaging (NMRI) to characterize seepage channels within rare earth ore bodies and established a numerical model of in-situ leaching to capture the coupled processes of seepage, ion exchange, and solute migration. This model was further applied to elucidate the evolutionary dynamics of these processes under different injection intensities and leaching agent concentrations21. Wang et al. constructed a fully three-dimensional numerical model of an experimental stope to simulate the in-situ leaching process of ion-adsorbed rare earth, thereby reproducing ore-body leaching behavior under the coupled effects of seepage, ion exchange, and solute migration in porous media22. Long et al. employed the convection–dispersion equation (CDE) to describe the transport behavior of ammonium ions (NH₄⁺) and rare earth ions (RE³⁺) during column leaching and further analyzed the influence of the leaching agent concentration on the rare earth recovery rates23. Zhou et al. conducted column leaching experiments on rare earth samples and explored strategies to enhance the leaching efficiency by applying an external electric field across the sample ends24. In a related study, Zhou et al. investigated the role of surfactants in promoting the leaching of weathered crust eluvial rare earth ores by introducing surfactants into the leachate25,26. Collectively, these studies have elucidated the coupled processes of seepage, ion exchange, and solute migration through laboratory infiltration tests, column leaching experiments, physical modeling, and numerical simulations, with a predominant focus on improving the leaching efficiency. However, field-scale in situ investigations addressing the geological disaster risks induced by poor permeability during the leaching process remain scarce.

Currently, the identification and early warning of geohazards in the leaching process of ion-adsorbed rare earth ores primarily depend on embedded displacement meters, inclinometers, and similar monitoring instruments, which are employed to detect soil deformation and movement characteristics for disaster prediction24. Conventional monitoring methods collect data from single measurement points, making it difficult to capture the regional characteristics of landslide hazards28. Moreover, soil deformation and displacement data obtained from individual sensors typically emerge only during the incipient stages of slope failure, and thus cannot provide early identification of hazard sources during in situ leaching. In particular, the dynamic surge of leaching solutions—an important triggering factor of geohazards—remains undetectable at an early stage. Embedded sensors also pose additional limitations, including high installation costs and drilling-induced damage to seepage channels. These constraints highlight the urgent need for a non-destructive monitoring technology capable of wide-area coverage, real-time data acquisition, and low operational cost.

High-density resistivity method, an emerging geophysical exploration technique, offers such advantages. Based on the arrayed electrode configurations and high-precision data acquisition systems, this method enables the non-invasive, efficient, and accurate detection of subsurface structures. With its capability for synchronous multi-parameter monitoring, high-density resistivity method has been widely applied in engineering practices such as slope stability assessment, subsurface cavity detection, and groundwater exploration29,30,31,32. For example, Zhang et al. delineated the extent and location of water accumulation in a coal mine goaf in Datong, Shanxi Province, using high-density resistivity method [334]. Similarly, Li et al. investigated a surface collapse zone in a Shandong mine, integrating high-density resistivity method with shallow seismic reflection surveys, and identified the distribution of groundwater and hidden voids through drilling verification34. Han et al. further combined multi-shaft pumping and water-level recovery tests with high-density resistivity method at a landfill site, successfully characterizing the anisotropy and preferential flow patterns of leachate migration35. Collectively, these studies demonstrate that high-density resistivity method provides an effective approach to revealing the migration pathways and spatial distribution of water in soil and rock masses.

Accordingly, this study investigates a test block from an ion-adsorbed rare earth ore in southern China. A high-density resistivity method monitoring system, in conjunction with soil moisture sensors, was deployed to continuously monitor apparent resistivity and volumetric water content along a representative soil profile. Complementary laboratory tests, including humidity measurements and triaxial compression experiments on remolded soils with varying water contents, were performed to examine the relationships among the soil resistivity, water content, and shear strength parameters. By establishing a correlation model between apparent resistivity and the water content, the zones of heterogeneous infiltration and solution accumulation during the leaching process were delineated. These findings can be used to identify and assess geohazard-prone areas within the ore body, providing a scientific basis for disaster prevention and control in in situ leaching operations.

Properties of the test ore blocks and soil samples

The test ore block is located within an ion-adsorbed rare earth ore in southern China, which is representative of a typical granite weathering-crust type light rare earth ore. A comprehensive understanding of the site-specific geological and geotechnical framework is a prerequisite for accurate monitoring and analysis, a principle demonstrated in similar mining geomechanics studies36. The mining area lies in a subtropical monsoon climate zone characterized by low mountains and hilly terrain. The topography is gently undulating, with broad ridges and average slope gradients of 20°~30°. Vegetation is well developed on the slopes. Stratigraphically, the profile consists of a surface soil layer, a fully weathered layer, a semi-weathered layer, and underlying bedrock. The upper slope is dominated by the fully weathered zone, while the lower part contains relatively developed bedrock fractures. The ore body is hosted within the fully weathered section of the granite weathering crust.

Fig. 1
Fig. 1
Full size image

Experimental ore block (Noted, the base map was sourced from Earth Online: https://map.bmcx.com/ganzhou__map/).

To characterize the soil properties of the test block, moisture sensors were pre-embedded along the resistivity survey line at the injection front (Fig. 1). Boreholes were drilled at the monitoring points using a Luoyang shovel, and soil samples were collected for laboratory testing in accordance with the Standard for Geotechnical Testing Method. The basic physical properties and particle size distribution of the samples taken from a depth of 3 m in monitoring hole No. 11 were tested (Tables 1 and 2). According to the Standard for Engineering Classification of Soil (GB/T 50145 − 2007)37, the soil is classified as silty clay.

Table 1 Basic physical properties of the samples.
Full size table
Table 2 Particle size gradient of test samples.
Full size table

Experimental scheme

On-site monitoring of the high-density resistivity method

In this study, the high-density resistivity method was conducted to monitor the resistivity variations during the leaching process at a certain ion-adsorbed rare earth ore test block. The principle of high-density resistivity method is based on that subsurface resistivity anomalies in rock and soil masses can be inferred from anomalies in apparent resistivity measured at the ground surface. The apparent resistivity is calculated according to the following expression34:

$$\rho _{s} = K\frac{{\Delta U_{{MN}} }}{I}$$
(1)

Where K denotes the geometric factor of the electrode array, ΔUMN represents the potential difference between the measuring electrodes M and N, and I is the injected current supplied to the ground.

As a heterogeneous medium, the resistivity of the soil and rock varies spatially within the subsurface. Assuming that the medium is composed of n units with true resistivities ρ 1, ρ 2, ρ 3…ρ n, the apparent resistivity measured at the surface does not correspond to the true resistivity of any single unit. Instead, it reflects a composite response governed by the resistivities of the individual units and their spatial distribution. The principle of high-density resistivity method is based on these resistivity contrasts within the subsurface. By injecting current into the ground through electrodes to establish a stable artificial electric field and subsequently measuring the resulting potential field, the resistivity distribution of the subsurface medium can be delineated, thereby enabling inference of its geological structure.

Based on the ore bed thickness and the required exploration depth, the survey line configuration, number of lines, and electrode spacing were determined. Three survey lines were established (Fig. 2). Owing to topographic constraints, Survey Line 1 was oriented at an angle of 65° to the ridge, whereas Survey Lines 2 and 3 were arranged perpendicular to it. Each survey line extended 180 m in length and comprised two cable arrays, with 30 copper electrodes installed at 3 m intervals along each line. In the in situ leaching experiment, the injection wells were spaced at 2-m intervals, and a 2% magnesium sulfate solution was employed as the leaching agent. Monitoring was carried out from August 6 to November 2, 2024. Solution injection commenced on August 7 with the opening of three rows of injection valves at the mountaintop, followed by the opening of five rows on August 8 and seven rows on August 9. Between August 10 and September 22, the injection was suspended during periods of heavy rainfall but proceeded continuously under normal weather conditions. On September 23, injection resumed with all valves on the gentle slope side opened, while those on the steep slope side remained closed. Clarified solution injection began on October 7, and the leaching operation was terminated on October 30.

Fig. 2
Fig. 2
Full size image

On-site monitoring of the high-density resistivity method.

The high-density resistivity method monitoring host and substations were installed at a relatively vegetation-free location on the mountaintop. Power was supplied through a hybrid system of solar panels and storage batteries, while data transmission was achieved via a 4G network. This configuration enabled the establishment of a real-time, online monitoring system for the high-density resistivity method during the leaching process.

Three moisture sensors were installed along the survey line 3 (Fig. 3), specifically at positions 10#−1 m, 11#−5 m, and 13#−1 m. The 10#−1 m sensor corresponds to Monitoring Hole No. 10, with an installation depth of 1 m.

Fig. 3
Fig. 3
Full size image

Sensor positions along profile 3 of the survey line.

Triaxial shear test

To investigate the influence of water content on the shear strength parameters of ore-bearing soils during the leaching process of rare earth ores, remolded soil samples with varying water contents were prepared. Approximately 2 kg of undisturbed soil was collected from the test block, air-dried, crushed, and the impurities were removed. Water was then added according to the target water content, and the mixture was homogenized before being sealed in a container for 24 h. The prepared material was subsequently compacted into remolded cylindrical specimens with a diameter of 39.1 mm and a height of 80 mm using a membrane-bearing mold and compactor. Based on the natural and saturated water contents of the test block, five target water contents—10%, 15%, 20%, 25%, and 38% (the saturated condition)—were selected for testing. Three replicate specimens were prepared for each condition. Unconsolidated–undrained (UU) triaxial shear tests were conducted on the samples using a TSZ-1 strain-controlled apparatus under confining pressures of 50, 100, and 200 kPa (Fig. 4). Axial loading was applied at a displacement rate of 0.8 mm/min, and testing was terminated when either specimen failure occurred or an axial strain of 20% was reached. The shear strength parameters, including the cohesion and internal friction angle, were determined from the Mohr stress circles constructed from the test results.

Fig. 4
Fig. 4
Full size image

Triaxial shear test process (note: this figure was a composite image created by combining the photographs of the apparatus and the samples using PowerPoint software).

Humidity test

The moisture sensor employed in this study consists of a stainless steel probe and a waterproof probe head. Its working principle is based on measuring the soil capacitance between the electrodes, which varies with the dielectric properties of the medium and thus reflects the volumetric water content. During in situ leaching, changes in the liquid content of the rare earth ore layer modify the soil dielectric constant, thereby affecting the measured capacitance. It should be noted, however, that the sensor records the volumetric water content over a localized range rather than point-specific values. Owing to the low clay content of the ore soil on site, it was not possible to obtain intact samples for determining the dry density. To establish a conversion relationship between the sensor readings and the true water content, laboratory calibration tests were performed on remolded ore soils with different water contents (Fig. 5). Undisturbed soil collected from the field was first dried and then reconstituted into specimens with target water contents of 5%, 10%, 15%, 20%, 25%, 30%, 35%, and 40%, prepared in acrylic cylinders with a diameter of 39.1 mm and a height of 80 mm. The moisture sensor probe was subsequently inserted into each specimen to measure the corresponding humidity values.

Fig. 5
Fig. 5
Full size image

Humidity test (note: this figure was a composite image created by combining photographs taken by the author using PowerPoint software).

Results

Characteristics of apparent resistivity variation in the survey line profiles

Given space constraints, this paper focuses on an in-depth analysis of the resistivity distribution along the survey lines 3 obtained from the high-density resistivity method. Figure 6 presents the temporal variations in apparent resistivity along the survey lines 3. At the onset of solution injection, several rows of valves located along the mountain crest were opened to initiate leaching. The infiltrating solution migrated downslope beneath the ridge, generating a droplet-shaped low-resistivity anomaly directly below the crest. Surrounding this low-resistivity core, the resistivity increased progressively outward, with high-resistivity zones concentrated along both flanks of the slope. This distribution suggests that, during the initial stage, the leaching solution flow was largely confined to the subsurface beneath the ridge and had not yet propagated laterally into the slope flanks.

As the injection period progressed, the initial droplet-shaped low-resistivity zone expanded, gradually extending beneath the gentle slope on the left side of the ridge and evolving into a peach-shaped anomaly. In contrast, the steep slope on the right side of the ridge remained dominated by extensive high-resistivity zones, primarily because the injection wells on the gentle slope side allowed a greater volume of the leaching solution to infiltrate downward. With continued injection, the low-resistivity zone further expanded, both laterally and vertically, eventually developing into a bamboo-hat-shaped anomaly. Beneath this zone, the resistivity increased systematically with depth, likely reflecting the transition to more intact rock and soil structures with fewer fractures, which constrained further downward percolation of the solution. Notably, the blue low-resistivity anomaly beneath the gentle slope extended upward to the surface, indicating that the leaching solution had seeped out along the slope face. Field observations corroborated this interpretation, as solution overflow was detected in several injection wells. Such persistent seepage can induce soil liquefaction at the surface, thereby substantially elevating the risk of slope instability and related geohazards.

Fig. 6
Fig. 6Fig. 6Fig. 6
Full size image

Spatial distribution of apparent resistivity along survey line 3 (note: this figure was generated by the Res2dinv software.)

Monitoring the results of the sensors

The monitoring results from sensor 13#−1 m show negligible variation, indicating that the leaching solution did not infiltrate to the depth corresponding to this sensor (Fig. 7).

At the 11#–5 m sensor, the volumetric moisture content exhibited a sharp increase, peaking at 96% on September 14, before subsequently declining to its initial level of 20%. This abrupt rise was likely driven by fluctuations in the local injection intensity, indicating that the injection rate in the vicinity of this borehole underwent a short-term surge followed by a reduction. Such behavior suggests that the injection process within the test block was unstable, resulting in transient variations in the volumetric water content of the ore body. At the 10#–1 m sensor, the volumetric moisture content reached saturation (100%) on September 25. This increase can be attributed to the full opening of the injection valves on the gentle slope side on September 23, which facilitated the gradual infiltration of the leaching solution into the near-surface soil at the slope toe. As a result, the monitoring value at the 10#–1 m sensor increased sharply and subsequently remained at a consistently elevated level. As of November 2, the humidity readings at the 13#–1 m sensor remained consistently low, indicating that the leaching solution had not infiltrated to the depth of the sensor.

Fig. 7
Fig. 7
Full size image

Sensor humidity along the third profile of the measurement line.

Variation of shear strength parameters with the water content

Figure 8 illustrates the variation of the soil cohesion and internal friction angle with respect to the water content. As shown, soil cohesion initially increases and then decreases with increasing water content, suggesting the presence of an optimal water content for the ore-bearing soil. This phenomenon can be explained by the synergistic mechanism of matrix suction’s adsorption effect and capillarity. During the low moisture content stage (< 15%), as the moisture content increases, the adsorption of water molecules by the abundant clay minerals and rare earth cations in the rare earth mineral soil intensifies, leading to the formation of a progressively thickening double electric layer structure. Strong and weak bound water films develop orderly on the particle surfaces, enhancing the inter-particle bonding strength. Concurrently, the number of curved liquid surfaces in the soil pores increases, resulting in greater capillary forces and a significant rise in matrix suction, both of which contribute to the increase in apparent cohesion. When the moisture content reaches 15%, the thickness of the adsorbed water film and the distribution of capillary channels achieve an optimal balance, maximizing the synergistic effect of matrix suction and inter-particle bonding, resulting in peak cohesion. However, when the moisture content exceeds 15%, excess water causes the bound water on the particle surfaces to transition into free water, diminishing the adsorption effect due to dilution of ion concentration. Simultaneously, larger pores are preferentially filled with water, reducing the number of curved liquid surfaces, lowering capillary pressure, and leading to a sharp decline in matrix suction38,39. This dual-action mechanism explains the nonlinear response characteristics of cohesion in ionic rare earth mineral soils during variations in moisture content. Additionally, the presence of excess water at high water levels may dissolve the cementing agents and soluble salts between the particles, further weakening the interparticle bonding and accelerating the decline in the cohesion. In agreement with red-bed mudstone landslide research40, the internal friction angle of ion-adsorbed rare earth ore soil similarly exhibits a progressive reduction with increasing water content The internal friction angle also exhibits a progressive decrease with increasing water content (Fig. 8). This trend may be explained by the formation of thin water films on the particle surfaces, which reduce the direct particle-to-particle contact and act as lubricating layers, thereby lowering the interparticle shear resistance. Moreover, the dissolution of the cementing substances and salts under high- water conditions produced smoother particle interfaces, further reducing the internal friction angle of the ore-bearing soil.

The experimental results further demonstrate that under fully saturated conditions, both the cohesion and internal friction angle reached their minimum values, indicating that the soil mass exhibited the weakest shear strength and the lowest resistance to failure. At this stage, the mechanical stability of the ore-bearing slope was substantially reduced, implying a markedly increased risk of slope instability and potential geohazard occurrence in the stope area.

Fig. 8
Fig. 8
Full size image

Cohesion and internal friction angle versus water content.

Functional relationship between apparent resistivity and water content

Based on the calibration experiments conducted on remolded ore soil samples with varying water contents, the corresponding discrete data points of the volumetric water content and sensor-measured humidity were obtained. A nonlinear regression analysis was performed on these distinct data, and the fitted curve depicting the relationship between the volumetric water content and humidity is presented in Fig. 9. The empirical functional relationship between the two variables can be expressed by

$$\omega _{{{\text{humidity}}}} = {\text{2}}.{\text{88}}\omega _{0} - 0.{\text{1357}}$$
(2)

In the formula, ω0 represents the water content of the ore soil; ω humidity denotes the humidity value of the ore soil, which is the reading obtained from the moisture sensor.

Fig. 9
Fig. 9
Full size image

Relation between humidity and water content.

Previous studies by Córdova et al. demonstrated that apparent resistivity exhibits a power–law relationship with the soil moisture content41,42,43. In this study, Statistical analysis of the apparent resistivity values corresponding to different time points from the moisture sensors will be conducted on a one-to-one basis which was used to construct a scatter distribution in a two–dimensional coordinate system. The curve fitting of these data points yielded the quantitative relationship between apparent resistivity and the moisture content (Fig. 10). The functional relationship is expressed by

$$\rho = {\text{2}}.{\text{455}}*\omega _{{{\text{humidity}}}} ^{{ - {\text{4}}.{\text{8494}}}}$$
(3)

In this formula, ρ denotes the apparent resistivity, while ω humidity represents the humidity of the ore soil, as measured by the moisture sensor. Equation (3) indicates that apparent resistivity exhibits a power–law relationship with the humidity values obtained from the moisture sensors, a finding consistent with the results reported by Córdova and other scholars34.

Fig. 10
Fig. 10
Full size image

Relation between the apparent resistivity and humidity.

Combining formulas (2) and (3), the quantitative relationship between apparent resistivity and water content is derived as follows:

$$\omega _{0} = 0.{\text{4178}}\rho ^{{ - 0.{\text{2}}0{\text{62}}}} 0.0471$$
(4)

From formula (4), it can be seen that the apparent resistivity is in a power function relationship with the water content.

Discussion

Dynamic changes in the water content of ion-adsorbed rare Earth ore bodies

By deploying water sensors along high-density resistivity method survey lines at different locations and depths within the test ore block and performing nonlinear fitting between the apparent resistivity and the water content at various measurement points, a correlation model was established for this type of ore soil. Based on this model, the spatial distribution of the water content within the ore body can be inferred. The integration of high-density resistivity method monitoring with multi-source data fusion thus provides a novel approach for tracking water migration and distribution in ion-adsorbed rare earth ores. From the inferred spatial distribution of water, saturated surface zones as well as low- water regions adjacent to high- water zones can be identified. These zones are interpreted as potential liquefaction-prone areas and incipient slip bands within the ore mass. Such identification offers important technical guidance for ensuring the safety of in situ leaching operations, mitigating geohazard risks, and reducing the unnecessary loss of rare earth resources.

Law of water migration in ore bodies

The hydrogeological conditions of the mining area directly influence the transport patterns of leachate and the stability of the slopes. Drawing on integrated hydrogeological research methods44, this study systematically analyzes the interactions among geological structures, pore structures, and moisture transport, providing a hydrogeological basis for explaining the anomalies in resistivity. The variation in the infiltration area within the ore body is influenced by multiple factors, including the distribution of the surface injection holes, the topographic conditions, the injection sequence, the injection intensity, and the underlying geological structure. As shown in Fig. 6, the test ore block exhibits a gentler slope on the left side and a steeper slope on the right side. The injection holes were primarily arranged on the gentle slope, while only three rows were on the steep slope near the mountain crest. The sequence of injection began at the mountaintop and progressively extended toward both flanks. These constraints reasonably account for the observed morphological evolution of the infiltration zone—from the initial “droplet” shape to a “peach” shape, and eventually to a “bamboo hat” configuration. The “droplet” and “peach” morphologies reflect the influence of the initial injection pattern. Prior to September 11, only a few rows of mountaintop injection wells were opened, resulting in a limited infiltration extent. The leaching solution percolated downward primarily along a confined linear pathway, producing a narrow infiltration zone with little lateral expansion. During the initial stage of infiltration, the downward migration rate of the leaching solution exceeded the lateral diffusion rate, resulting in the formation of a “droplet-shaped” infiltration zone. Once the leaching agent reached the semi-weathered layer, the vertical infiltration velocity diminished, while lateral diffusion became dominant, giving rise to a “peach -shaped” infiltration pattern. With the progressive opening of the injection wells along the gentle slope, the injection area evolved from a linear to a planar distribution. Consequently, the ore body beneath the gentle slope experienced rapid infiltration, and the infiltration zone expanded into a “bamboo hat-shaped” configuration. The dynamic evolution of these infiltration patterns highlights the influence of the control of bedrock integrity on vertical fluid migration. Specifically, the closure of fractures in the underlying rock mass increases the hydraulic resistance, forcing the leaching solution to spread laterally. In certain cases, the solution exfiltrates at the ground surface, thereby elevating the risk of surficial soil liquefaction and compromising slope stability.

Research on the identification of mine risk zones

Based on the inversion results of the comprehensive resistivity survey in conjunction with the monitoring data obtained from the moisture sensors, three zones within the test block were identified as having elevated disaster risks: Risk Zone 1 (red), located on the northeast slope; Risk Zone 2 (red), situated on the southwest slope; and Risk Zone 3 (blue) within the central section of the ore body. This approach of integrating field data to produce a spatial risk zonation aligns with advanced methodologies used in other geohazard assessments45. Although no landslide incidents have occurred in these three risk areas yet, with the continued injection of fluids, landslide incidents are inevitable. For example, in coal mine surrounding rocks, over time, due to persistent disturbances from blasting, pore water pressure, and tectonic forces, the rock mass will ultimately fail46,47.

In Risk Zone 1, the upper part of the ore body shows a high-resistivity red zone, reflecting low water content, whereas the underlying blue low-resistivity zone is largely continuous, indicating that the leaching agent has migrated downslope to the mountain foot. This continuous low-resistivity belt is highly susceptible to the development of potential slip surfaces, which may induce landslide hazards. In Risk Zone 3, the blue low-resistivity area extends to the surface, confirming the infiltration of the leaching solution. Given the steep slope in this zone, insufficient drainage of pore water can easily result in soil liquefaction, further increasing the likelihood of liquefaction disasters and liquefaction-induced landslides (Figs. 11 and 12) These findings underscore the need for heightened safety awareness among workers operating in these risk-prone areas, as well as strengthened hazard inspections. Where necessary, adjustments to the injection scheme or the implementation of temporary reinforcement measures should be carried out to mitigate disaster risks. The spatial distribution of the risk zones is shown in Fig. 11.

The monitoring period of this project was relatively short, lasting only three months, and thus did not adequately capture the cumulative effects of long-term rainfall during the rainy season on slope stability. Other geophysical exploration methods also exhibit certain errors, as there are limitations in both the detection techniques and data processing methods48. Furthermore, the electrode spacing of 3 m used in the high-density resistivity method introduced certain limitations in detection accuracy, making it difficult to detect preferential flow paths through fractures or cracks at the decimeter or centimeter scale, which can be critical initiators of instability. Therefore, future studies on resistivity monitoring should extend the observation period and enhance the spatial resolution to provide a more accurate assessment of water migration and a more reliable analysis of mine slope stability. The high-density resistivity method provides a “smoothed” or “bulk” view of moisture content would provide a more nuanced interpretation of the results.

Fig. 11
Fig. 11
Full size image

Profile of the disaster-prone risk zone along survey line 3 (note: this figure is generated by the Res2dinv software).

Fig. 12
Fig. 12
Full size image

Leaching agent overflowing outside the injection hole (note: this figure was taken by the authors). a Leaching agent overflowing, b Surface infiltration.

Conclusions

This study applies high-density resistivity method prospecting technology to monitor water migration in ion-adsorbed rare earth ore ores. By deploying survey lines in the field, the temporal and spatial variation of apparent resistivity under different injection periods was revealed. Coupled with laboratory calibration tests, the functional relationship between apparent resistivity and the water content was established, and the influence of the water content on the soil shear strength parameters was clarified. This integrated approach enables efficient and accurate identification of the water content distribution within the ore body, thereby providing a scientific basis for optimizing the injection intensity and offering early-warning support for the prevention of slope instability and landslide hazards in rare earth ores.

  1. (1)

    On-site high-density resistivity method monitoring of the experimental ore blocks in ion-adsorbed rare earth ores revealed that water migration within the ore body is strongly influenced by the sequence, distribution, and depth of the injection holes, as well as by the physical properties of the ore soil and the structural characteristics of the ore body. During the early stage of liquid injection, water migration is dominated by vertical infiltration, forming a “droplet-shaped” infiltration zone. In the middle stage, lateral diffusion becomes predominant, causing the infiltration zone to expand toward the slope toes on both sides and evolve from a “peach-shaped” to a “bamboo hat-shaped” pattern. These spatial variations in infiltration are primarily attributed to the high permeability of the ore soil, the low permeability of the underlying bedrock, and the staged injection process.

  2. (2)

    Field experiments combining high-density resistivity method monitoring with moisture sensors demonstrate that this approach can not only identify zones of water migration but also quantify the distribution of water content within different apparent resistivity regions of the ore body. This provides valuable guidance for mine operators in evaluating the uniformity of seepage and making timely adjustments to the injection intensity. The experimental results further indicate that excessive injection intensity on the gentle slope side is the primary cause of leachate overflow in the surface soil layer.

  3. (3)

    On-site monitoring with high-density resistivity method and moisture sensors, combined with laboratory calibration experiments, revealed a positive correlation between the ore body water content and sensor-measured humidity, expressed as follows: ωhumidity = 2.88ω0−0.1357. Additionally, a power-function relationship was established between apparent resistivity and the water content: ω0 = 0.4178ρ−0.2062 ་0.0471.

  4. (4)

    Indoor triaxial shear tests showed that with increasing water content, the cohesion of the ore body first increases and then decreases, reaching a peak at 15% water content, while the internal friction angle decreases continuously. Once the ore body becomes saturated, both the cohesion and internal friction angle drop to their lowest levels, rendering the slope highly susceptible to failure. This highlights the necessity of strengthening mine safety inspections.

  5. (5)

    In the middle to late stages of solution injection, a distinct high–water zone adjacent to a low–water region was observed on the steep slope side of Survey Line 3. The transitional interface between these contrasting zones is highly susceptible to evolving into a potential weak plane for slope instability. Accordingly, the injection intensity on this slope should be reduced, and the slope safety inspections must be strengthened.