Mining-Induced Deformation and Slope Stability in Steep Mountainous Areas Based on InSAR Monitoring and Rock Movement Theory: A Case Study from Southwestern China

Highlights

What are the main findings?

• Mining-induced deformation in extremely steep mountainous terrain is jointly controlled by mining depth, slope gradient, and structural plane configuration, forming a topography–structure–mining coupled mechanism.
• InSAR-derived deformation boundary angles exceed theoretical predictions, indicating that complex topography and rock mass structure constrain deformation propagation.

What are the implication of the main findings?

• Traditional rock movement theory has limited applicability in extremely steep mountainous conditions and may misestimate deformation influence ranges.
• A protective coal pillar (~160 m) can effectively reduce the transmission of mining-induced stress toward steep slopes and mitigate impacts on existing landslides.

Abstract

Geological disasters are frequently triggered in steep mountainous mining areas due to the coupling effects of underground excavation and slope stability, yet the applicability of traditional rock movement theories in such terrains remains unclear. This study investigates an extremely steep coal mine in southwestern China, integrating engineering geological surveys, unmanned aerial vehicle (UAV) measurements, InSAR monitoring, and rock movement theoretical calculations to analyze the impact of mining on mountain deformation and slope stability. The results show that the study area exhibits steep slopes (55–85°) and gently inclined, reverse-layered rock masses controlled by structural fracture zones, creating a geological environment prone to mining-induced landslides. The 1151 working face lies at a depth of 286–470 m, with a protective coal pillar of approximately 160 m left between the excavation and the cliff zone. InSAR monitoring indicates cumulative LOS deformation rates of −0.98 to 0.55 cm/a, with subsidence concentrated above the working face, while existing landslides in the cliff zone show no significant deformation. Comparison between theoretical calculations and InSAR inversion reveals that InSAR boundary angles (downslope 61–68°, upslope 67–73°) exceed theoretical predictions (downslope 48–52°, upslope 55°), indicating that complex topography and rock mass structure constrain mining-induced deformation propagation. The findings demonstrate that appropriately designed protective coal pillars and avoidance of unstable slopes can effectively mitigate the impact of mining-induced disturbances on existing hazards. This study provides valuable reference for landslide risk assessment and disaster prevention in extremely steep mining regions.

1. Introduction

Coal remains an important energy resource supporting industrial production and socioeconomic development [1,2]. China has long been one of the world’s leading coal producers, and the southwestern region plays an important role in the national coal supply system [3,4,5]. Coal-bearing strata are widely distributed in this region, which tectonically belongs to the Yangtze Platform. Influenced by multiple phases of tectonic uplift and intensive river incision, the region has developed complex geomorphological landscapes dominated by karst mountains. Extremely rugged terrains, including steep mountains, deeply incised valleys, and high cliffs, are widely distributed [6,7]. Under such complex topographic and geological conditions, underground coal mining may exert more pronounced disturbances on the regional geological environment.

In recent years, with the widespread application of mechanized longwall mining technology, coal extraction has been characterized by increased mining height, wider working faces, faster advance rates, and higher mining intensity [8,9]. Large-scale underground mining alters the stress state of the overlying strata, leading to rock mass fracturing, displacement, and stress redistribution, which may further cause structural instability of the overburden and surface deformation [10,11]. In the mountainous regions of southwestern China, where topographic relief is significant, mining-induced disturbances often interact with natural slope conditions, increasing the likelihood of geological hazards such as rockfalls, landslides, debris flows, ground fissures, and surface subsidence [12]. Several typical landslide and rockfall events associated with historical coal mining have been reported in the steep mountainous areas of southwestern China, highlighting that geological safety issues arising from the coupling of mining disturbances and complex terrain should not be overlooked [13,14].

Previous studies have mainly focused on mining subsidence characteristics and hazard prevention technologies in plains or gently inclined terrains, where the theoretical framework has become relatively mature. According to classical mining subsidence theory, the overlying strata after underground extraction are generally divided into several zones, including the caving zone, fracture zone, and bending zone [15]. However, most theoretical models and engineering specifications are developed under the assumption of approximately flat ground surfaces, with limited consideration given to the influence of complex topography on the spatial distribution of rock movement fields and stress transfer paths. Compared with flat terrains, mountainous areas are strongly controlled by rugged topography and slope structures, and surface subsidence often exhibits pronounced spatial and temporal heterogeneity and asymmetry. As a result, the deformation mechanisms of mountainous slopes are more complex, and the uncertainty in hazard prediction is significantly increased [16]. In particular, under extremely steep mountainous conditions, the coupling relationship between goaf collapse effects and slope engineering geological structures remains insufficiently understood. In many mountainous mining areas, old goafs formed by roadway mining or room-and-pillar methods are commonly left beneath slopes with typical “boot-shaped” geomorphological features. With the expansion of disturbance range and intensity associated with modern mechanized longwall mining, the superimposed effects of new and old goafs further aggravate slope stability problems. Therefore, issues such as the stability of existing goafs, the rational design of protective coal pillars under re-mining conditions, and the influence of mining-induced activities on surface deformation and geological hazard evolution have become key engineering geological challenges that urgently need to be addressed for safe mining and sustainable resource utilization in mountainous coal mines.

In terms of deformation monitoring, traditional techniques such as Global Positioning System (GPS) and total station measurements have been widely used for observing surface deformation in mining areas. However, these methods are often limited by sparse monitoring point distribution, relatively high costs, and intensive fieldwork, making it difficult to achieve large-scale and continuous dynamic monitoring [17,18]. In recent years, Interferometric Synthetic Aperture Radar (InSAR) has been increasingly applied in studies of mining subsidence and slope stability due to its wide spatial coverage, high measurement precision, capability of retrieving historical deformation information, and all-weather observation capability [19,20,21]. In particular, the Stacking-InSAR method can effectively extract long-term deformation trends, providing valuable technical support for revealing surface evolution processes under mining disturbances [22,23,24]. However, in mountainous regions of southwestern China, the high vegetation coverage often reduces radar coherence, while complex meteorological conditions may introduce significant atmospheric delay effects. As a result, deformation results derived from a single InSAR analysis may still contain uncertainties [25,26]. Therefore, it is necessary to integrate InSAR observations with field engineering geological investigations, theoretical analysis of mining-induced rock movement, and other monitoring approaches to improve the reliability of deformation identification and hazard assessment in mountainous mining areas [27].

Based on the above background, this study selects a coal mine located in an extremely steep mountainous area in Zhenxiong County, Zhaotong City, Yunnan Province, southwestern China, as the study site. The research focuses on the coupling mechanism between the goaf–slope system and the development characteristics of surface geological hazards. A comprehensive methodological framework integrating field engineering geological investigation, Unmanned Aerial Vehicle (UAV)-based topographic survey, InSAR deformation monitoring, GIS spatial analysis, and theoretical calculations based on rock movement theory was employed. Using these multi-source approaches, the spatiotemporal evolution of mining-induced surface subsidence from August 2024 to April 2025 was analyzed, and the response relationship between mining activities and surface geological hazards was investigated. This study develops a coupled analysis method for mining-induced rock movement and slope stability in complex high-steep mountainous geological settings through the integration of multi-source data and rock movement theory. The results are expected to provide a scientific basis and technical support for safety evaluation of re-mining activities and geological hazard risk prevention, while also offering theoretical insights for the coordinated development of geological safety and ecological environment in mining areas of southwestern China.

2. Study Area

The study area is located on the northern margin of the Yunnan–Guizhou Plateau, within a tectonic slope zone on the left bank of Labagou Valley. It lies in the transitional zone between an ancient alluvial fan and a tectonic slope. The original slope gradient ranges from approximately 15° to 20°. The study area covers about 5.99 km², with elevations ranging from 1205 to 1770 m (Figure 1a,b). The terrain exhibits a distinct stepped morphology. A pronounced slope break with gradients of 40–50° occurs near the valley flank. At the mountain ridge front, a tectonic unloading zone has developed, forming near-vertical rock slopes with inclinations greater than 60°. The lower part of the slope, extending approximately 150–300 m downslope, is mainly composed of multi-stage colluvial and landslide deposits. The regional structural trend is predominantly NE–SW, and the overall topography is characterized by higher elevations in the northwest and lower elevations in the southeast.

Tectonically, the study area is located on the southeastern limb of the Mahe Syncline within the northeastern Yunnan fold belt. The geological structure is characterized by a monocline, with bedding dip angles ranging from 5° to 30°. No major faults have been identified within the study area. However, four small normal faults and two sets of steeply dipping shear joints were observed, which may locally influence slope stability. The exposed strata in the area mainly include Triassic argillaceous limestone (T₁y), Triassic siltstone (T₁f), Permian limestone (P₁m), Permian basalt (P₂β), and coal-bearing Permian fine sandstone (P₂x). Overall, the geological structure of the area is relatively simple, and the strata exhibit stable attitudes (Figure 1c). Climatically, the study area belongs to the plateau subtropical monsoon climate zone of the northeastern Yunnan highland. The region is characterized by relatively low temperatures, high humidity, frequent cloudy and rainy weather, a long frost period, and abundant precipitation with distinct seasonal variations.

Two coal seams, C₅ and C₆, are developed in the study area and are both mineable. According to field mining conditions, the C₅ and C₆ seams are mined as a combined seam during production, with an average thickness of approximately 5 m and a gangue interlayer thickness of about 20–50 cm. The coal seam dip angle is approximately 8°. Coal resource exploitation in this area has a long history, although the mining scale has generally been relatively small. Since 2024, three working faces—1151, 1155, and 1154—have been arranged along the steep mountain slope zone for mining operations. Part of the 1151 working face had previously been mined for the C₆ seam between 2013 and 2014. The remaining portion of the 1151 working face began extraction in August 2024 and is scheduled to be completed in August 2025. All working faces adopt the longwall mining method, and the roof is managed using the full caving method.

3. Data and Methods

This study adopts an integrated research approach combining field investigation, rock mass movement theory analysis, and InSAR deformation monitoring to systematically evaluate the effects of underground coal mining on mountain stability and existing geological hazards in extremely steep mountainous areas (Figure 2). First, based on coal mine exploration reports and regional engineering geological background data, the formation conditions and controlling factors of mining-induced geological hazards in the study area were analyzed. Field investigations and UAV aerial surveys were conducted to obtain high-precision topographic data, from which a Digital Elevation Model (DEM), Digital Orthophoto Map (DOM), and realistic three-dimensional model of the study area were constructed. These datasets were further used to identify and interpret the geomorphological characteristics of the mountain terrain and the failure features of geological hazard bodies. Second, the surface influence range of the goaf was calculated based on the theory of mining-induced rock mass movement. In addition, surface deformation information was retrieved using InSAR time-series techniques to obtain the spatial distribution, influence extent, and deformation intensity characteristics of mining-induced ground movement. Finally, by comparing the InSAR monitoring results with the theoretical calculations of mining subsidence, the surface deformation response and associated hazard risks caused by the goaf under the condition of protective coal pillars were evaluated. Furthermore, the potential impacts of future mining activities on the stability of extremely steep mountain slopes and the evolution of existing geological hazards were comprehensively analyzed. The data sources used in this study are summarized in

Table 1. Data types and parameters.

Data Type	Data Source/Processing Method	Data Parameters
InSAR	Sentien-1, Stacking method	Spatial resolution: 5 m × 20 m Temporal resolution: 12 days Accuracy: ±5 mm/year
UAV aerial survey	DOM, DSM, Photogrammetric Model	Ground resolution: 0.05 m Forward overlap: 80% Side overlap: 70%
Geological map	1:10,000 regional geological survey	Stratigraphic units: J1z, J2s, P1m, P1q, P2β
Coal mine exploration report	Borehole data, geophysical logging	Lithology: sandstone, mudstone, limestone, coal seams
Geological map	Underground roadway mapping	Borehole spacing: 500 m × 500 m

.

3.1. Engineering Geological Investigation and UAV Data Acquisition

To systematically understand the characteristics of historical coal mining activities, regional engineering geological conditions, and the current development status of geological hazards in the study area, a multi-stage and systematic field investigation was conducted from September 2024 to April 2025. Through field reconnaissance and the collection of historical data, fundamental information regarding mining methods, mining extents, and extraction elevations during different periods was compiled. Existing geological exploration data and geological hazard investigation reports were also integrated to establish a fundamental database describing the evolution of mining activities and the geological environment in the study area. In addition, based on the 1:10,000 digital geological map and field engineering geological surveys, the stratigraphic classification, lithological boundaries, and structural characteristics of the study area were verified and revised. At key locations, in situ measurements of stratigraphic occurrence were carried out to obtain parameters including strike, dip direction, and dip angle of rock strata, while the structural characteristics and tectonic features of the rock layers were systematically recorded. Field verification and positioning surveys were also conducted for existing geological hazard bodies in the study area, such as ground fissures and collapse masses, in order to obtain information on their types, scales, and spatial distribution.

To obtain high-precision terrain and surface morphology data, low-altitude photogrammetry was carried out using a DJI M300 (DJI, Shenzhen, China) unmanned aerial vehicle (UAV). The flight was conducted at a relative altitude of 240 m with a speed of 10 m/s, while the forward and side overlaps of the flight routes were set to 80% and 70%, respectively. A total of 3219 valid aerial images were acquired under clear weather conditions with no cloud cover and low wind intensity. The collected images were processed using DJI Terra (Version 5.0.2) software for aerial triangulation and three-dimensional model reconstruction. Subsequently, a digital orthophoto map (DOM), a digital surface model (DSM), and a realistic three-dimensional model with a spatial resolution of 0.05 m were generated [28].

3.2. Spatial Characteristics Analysis of Slope Gradient and Rock Mass Structure

To clarify the influence range of mining activities and the controlling factors of slope stability, this study conducted spatial extraction and comprehensive engineering geological analysis of key parameters, including mining depth, slope gradient, and slope structure, based on DEM data, geological maps, and mining engineering data. First, regarding the acquisition of mining depth parameters, a three-dimensional spatial relationship model between the ground surface and the coal seam floor was constructed based on the DSM generated from UAV surveys and the terrain elevation model of the mining area, combined with underground mining engineering plans and coal seam burial depth data. Through spatial overlay analysis, the difference between the surface elevation and the mining elevation of the coal seam was calculated to obtain the spatial distribution map of mining depth, thereby achieving a rasterized representation of mining depth. The calculation formula for mining depth is expressed as follows:

$$H\left(x,y\right)=Z_s\left(x,y\right)-Z_c\left(x,y\right)$$

(1)

where $H\left(x,y\right)$ represents the mining depth at any location, $Z_s\left(x,y\right)$ denotes the ground surface elevation, and $Z_c\left(x,y\right)$ represents the mining elevation of the coal seam at the corresponding location. Through spatial statistical analysis, the variation characteristics of mining depth across different slope sections and within different working face areas were identified.

Subsequently, regarding the extraction of the slope factor, terrain analysis was performed using high-resolution DEM data. The slope raster dataset was calculated using the GIS spatial analysis module, and slope units were further delineated by integrating slope aspect information [29]. The slope was calculated using the following expression:

$$S=\mathit{arctan}\left(\sqrt{p^2+q^2}\right)$$

(2)

where $p$ and $q$ represent the elevation change rates of the terrain in the x and y directions, respectively. Based on slope classification statistics, the distribution characteristics of mining depth and the differences in hazard development within different slope gradient intervals were analyzed, enabling the identification of areas where steep slopes overlap significantly with mining-induced disturbances. Regarding the analysis of slope structure and engineering geological conditions, stratigraphic occurrence, lithological assemblages, and the distribution of structural lineaments were obtained based on the 1:10,000 geological map and field measurement data. A slope structural model was then established using spatial overlay analysis. Particular attention was given to the geometric relationship between the dip direction of rock strata and the slope aspect, including dip slopes, anti-dip slopes, and oblique slopes, in order to identify potential structurally controlled instability modes. Furthermore, by integrating the spatial relationship between fault structures and goaf distribution, the controlling effects of structural weak planes on stress redistribution and the expansion of the mining-induced rock movement field were analyzed.

3.3. Calculation of the Influence Range of Mining-Induced Rock Movement in the Goaf

According to the general laws of mining subsidence, as the scale of goaf excavation expands, mining-induced disturbances are progressively transmitted to the overlying strata, resulting in a continuous expansion of the range of rock mass movement and deformation [30,31]. When the working face advances to a certain scale, the movement of the overlying strata evolves from localized bending failure to overall subsidence and structural reorganization, eventually propagating to the ground surface and causing surface subsidence, tilting, and curvature deformation [32,33,34]. After the longwall working face has been fully mined and the movements of the overlying strata and ground surface gradually stabilize, a subsidence basin typically forms on the ground surface with a spatial extent larger than the projected mining area. The outer boundary of this basin can be characterized by the surface boundary angle [35,36]. The more sufficient the mining extraction and the more complete the surface movement process, the larger the spatial influence range of the subsidence basin [37]. Previous studies have shown that the influence range of surface rock movement is mainly controlled by factors such as mining depth, mining thickness, overburden structure, and movement angle parameters. Under conditions of sufficient mining disturbance, the boundary of the rock movement influence zone can be estimated using the geometric relationships of the boundary angle or movement angle [38].

Let $H$ denote the mining depth, $L_0$ the projected length of the working face along the strike direction, and $B_0$ the projected width along the dip direction. Let $\delta$ represent the boundary angle in the strike direction, while ${\gamma }_0$ and ${\beta }_0$ denote the uphill boundary angle and downhill boundary angle, respectively. Under these conditions, the influence range of mining-induced rock movement can be calculated based on geometric extension relationships.

The influence length along the strike direction is expressed as:

$$L=L_0+2H\mathit{cot}{\delta }_0$$

(3)

The influence width along the dip direction is expressed as:

$$B=B_0+H\mathit{cot}{\gamma }_0+H\mathit{cot}{\beta }_0$$

(4)

where $H\cot {\gamma }_0$ and $H\cot {\beta }_0$ represent the influence distances extending beyond the projected working face in the uphill and downhill directions, respectively. For inclined coal seams, the boundary angle in the downhill direction decreases with increasing coal seam dip angle and can be corrected using the following expression:

$${\beta }_0={\delta }_0-0.5\alpha$$

(5)

where $\alpha$ represents the dip angle of the coal seam. This correction relationship reflects the influence of the seam dip on the asymmetric expansion of the rock movement influence range.

3.4. InSAR Deformation Monitoring and Data Processing

Time-series InSAR technology is an important approach for retrieving surface deformation information using multi-temporal SAR data and enables precise monitoring of large-scale ground deformation [39]. To obtain stable surface deformation characteristics in the study area, the Stacking-InSAR method was adopted to process and analyze multi-temporal SAR data. A total of 29 Sentinel-1 ascending SAR images acquired from the European Space Agency (ESA) between May 2024 and May 2025 were used in this study, with a minimum revisit interval of 12 days. The image acquired on 1 May 2024 was selected as the master image to construct the interferometric network. Topographic phase correction was performed using the SRTM DEM with a spatial resolution of 30 m. During interferometric processing, the multilook factors in the range and azimuth directions were set to 2 × 5, respectively, to improve the signal-to-noise ratio of the interferometric phase and reduce phase noise. In the Stacking-InSAR processing workflow, interferometric pairs were first generated from the multi-temporal SAR images, followed by multilooking, topographic phase removal, phase filtering, phase unwrapping, and orbit error correction. Subsequently, the interferometric phases from multiple acquisitions were stacked and averaged to reduce the effects of random noise and residual atmospheric delays. Finally, the average deformation rate field in the radar line-of-sight (LOS) direction was obtained for the study area, which was used to characterize the spatial distribution of mining-induced ground deformation. The specific parameters are shown in

Table 2. Parameters of SAR data.

SAR Satellite	Heading/Track Angle	Date Image	Original Resolution (m)		Multi-Looking Factor		Incidence Angle (°)
			Azimuth	Range	Azimuth	Range
Sentinel-1	Ascending 128/−12°	20240501~20250501	2.8	2.33	5	2	32.23

.

4. Results

4.1. Field Investigation Results

4.1.1. Characteristics of Historical Coal Mining Activities

Investigation indicates that coal mining in the study area has a long history, but the overall scale of exploitation has been relatively small. Eight historical mining sites are identified (Figure 3), primarily targeting the C6 coal seam. All old workings were developed using horizontal adits along the coal seam strike, with manual pick-and-shovel excavation, relying on natural ventilation and drainage. The mining elevation ranged approximately from 1025 to 1265 m, and the length of each adit along the strike was about 100–150 m. Overall, the operations exhibited seasonal, small-scale production characteristics. Although the historical mining scale was limited, it still caused localized ground fissures and collapses. Moreover, the long-term existence of old goaf areas altered the original stress equilibrium of the rock mass, creating potential hazards for subsequent mechanized mining. In the context of steep mountainous terrain, disturbances in the goaf areas can easily couple with slope structural planes and faulted zones, thereby amplifying slope deformation responses. The DEM and slope map of the study area are shown in Figure 4 and Figure 5.

Slope structure and profile analysis (Figure 6 and Figure 7) indicate that the strata in the study area generally exhibit a monoclinic, gently inclined structure, with bedding orientations of approximately 280–303°∠17–21° (Figure 6, Figure 7 and Figure 8b,c), overall dipping gently toward the northwest. The 1151 working face is arranged along the dip of the coal seam, and mining-induced disturbances primarily propagate along the bedding planes. Spatially, the middle and upper parts of the working face are dominated by gently inclined layered rock masses, whereas the cliff zone exhibits a sudden increase in slope, forming extremely steep segments with pronounced elevation differences. The strata are exposed at the cliff front, with well-developed vertical joints. Profiles AA′ and BB′ (Figure 6 and Figure 7) show an inverse relationship between slope surface and bedding dip. Such extremely steep reverse-dip slope segments provide certain structural constraints under gravity; however, they are susceptible to tensile cracking at the crest and localized block instability at the shoulder and foot of the slope due to mining-induced disturbances and joint-fracture cut effects.

4.1.2. Engineering Geological Conditions of the Slope

The stratigraphy of the study area, from top to bottom, consists of Triassic, Permian, and Quaternary deposits. The upper part is dominated by the Yongningzhen Formation (T₁y) of the Triassic, composed of limestone and calcareous mudstone, with a residual thickness of 31–81 m (average ~56 m), in conformable contact with the underlying Feixianguan Formation. The middle-upper section includes the Feixianguan (T₁f) and Kayitou (T₁k) formations of the Triassic; the Feixianguan Formation constitutes the main strata of the steep slope segments, with a thickness of 349–564 m, whereas the Kayitou Formation is 22–64 m thick, mainly composed of interbedded fine sandstone, siltstone, and mudstone. The lower part is the Xuanwei Formation (P₂x) of the Permian, a coal-bearing unit with a thickness of 192–215 m, representing the primary coal-bearing horizons in the area. In gentle slope and valley zones, Quaternary Holocene (Q_h) loose deposits are widely distributed, with thicknesses of 0–20 m, mainly consisting of residual slope, alluvial, and colluvial materials. These deposits are loosely structured with low shear strength, making them susceptible to collapses and surface subsidence. The study area exhibits a monoclinic structure, with bedding orientations of 255–350°∠5–30° (Figure 8b,c), and dip angles gradually decreasing from south to north.

4.1.3. Fault Structures and Rock Mass Integrity

No regional-scale major fault structures have developed in the study area, and the overall tectonic activity is relatively low. However, multiple small- to medium-scale fractured zones are exposed underground. Four minor normal faults were identified in the production roadways, with displacements of 1.5–3 m, strikes toward NE, dips toward NW, and dip angles of 50–70°. Although these faults are limited in scale, they locally disrupt coal seam continuity and roof–floor integrity, weakening the overall rock mass structure and potentially serving as zones of stress concentration and fracture propagation induced by mining. In addition, two sets of high-angle shear joints are developed in the area (300°∠80° and 85°∠50°), producing a blocky rock mass due to joint intersection. Under extremely steep slope conditions, the combination of structural planes and slope surfaces controls the failure mode of the rock mass. Joint planes provide potential slip surfaces for block sliding and tensile deformation. When mining-induced disturbances are superimposed with gravitational effects, the controlling role of structural planes is amplified, easily triggering localized collapses and shallow landslides.

4.2. Geological Hazard Types and Spatial Distribution

The geological hazards in the study area mainly include collapses, ground fissures, and localized surface subsidence. Ground fissures are primarily distributed linearly along ridges, showing continuity, with nine major fissure zones identified near the 1154 and 1151 working faces (Figure 8a,b,e,f), reflecting clear structural control and superimposed mining-induced effects. Three relatively large collapse–landslide areas are developed in the area (Figure 8d–f), mainly occurring in fractured rock masses, jointed zones, and steep slope sections. Although the hazardous rock masses have been incorporated into monitoring programs, they still pose potential threats to roads and residential areas at the slope foot. Additionally, surface subsidence has been observed in localized zones with thick loose deposits. While these subsidence areas have not yet directly affected buildings or transportation facilities, their spatial distribution appears to correlate with historical goaf areas and variations in the thickness of Quaternary deposits.

4.3. Spatial Impact of Mining-Induced Subsurface Disturbance

A three-dimensional spatial model was constructed based on UAV-derived DEMs, mining engineering data, and geological maps to systematically analyze the mining depth, slope, and slope structural characteristics in the study area. The mining depth calculations (Figure 9) indicate pronounced spatial variability correlated with topographic undulations, with significant differences among working faces. Shallow burial zones are mainly distributed in valleys and slope foot areas, whereas deep burial zones are concentrated in the northwest region behind the cliff zone. The 1151 working face has a burial depth of 286–470 m, representing relatively shallow conditions within the study area. Profile analysis (Figure 3, Figure 6 and Figure 7) shows that local gradients in mining depth occur near the working face boundaries and structurally developed zones, providing geometric conditions for the redistribution of mining-induced stress and differential rock movement.

Slope analysis (Figure 5) indicates that the study area is dominated by medium- to high-gradient terrain, with cliff zones exhibiting slopes of 55–85°, representing typical extremely steep mountain landforms. The strata in this region form a reverse-dip slope, where the slope surface and bedding dip in opposite directions. The bedding orientation provides a certain “locking” effect on the slope; however, under mining-induced disturbances and joint intersection, tensile cracks readily develop at slope shoulders and crests. High-slope areas largely coincide spatially with existing fissures and collapse zones, representing potential instability-sensitive zones.

Comprehensive analysis suggests that under underground mining conditions in extremely steep mountain areas, mining depth governs the intensity and propagation depth of mining-induced stress, slope gradient controls the gravitational component and the direction of deformation release, and slope structure determines potential slip modes and failure paths. The spatial coupling of these three factors shapes the distribution pattern of mining-affected sensitive zones. For the 1151 working face, with a burial depth of 286–470 m, a protective coal pillar 160 m wide was retained between the working face, the cliff zone, and historical hazard areas. This measure effectively mitigates the propagation of mining-induced disturbances toward the extremely steep reverse-dip slope, limits the outward expansion of the displacement angle and stress concentration, thereby reducing the risk of landslide initiation and enhancing overall slope stability.

4.4. Comparison Between Rock Movement Theoretical Calculations and InSAR Monitoring Results

The rock movement influence range of the goaf was calculated using a boundary-angle geometric model, and a vector map of the influence boundary under mountainous conditions was constructed by applying a topography correction factor (Figure 10). Field investigations indicate that the coal mine in the study area contains two exploitable and partially exploitable coal seams (C₅ and C₆) within the Upper Permian Xuanwei Formation (P₂x), composed mainly of siltstone, silty mudstone interbedded with mudstone, and coal seams, with relatively hard overlying strata. The 1151 working face adopts longwall mining for both C₅ and C₆ seams, with an average mining height of approximately 5 m. Measured bedding orientations of the coal seams range from 5° to 15°. Based on field mining feedback and coal seam contour lines (6–15° dip), the dip angle α of the 1151 coal seam on profile BB′ is taken as 14°, and on profile FF′ as 6°, giving an average dip α of 10° for the working face. Considering the average values, the strike boundary angle δ₀ of the 1151 working face is set to 55°, the upslope boundary angle γ₀ on transverse profiles is 55°, and the downslope boundary angles are 48° (BB′) and 52° (FF′), with an average downslope boundary angle of 50°. It should be noted that the theoretical rock movement boundaries were calculated using empirical boundary-angle parameters and under the assumption of a homogeneous geological medium. Consequently, the influence length along the strike direction ranges from 818 to 1076 m, with single-side extensions of 200–329 m; the influence width along the dip direction ranges from 557 to 840 m, with upslope extensions of 200–329 m and downslope extensions of 240–394 m, resulting in a rock movement influence area of 456,000–904,000 m².

Stacking-InSAR results indicate that the main deformation zones in the study area are concentrated near the 1151 working face, with radar line-of-sight (LOS) deformation rates ranging from −0.98 to 0.55 cm a⁻¹. These results effectively reflect the spatial distribution of surface deformation induced by mining. It should be noted that the deformation values presented here are along the LOS direction and have not been converted to vertical or horizontal components. Spatial comparison between the theoretical rock movement boundaries and the InSAR-inverted boundaries shows a high overall degree of agreement. Areas of maximum cumulative deformation are mainly located at the 1151 working face and adjacent goaf zones, distributed within the theoretical boundary limits. The InSAR-derived rock movement boundaries generally fall within the theoretically predicted range, with only minor local contractions, and no anomalous subsidence patches exceeding the boundary-angle control lines are observed.

Along the B–B′ profile (Figure 11), variations in the deformation gradient are also primarily controlled by the boundary angles. Areas of concentrated subsidence correspond well spatially with the projection of the goaf and collapse zones. Although localized deformation responses are observed around existing goaf areas and mining-disturbed zones formed before 2013, no significant outward expansion beyond the model-predicted boundaries is evident.

Further analysis based on the BB′ and FF′ profiles (Figure 11 and Figure 12) indicates that the spatial positions and extents of InSAR-inverted rock movement boundaries differ somewhat from the theoretical boundary angle calculations. Overall, the InSAR-derived boundary angles are slightly larger than the theoretical values, whereas the surface influence range controlled by the theoretical boundary angles (red dashed lines) is generally slightly larger than the InSAR-identified boundaries (green dashed lines). Above the 1151 working face, the theoretical model predicts a more pronounced outward extension of the boundary toward the surface, whereas the actual deformation response identified by InSAR exhibits a slight inward contraction, indicating that the real surface response is more constrained than the theoretical prediction.

Near the steep slope and the Zhenfeng Road side, the rock movement boundaries determined by both methods extend beneath existing geological hazard zones. However, monitoring results show no signs of accelerated deformation or significant subsidence within these hazard areas. This observation is closely related to the relatively large burial depth of the 1151 working face (286–470 m) and the retention of a 160 m wide protective coal pillar. The protective coal pillar effectively reduces the propagation intensity of mining-induced stress toward the extremely steep slope, limits outward expansion of the displacement angle, and causes the mining-induced disturbance to attenuate within the slope mass. Profiles also indicate that the cliff zone is mainly composed of thick limestone and medium-thick layered sandstone, with generally competent lithology. The overlying Feixianguan and Yongningzhen formations have considerable thickness, providing a buffering and dispersing effect on mining-induced deformation. The relatively intact thick-layer lithological combination partially inhibits the propagation of mining-induced fractures to the surface, preventing the theoretical maximum influence range from fully translating into actual deformation. The theoretical boundary-angle model reflects the potential maximum influence under full mining conditions, whereas the InSAR results reveal the actual rock movement response observed in the field. The discrepancy between the two indicates that, under current extremely steep mountain conditions, mining-induced disturbances have not overcome the stability constraints imposed by the structural–lithological combination, and the existing geological hazard masses remain relatively stable.

The comparison results in

Table 3. Theoretical vs. InSAR Rock Movement Angles and Mining Influence Scope.

Method	BB′ (${\mathit{\gamma }}_0$)	BB′ (${\mathit{\beta }}_0$)	FF′ (${\mathit{\gamma }}_0$)	FF′ (${\mathit{\beta }}_0$)	Rock Movement Influence Area (m2)
Theoretical	55	48	55	52	560,000
InSAR	67	61	73	68	272,000

Note: The units of ${\gamma }_0$ and ${\beta }_0$ in the table are degrees (°).

show that the InSAR-inverted boundary angles β₀ range from 61° to 68°, approximately 13–16° larger than the theoretical downslope boundary angles, while γ₀ ranges from 67° to 73°, about 12–18° larger than the theoretical upslope boundary angles. The corresponding rock movement influence areas differ by approximately 288,000 m². This discrepancy indicates that, under the current mining conditions, the theoretical model tends to overestimate the rock movement influence range in extremely steep mountainous terrain.

5. Discussion

5.1. Control Mechanisms of Terrain and Structure on the Propagation of Mining-Induced Deformation

The study area is characterized by extremely steep slopes in an underground coal mining environment, where mining-induced disturbances exhibit a pronounced coupling with slope structural conditions. Historical small-scale goaf areas have altered the original in situ stress field during prolonged mining, placing the overlying strata in a state of continuous stress relaxation and redistribution. Under adverse topographic and structural conditions, these zones have triggered two significant historical landslides, serving as major potential hazard sources within the study area. From an engineering geological perspective, the slopes generally exhibit a gently inclined reverse-dip structure and develop two sets of high-angle joint fractures, giving the rock mass a distinctly blocky character. Local small faults and fractured zones further compromise the integrity of roof and floor strata, while the high slope angles (55–85°) reduce the shear strength of the slope mass. Under such structural conditions, mining-induced stress disturbances are more readily transmitted along bedding planes and joint fractures, forming tensile–shear cracks from the slope shoulder to the crest. Compared with flat mining areas [40,41], mining-induced surface deformation in mountainous terrain not only manifests as vertical subsidence but may also superimpose lateral deformation due to slope gravity effects, producing a compound influence on slope stability. When the mining influence range approaches or encroaches upon landslide masses, potential slip surfaces and structural weak planes within the slope may be activated or further extended, leading to gradual relaxation of the rock mass structure and evolution toward instability [42,43]. Therefore, under extremely steep mountainous conditions, the release path of mining-induced deformation is typically controlled jointly by mining depth, slope gradient, and structural plane combinations, forming a topography–structure–mining coupled control mechanism. However, compared with typical mining-disturbed slopes [44] (e.g., the Zhenxiong “1·22” landslide [45,46], Junlian “2·8” landslide [47], and Pusa landslide [48]), the study area has not experienced rapid overall slope failure. This indicates that, under the current mining depth and with the retention of protective coal pillars, mining-induced disturbances have not yet exceeded the structural stability threshold of the slope.

5.2. Discrepancies Between Theoretical Rock Movement and InSAR Monitoring Results Under Extremely Steep Mountain Conditions

Traditional rock movement theory is primarily based on the assumptions of homogeneous, continuous, and isotropic strata, and is often applied in relatively gentle mining terrains [49,50,51]. Under such conditions, mining-induced deformation can be predicted in terms of its spatial influence using parameters such as influence angles or boundary angles. However, in extremely steep mountainous environments, complex topography and rock mass structures often alter the propagation paths of mining-induced deformation, limiting the applicability of theoretical models. In this study, the InSAR-inverted rock movement boundary angles are generally larger than the theoretical calculations, whereas the corresponding deformation influence areas are smaller than the theoretical predictions. This discrepancy indicates that, in steep mountain conditions, actual mining-induced deformation is constrained by multiple factors during propagation through the overlying strata. Firstly, the rock mass in the study area exhibits significant heterogeneity and anisotropy due to joints, interbedded weak layers, and local fractured zones. Under such conditions, mining-induced disturbances tend to dissipate or attenuate along structural planes during upward propagation, limiting further deformation expansion. Secondly, the combined effects of lithology and topography also significantly influence the propagation of mining-induced deformation. The Feixianguan and Yongningzhen formations are composed mainly of thick limestone and sandstone layers with relatively high overall strength, providing a buffering and dissipating effect on mining-induced deformation. In addition, the reverse-dip slope morphology constrains horizontal displacement, preventing the overlying strata from reaching the full expansion predicted under ideal mining conditions. Therefore, the observed discrepancy indicates that the propagation of mining-induced deformation in steep mountainous environments is substantially constrained by topographic conditions and structural discontinuities, leading to a smaller actual deformation extent than that predicted by conventional rock movement theory. It should be emphasized that the theoretical boundary angle reflects the maximum potential surface influence range after mining completion and stabilization of ground movement, whereas the InSAR-derived results represent cumulative deformation recorded during a limited monitoring period. In this study, monitoring was conducted from May 2024 to May 2025, with mining of the working face occurring from August 2024 to April 2025, which may be insufficient to fully reflect long-term cumulative and delayed deformation effects. Additionally, in mountainous environments, SAR side-looking imaging is prone to geometric distortions such as layover and shadowing, potentially causing local decorrelation or phase anomalies that can interfere with the identification of deformation boundaries [52]. Consequently, the currently identified rock movement boundaries are still subject to temporal and observational accuracy limitations. Overall, the discrepancy between the theoretical and InSAR results essentially reflects the difference between the “maximum possible theoretical influence range” and the “actual controlled response range,” indicating that the practical applicability of traditional rock movement theory in extremely steep mountainous areas requires further investigation.

5.3. Slope Stability and Hazard Risk Under Present Mining Conditions

Historically, two steep landslides in the study area were triggered by early coal mining activities, closely associated with adverse slope engineering geological conditions and mining-induced disturbances. The landslide masses are primarily located in high-gradient cliff zones, where the rock mass structure is strongly controlled by bedding planes, joint fractures, and localized weak interlayers (Figure 13). Under prolonged gravitational stress and rainfall influence, these zones have developed slope structures with relatively poor stability [53]. Theoretical rock movement calculations indicate that the current mining influence boundaries approach the upper edges of the landslide masses. However, a protective coal pillar approximately 160 m wide was retained between the working face and the projected landslide area, which was not mined. This significantly attenuated the propagation of mining-induced disturbances, and rock movement deformation experienced a degree of decay before reaching the slope masses. In combination with InSAR deformation monitoring results, no significant overall deformation was observed in the two historical landslide zones. Nevertheless, minor deformation could be detected locally at the rear edge of the landslide masses, suggesting that mining-induced effects may have disturbed the shallow slope structure to some extent. Under the combined effect of the protective coal pillar and the relatively large mining depth, the overall slope remains relatively stable, although mining-induced rock mass damage is irreversible. Over time, continued gravitational tension, rainfall, and weathering may gradually degrade the slope structure, increasing the potential risk of future landslide reactivation. Therefore, during subsequent mining operations, it remains necessary to maintain continuous surface deformation monitoring and engineering geological surveys, establishing a multi-source monitoring system such as “GNSS points—ground inspection—InSAR surface”, with InSAR as the primary remote sensing component, to track slope deformation evolution in a timely manner. Concurrently, engineering protection measures, including removal of unstable rocks, flexible protective nets, and retaining structures, should be implemented for potential landslide masses in a graded management approach, ensuring slope stability and mining safety under extremely steep mountain mining conditions.

6. Conclusions

(1)

Severe mining induced hazard conditions in extremely steep mountainous terrain. The study area is located in the extremely steep mountainous region of southwestern China, where slopes are generally steep and cliff zones reach gradients of 55–85°. The gently inclined layered rock masses are significantly affected by fractures and joints, and the superposition of reverse-dip and oblique slopes creates pronounced topographic relief. These conditions make slope stability highly sensitive to underground mining disturbances, forming a potentially unfavorable engineering geological environment for mining-induced landslides.

(2)

Historical mining is the primary control on existing geological hazards. The spatial distribution of cracks, collapses, and localized subsidence corresponds closely with historical goaf areas and early mining zones, indicating that past coal mining altered the original in situ stress field and weakened slope structural integrity. These factors are major controls on geological hazard development in the area. Currently, existing hazard masses are either evolving slowly or remain largely stable, with no evidence of accelerated deformation.

(3)

Protective coal pillars effectively limit mining-induced disturbances. The 1151 working face has a burial depth of 286–470 m, and a protective coal pillar approximately 160 m wide was retained between the working face, cliff zones, and existing hazard areas. InSAR monitoring shows cumulative LOS deformation of −0.98 to 0.55 cm, concentrated primarily within the working face and adjacent goaf zones, with no significant abnormal deformation detected in the existing landslide masses. This indicates that protective coal pillars effectively constrain mining-induced disturbances from propagating into extremely steep slope sections.

(4)

Theoretical rock movement influence exceeds InSAR-observed range. Comparison of theoretical calculations with InSAR inversion results indicates that the InSAR-inverted boundary angles are generally larger than the theoretical values (downslope: 61–68° vs. theoretical 48–52°; upslope: 67–73° vs. theoretical 55°), whereas the corresponding rock movement influence area is smaller than predicted. This suggests that, under extremely steep mountain conditions, complex geological structures and relatively intact lithology constrain overlying strata movement, resulting in an actual rock movement influence smaller than that predicted by conventional rock movement theory.

In summary, this study integrates InSAR deformation monitoring with theoretical rock movement analysis to reveal the discrepancies between predicted and actual mining-induced influence ranges in extremely steep mountainous areas, and to validate the effectiveness of protective coal pillars in limiting the propagation of mining-induced disturbances toward steep slopes. The results indicate that time-series InSAR provides a valuable tool for detecting slope deformations and delineating deformation boundaries in complex terrain, offering practical guidance for combining satellite remote sensing with geomechanical analysis. These findings provide a technical reference for slope stability assessment and hazard mitigation under underground coal mining conditions in the extremely steep mountains of southwestern China, and offer insights for remote sensing-based monitoring and risk management of mining-induced hazards in similar high-gradient mountainous regions.