Rock engineering is frequently associated with various geological disasters, with landslides being the most prevalent and causing substantial damage1,2,3,4. In 2017, landslides were the third most frequent cause of mining accidents in China, both in terms of incidents and fatalities5. Many of these landslides occur in complex geological settings, where the presence of faults and structural planes within the slopes results in significant deformation during excavation6,7,8.

The occurrence of landslides is influenced by various factors, among which geological structure and lithology play a decisive role9,10,11,12,13,14. Over the past few decades, new modeling methods and techniques have been developed to study the failure mechanisms and deformation characteristics of slopes15,16,17,18,19,20. Zheng et al.21 demonstrated through numerical simulations that faults play a crucial role in shaping the formation and failure patterns of landslides in mountainous regions. Studies by Brideau and Stead22 and Huang et al.23 highlight that geological structures significantly influence landslides through the formation of separation surfaces or composite failure surfaces with low residual or peak strength, often caused by rock mass fatigue, faults, and joints. Zheng et al.24 employed UDEC to analyze the impact of fault orientation on slope stability, establishing a relationship between the safety factor of the slope and fault parameters. Further, research by Li et al.25 and Yuan et al.26 on the failure modes of anti-dip slopes in the Fushun West open-pit mine identified that structural planes and weak layers not only accelerate toppling failures but also induce shear failures. Physical modeling has also been effective in studying slope deformation. For instance, Adhikary and Dyskin27, Amini et al.28 and Zhang et al.29 demonstrated through physical models that landslides often occur due to joint displacement and connectivity. Zhao et al.30 examined the impact of faults on the stability of undercut slopes under excavation conditions using physical modeling. The results indicated that as the excavation depth increased, the size of the fault scarp and the deflection increased, leading to shear displacement of the fault, which made the slope more susceptible to bending failure. Additionally, numerous studies have examined slope failure mechanisms using mechanical experiments, deformation monitoring, and microseismic methods31,32,33,34,35,36.

Despite extensive research on landslide mechanisms, the impact of multiple faults on landslides has received limited attention. Slope landslides influenced by multiple faults present a complex geological challenge, making it difficult to fully understand the landslide mechanisms. Studying the failure sequence in such settings can help identify the key factors contributing to their formation and progression. This study focuses on a large-scale landslide, controlled by faults with multi-zone failures, on the east slope of the Nanfen open-pit mine in Benxi City, Liaoning Province, China. Preliminary analysis indicated that the faults divided the slope into distinct zones, compromising the overall integrity of the rock mass. Excavation at the base of the slope has triggered bedding sliding landslides in various regions, damaging multiple transportation routes.

To understand the evolutionary characteristics of zoning failures, we conducted extensive investigations into the geological conditions and engineering activities. InSAR monitoring was utilized to calculate the vertical and horizontal deformation time series within the study area, allowing for a detailed analysis of the deformation characteristics of these landslides. Additionally, numerical simulation analysis was employed to examine the landslide process and failure sequence, leading to the identification of critical zones. This comprehensive approach provided valuable insights into the evolution and failure sequences of the sliding bodies. We believe that our research methods and findings offer significant contributions to understanding the mechanisms of multi-regional landslides influenced by faults and in identifying potential risk areas.

The background of landslide areas

Landslide location

The Nanfen open-pit mine, located in Benxi City, Liaoning Province, China (41°05’33"N, 123°48’37"E), is the country’s largest iron ore mine, with an annual production capacity of 15 million tons, as illustrated in Fig. 1a. The mine covers an area extending 3.41 km from north to south and 1.75 km from east to west, reaching a maximum height of approximately 590 m. On June 19, 2021, a large-scale, multi-zone sequential bedding sliding landslide occurred on its eastern slope. By August 30, the maximum subsidence had exceeded 40 m. The landslide affected a zone approximately 400 m in height, 550 m in width, and 50 m in thickness, as illustrated in Fig. 1c. The total area impacted by the landslide spans 200,000 m².

A comparison of the deformation before and after the landslide is presented in Fig. 1b. During the excavation process, the right side of the sliding mass experienced initial collapse (Fig. 1b-1), while the left side exhibited tensile cracking and shear dislocations with a displacement of approximately 15 m (Figs. 1b-2, 1b-3, and 1b-4). The landslide induced numerous tensile cracks on the 710 m bench (Figs. 1b-5), and it resulted in uplift on the 298 m bench, raising it by approximately 9.7 m (Fig. 1b-6).

Fig. 1
figure 1

The general information of landslide in Nanfen open-pit mine. (a) The location of the Nanfen open-pit. (b) An overall view of the landslide, including surface deformation and cracking (b-1, b-2, b-3, and b-4). tension cracks at the rear edge (b-5), and uplift accumulation at the slope toe (b-6). (c) The panoramic view of Nanfen open-pit mine on Google Earth. The satellite imagery of (a) is obtained from ASF Date Search (https://search.asf.alaska.edu); The satellite imagery of (b) and (c) are obtained from Google Earth (https://earth.google.com).

Full size image

Geological setting of the site

The lithology within the mining area exhibits a classification of 12 distinct rock groups based on their engineering geological properties. However, within the landslide area, two primary rock types dominate: two-mica quartz schist and epidote amphibole schist, as shown in Fig. 2a and 2c. The two-mica quartz schist exhibits a dip ranging from 272° to 278°, with a dip angle ranging from 42° to 50°, and an average thickness of 400 m. In contrast, the epidote amphibole schist has a dip direction between 249° and 274°, with a dip angle varying from 27° to 57°, and an average thickness of 100 m.

Initially, only 45 faults were identified during the preliminary survey. However, as excavation progressed, additional faults were uncovered, leading to the identification of 86 new faults. This brings the total number of faults in the area to 131, as illustrated in Fig. 2a. The landslide zone is primarily affected by six significant faults, detailed in Fig. 2b and further described in Table 1.

Table 1 The fault characteristics of landslide area.
Full size table
Fig. 2
figure 2

Geological structure and mining information. (a) The plan of geological structure. (b) The enlarged map of the geological structure within the landslide area. (c) The section view of the landslide area. The satellite imagery of (b) is obtained from Google Earth (https://earth.google.com).

Full size image

Methodology

Survey of past slope failure locations

This work highlights our thorough examination of historical damage, emphasizing the characteristics of slope failures associated with varying excavation activities. By detailing the methodologies used during our site investigations, we aim to provide a comprehensive overview of how these activities have influenced the slope’s integrity over time.

Displacement monitoring

The monitoring results are fundamental for judging the landslide dynamics37,38,39. The InSAR technology is particularly effective in this context, as it provides high measurement accuracy for detecting small slope deformations and can capture large-scale displacement fields, enabling effective monitoring of deformation patterns prior to landslides4,40,41,42,43.

This study utilizes the multidimensional small baseline subset (MSBAS) technique, proposed by Samsonov44, to calculate the vertical and horizontal deformation time series in the study area. This method enhances the SBAS-InSAR technique by using overlapping temporal and spatial data from different radar sensors to derive two-dimensional deformation time series and deformation rate results. This approach not only provides a more accurate representation of the deformation within the study area but also improves the temporal resolution of the monitoring results45. The fundamental principles of this technique can be found in the referenced literature44.

Sentinel SAR data were employed in this study. Repeat orbit SAR images from 28 ascending and 26 descending passes acquired between September 2020 and June 2021 were processed to analyze the deformation of the eastern slope of the Nanfen open-pit mine. This approach allowed for the investigation of the spatial and temporal distribution as well as the evolution characteristics of ground surface deformation in the study area.

Numerical simulation

We use Rocscience’s RS3 and Rocslope to analyze the safety factors and deformation characteristics of the sliding masses in the east slope of the Nanfen open-pit controlled by faults, and obtain the failure sequence of different sliding masses.

The analyze of deformation characteristics

In this study, RS3, a three-dimensional finite element analysis program, is utilized to analyze the failure mechanisms and deformation characteristics of the slope46. This software can use the joint finite element method to analyze the established joint slope model. Its advantage is that it can consider the properties of joints and rock mass at the same time. Meanwhile, the software can quickly generate three-dimensional mechanical models and assign material parameters.

The slope under study is located within a primary rock formation, necessitating the inclusion of as much primary rock as possible in the slope model. However, incorporating extensive primary rock typically results in a model that is too large, leading to impractical computation times. To address this, the artificial boundary position in the model was determined based on the principle that the boundary should not significantly influence the slope’s failure mode25,47. Consequently, the model was set at dimensions of 1400 × 1200 × 700 m, containing 1,384,580 tetrahedral elements. Six faults were sequentially added according to their spatial orientation. To accurately represent the faults, they were modeled as fractured zones with specific thicknesses. Based on the physical properties detailed in Table 1, the faults were categorized as either hard or soft: those containing fault gouge were treated as soft faults, while those containing only breccia were considered hard faults. The interface between the fault and the rock mass was modeled as a joint element.

The model was subjected to displacement boundaries, with a fixed constraint applied at the base to prevent movement. Horizontal displacement of the rock mass was restricted by applying x-direction and y-direction constraints to the model’s left and right boundaries. The slope surface was designated as a free boundary, allowing it to move freely under the influence of gravity and other external forces, as illustrated in Fig. 3. The monitoring points (P1–P8) in the figure are primarily used to monitor the stress variation along the fault. The in-situ stress was modeled using a gravitational stress field. For this calculation, the Mohr-Coulomb elastoplastic constitutive model was selected. The mechanical parameters for various materials were derived from previous design reports and were adjusted multiple times based on computational results, as presented in Tables 2 and 3.

Fig. 3
figure 3

The analysis model for deformation characteristics.

Full size image
Table 2 The mechanical parameters of lithology.
Full size table
Table 3 The mechanical parameters of slope joints.
Full size table

Limit equilibrium analysis of 3D blocks and analysis of failures sequence

RocSlope is a 3D limit equilibrium slope stability program for evaluating the safety factor of structurally-controlled block failures in rock slopes. By employing block kinematics principles and failure criteria, RocSlope rapidly identifies potential sliding areas on the slope. The software also determines the location and depth of the critical block based on the changing characteristics of the failed block. The failure criterion in RocSlope is based on the Mohr-Coulomb criterion, where the sliding force is primarily influenced by the weight of the rock mass, and the anti-slip force is mainly provided by the strength of the structural plane. Fault data are incorporated into the RocSlope model by importing them individually and designating them as joint elements. Similarly, measured sliding surfaces are imported as a joint element to form the blocks. In the case of the Nanfen open-pit mine, this approach resulted in the formation of seven distinct blocks, as shown in Fig. 4.

Fig. 4
figure 4

The analysis model for safety factors and failure sequence.

Full size image

Results

Evolution of excavation locations and historical landslide relationships

Figure 5a is the sectional view of the landslide area. The study area has experienced several landslides, with the temporal development of these events from 2017 to 2021 documented through satellite imagery and on-site investigations. In 2017, expansion and excavation on the eastern slope of the mining area reached the 434 m bench. This led to a localized surface landslide on the right side of the 434–554 m bench. However, the failure did not extend downward due to the stabilizing influence of the 54 m wide platform at the 434 m bench, as shown in Fig. 5b.

In 2018, further expansion of the eastern slope to the 410 m bench resulted in another surface landslide in the same region, affecting the 410–574 m bench. Once again, the 48 m wide platform at 410 m bench prevented the landslide from progressing downward, as depicted in Fig. 5c. Figure 5d shows the landslide area in 2020. By 2021, additional excavation extended the slope to the 298 m bench, triggering a similar surface landslide along the 298–574 m bench. Like previous occurrences, the landslide did not penetrate vertically due to the 40 m wide platform at the 298 m bench, shown in Fig. 5e. However, on June 19th, a significant landslide affected the entire slope, with deformation continuing until August, as illustrated in Fig. 5e.

Throughout these events, three main types of cracks were observed: tension cracks, tension-shear cracks, and swelling cracks, as illustrated in Fig. 5f. Tension cracks predominantly appeared above the 646-meter bench, in areas where passive deformation from large landslides or secondary sliding occurred, indicating downward movement. Tension-shear cracks mainly developed on the 358–646 m bench, especially on the northern side of the landslide. These cracks were extensive and wide, significantly influencing the northern boundary of the landslide. The central portion of the sliding mass showed fewer cracks, with the rock mass remaining relatively intact. The cracks here typically inclined northward, suggesting stress and deformation directed southward and downward. Swelling cracks were mostly found on the 298 m bench, forming undulating patterns that spread from the central area to the outer edges of the landslide, with deformation oriented northwestward.

During the failure process, the cracks are mainly divided into three types: tension cracks, tension and shear cracks, and swelling cracks, with their distribution shown in Fig. 5f. Tension cracks mainly occur on slope above the 646 m bench. These cracks primarily occur in areas subjected to passive traction deformation induced by extensive landslides or secondary sliding, with the deformation direction indicating a downward movement. The tension and shear cracks primarily materialize on the 358–646 m bench, particularly on the northern side of the landslide. These cracks exhibit prolonged extents and notable aperture, exerting significant control over the northern boundary of the landslide. The central portion of the sliding mass experiences fewer cracks compared to the northern side, and the rock mass remains relatively intact. These cracks are characterized by a prevalent oblique northward inclination, signifying that stress and deformation are directed southward and downward. Swelling cracks predominantly emerge on the 298 m bench, appearing as undulations extending from the central area towards the outer periphery of the landslide, with deformation oriented northwestward.

Fig. 5
figure 5

Investigation of landslide process and crack characteristics. (a) The sectional view of the landslide area. (b), (c), (d), and (e) show images of the landslide from different periods. (f) The cracks characteristics of landslide. The satellite imagery is obtained from Google Earth (https://earth.google.com).

Full size image

The result of monitoring

Figure 6 presents the InSAR deformation results for the east slope of the Nanfen open-pit mine, covering the period from September 5, 2020, to June 19, 2021. Figure 6a shows the vertical displacement. By June 19, 2021, the southern side of the 434–574 m bench exhibited the most significant subsidence, approximately − 100 mm. In contrast, the northern side of the 298 m bench showed uplift, with a maximum displacement of + 30 mm.

Starting in September 2020, initial deformation was observed in the rock mass near the landslide area on the 434–574 m bench, with the displacement gradually increasing at an average rate of −3 mm per month. By November 2020, the maximum subsidence reached − 10 mm. From December 2020 onwards, the rate of subsidence accelerated, and the deformation zone expanded from south to north. By May 2021, the subsidence rate had peaked at approximately 15 mm per month.

In January 2021, the deformation area began to shift upwards, followed by significant subsidence of −30 mm on the 646–710 m bench in February. As the subsidence area and rock mass displacement increased, uplift was observed on the northern side of the 298 m bench. Figure 6b highlights the horizontal displacement, showing that the horizontal deformation trend mirrors the vertical deformation pattern but with greater magnitude in horizontal displacement. The InSAR monitoring results indicate that the right side of the 434–574 m bench was the first area to experience failure, serving as a critical trigger point for multiple landslides. The failure in this region led to significant deformation in other areas.

Fig. 6
figure 6

The monitoring results in InSAR. (a) The vertical displacement of InSAR. (b) The horizontal displacement of InSAR.

Full size image

The numerical results

The deformation characteristics and failure mechanism

Figure 7 presents the simulation results depicting the entire slope failure process. The right side of the 434–574 m bench serves as the primary deformation zone, while the left side predominantly experiences tensile shear failure. Additionally, tensile failure is observed at the slope crown, aligning with on-site survey findings.

Figure 7a illustrates the damage evolution of the landslide. In the initial stage, the zone comprising the F78 fault and quartz vein exhibits relatively high fracturing and low strength, primarily undergoing shear failure and forming an initial sliding mass. The formation of this sliding mass is the primary triggering factor for the overall slope landslide, with significant deformation concentrated in this zone (434–574 m bench). In the intermediate stage, influenced by the initial sliding mass, other faults gradually become activated, leading to tensile shear failure. The damaged area continuously expands towards the left and upward from the initial landslide zone, resulting in the emergence of multiple tensile and shear cracks on the left side. In the final stage, the upper rock mass of F60 fault experiences damage and tensile failure, culminating in the complete failure of the entire slope.

Figure 7b depicts the deformation evolution of the entire landslide, aligning with the observed failure pattern. In the initial stage, deformation first occurs on the right side of the 434–574 m bench, where the faults governing this area experience notable shear deformation. In the intermediate stage, influenced by the initial sliding body, other faults also begin to deform significantly, with the deformation zone expanding leftward and upward. In the final stage, the maximum shear strain in all faults controlling the slope shows significant changes. Simultaneously, the upper rock mass of F60 fault undergoes tensile failure and experiences large displacement, leading to the complete sliding of the slope. The calculated displacement field is consistent with InSAR observations, and the trend of displacement changes during the failure process is consistent with the actual outcomes. This further validates the effectiveness of the numerical simulation in replicating the initial deformation pattern and capturing the subsequent evolution of deformation throughout the entire failure process.

Fig. 7
figure 7

The failure evolution of landslide. (a) The damage evolution of landslide. (b) The deformation evolution of landslide.

Full size image

Figure 8 illustrates the variation in the differential principal stress (\(\sigma _{1}\)) at monitoring points along various faults during the computational process of different landslide stages. Specifically, P1 and P2 monitor stress changes in the F78 fault, P3 monitors the F60 fault, P4 monitors the F7 fault, P5 and P6 monitor the Fn-2 fault, and P7 and P8 monitor the Fn-1 fault. In the Fig. 8, negative values on the y axis indicate stress reduction, while positive values represent stress increase. The x axis corresponds to the computational steps at different landslide stages as shown in Fig. 7. Based on the failure process in Fig. 7, stress reduction is primarily observed at P1 and P2 during the initial stage of the landslide. The upper section of the F78 fault fails first, leading to a stress decrease at P1. Simultaneously, the failure in the upper section causes stress concentration in the lower section, resulting in stress increase at P2. As the computational steps progress, failure propagates downward along the fault, resulting in a significant stress reduction at both P1 and P2.

During the intermediate stage of the landslide, failure expands laterally from the left side, as reflected by stress reduction at P4 and P5-P6. For the Fn-2 fault, failure initiates at the lower section and propagates upward, with stress reduction at P6 preceding that at P5. The upward propagation of failure also causes stress reduction at P4 (F7 fault) to occur at a later stage. In the final stage, the Fn-1 fault exhibits bottom-to-top failure, with stress reduction at P7 preceding that at P8. This sequential pattern of stress reduction along the faults closely aligns with the landslide failure process. By analyzing variations in fault stress differentials, distinct failure stages can be identified, enabling effective predictions of the landslide evolution. This provides a crucial basis for monitoring and forecasting landslide hazards.

Fig. 8
figure 8

The stress differential of the fault between computational steps.

Full size image

The three-dimensional block safety factor and failure sequence

Figure 9 illustrates the three-dimensional block safety factor and failure sequence. Figure 9a showcases the safety factor contour when different zones experience movement. In this calculation, Zone 1 corresponds to the initial landslide area. When the safety factor of Zone 1 exceeds 1.15, no movement is observed. The safety factors of other zones are well above 1.5, indicating overall slope stability, as shown by the red curve in Fig. 9b. However, when the safety factor of Zone 1 drops below 1.15, movement initiates in that zone. The displacement of Zone 1 reduces lateral constraints on the right side of Zone 2, creating a free surface on the lower right side of Zone 2 and decreasing its anti-slip force. Consequently, the safety factor of Zone 2 decreases to 1.015, leading to movement in that zone. The displacement of Zones 1 and 2 further reduces constraints below Zone 3, significantly decreasing its anti-slip force and causing the safety factor of Zone 3 to drop sharply to 0.508, resulting in movement.

The displacement of Zones 1 and 2 also lessens constraints on Zone 4. However, since Zone 4 is the primary anti-slip section of the major landslide, its anti-slip force remains greater than the sliding force, maintaining a safety factor of 1.158 and preventing movement. As Zone 4 remains stable, the safety factors for the remaining zones stay above 1.5, as depicted by the blue curve in Fig. 9b. However, if the safety factor of Zone 4 falls below 1.15, movement occurs in that area as well. The displacement of Zone 4 significantly reduces the resistance of Zones 5 and 6, leading to safety factors of 0.503 and 1.105, respectively, and resulting in movement in those zones. Subsequently, the safety factor of Zone 7 decreases to 0.505, causing movement in that area, as illustrated by the green curve in Fig. 9b.

This analysis demonstrates that Zone 1 is a critical block, with its failure governing the movement of Zones 2 and 3. Zone 4 serves as the primary anti-slip zone, and its movement ultimately triggers slope failure, controlling the movement of Zones 5, 6, and 7. The sequence and direction of landslides within the entire slope are depicted in Fig. 9c.

Fig. 9
figure 9

The three-dimensional block safety factor and failure sequence. (a) The safety factor contour when zones move. (b) The safety factors for different zones. (c) The sequence and direction of landslides in different zones.

Full size image

Discussion

This study aims to deepen the understanding of the evolutionary characteristics of multi-regional landslides, particularly those controlled by faults. To achieve this, we conducted extensive investigations into geological conditions and engineering activities, utilizing InSAR monitoring to calculate vertical and horizontal deformation time series in the study area for a detailed analysis of landslide deformation characteristics. Additionally, we employed numerical simulation analysis to explore the landslide process and failure sequence, identifying critical areas. This comprehensive approach provides important insights into the evolution and failure sequences of multi-regional landslides.

Accurate deformation monitoring is essential for analyzing the failure mechanisms and deformation characteristics of slopes37,38,48. Environmental factors such as blasting, excavation, and rainfall can affect the data from individual measurement points, making sudden changes at a single point unreliable indicators of slope instability. Therefore, it is more effective to focus on substantial changes in regional monitoring data5. With the rapid advancement of modern surveying and mapping technologies, various regional monitoring methods have been developed and applied in mining, including S-SAR49, D-InSAR50, UAV51, and others. In this study, InSAR monitoring technology is employed to analyze the deformation characteristics of multi-zone failures. Significant changes in overall deformation indicate the presence of a penetrating sliding fracture surface, with the areas exhibiting early and pronounced deformation serving as the dominant regions of the landslide.

Simulation results indicate that slopes controlled by structural planes become unstable due to initial displacement and the failure of weak structural planes. This process reduces the lateral constraints and sliding resistance of the landslide mass, ultimately leading to slope failure, and the initial landslide damage inevitably leads to a decline in the stability of the overlying rock mass, consistent with previous studies25,29,52. Compared to previous research, our work demonstrates significant innovations in several areas: First, the study emphasizes the detailed analysis of landslide sequences and the identification of key areas. By accurately determining the sequence of block movements, we provide crucial insights into how faults interact to gradually destabilize large slopes. This sequence analysis, particularly the identification of key hazard zones (Zone 1) and critical stabilizing areas (Zone 4), is valuable for developing targeted reinforcement strategies. Additionally, these observations confirm that large-scale landslides involve continuous deformation and significant sliding forces throughout the initiation process. Consequently, we summarize the necessary and sufficient conditions for the development of multi-belt landslides as follows: (1) the formation of a continuous sliding surface; (2) adequate allowable sliding space for the sliding body; and (3) sufficient sliding force throughout the landslide initiation process.

Although this study significantly enhances the identification of landslide sequences and critical areas through 3D modeling and InSAR dynamic monitoring, some limitations still exist. First, InSAR monitoring may be insufficient in capturing rapid deformations, particularly as crucial data during instantaneous deformation phases may be overlooked53. Additionally, the assumption of material uniformity in numerical simulations may not align with the heterogeneous actual geological conditions, potentially affecting the accuracy of simulation results. Therefore, based on these limitations, the study results should be interpreted cautiously and applied in conjunction with field realities. Future research will focus on enhancing the temporal resolution and spatial coverage of monitoring technologies. We plan to employ slope radar for all-weather detailed monitoring to capture real-time dynamic deformations of the slopes. This method will compensate for the shortcomings of InSAR monitoring in detecting rapid deformations and provide more detailed deformation information. It is recommended to consider the heterogeneity and attenuation of geological materials in numerical simulations to improve model accuracy under complex geological conditions54.

Moreover, the painful lessons from this landslide incident highlight the importance of rapidly identifying hazardous areas for mining planning. For mine safety management, annual assessments of hazardous areas are essential, necessitating a comprehensive technological approach that includes high-precision UAV modeling, automated structural plane identification, mechanical mechanism analysis, and real-time dynamic deformation monitoring. This comprehensive understanding of potential slope instability factors and early deformation trends provides a scientific basis for safe mining practices and effectively reduces the risk of landslides.

Conclusion

This study employs a comprehensive approach, integrating geological surveys, InSAR monitoring techniques, and numerical simulation, to investigate the failure process and deformation characteristics of multi-zone landslides controlled by faults. The main conclusions are as follows:

  1. 1.

    The primary mechanism driving this landslide involves multiple faults that divide the slope into distinct zones, compromising the integrity of the slope’s rock mass. Excavation activities at the slope toe lead to unloading and stress release in the rock mass. Under shear stress from the upper slope, a shear outlet forms at the base of the sliding body. As excavation progresses, the initial landslide (Zone 1) experiences surface collapse with the formation of multiple benches, significantly influencing the deformation of the entire landslide.

  2. 2.

    After the initial landslide, the rock mass on the right side of the 434–574 m bench becomes the primary sliding zone in the subsequent landslide area, showing early sliding and significant deformation. The area above the 574 m bench experiences tension, while the area below the 434 m bench serves as the primary anti-sliding zone.

  3. 3.

    Monitoring data and numerical simulations indicate that during multi-zone landslides, the initial failure area not only provides displacement space for subsequent landslides but also reduces lateral constraints in adjacent areas, leading to a decrease in their anti-sliding force and a sharp decline in their safety factors.

  4. 4.

    Slopes controlled by structural planes require detailed geological exploration and mechanical analysis to assess potential failure sequences. By integrating InSAR monitoring with UAV oblique photogrammetry, we can accurately pinpoint critical failure zones and develop targeted reinforcement strategies and construction methods, effectively reducing the risk of large-scale landslides. This approach provides essential scientific support for slope stability management under complex geological conditions.