Progressive Shrinkage of the Alpine Periglacial Weathering Zone and Its Escalating Disaster Risks in the Gongga Mountains over the Past Four Decades

Abstract

The Alpine Periglacial Weathering Zone (APWZ) is a critical transitional belt between alpine vegetation and glaciers, and a highly sensitive region to climate change. Its dynamic variations profoundly reflect the surface environment’s response to climatic shifts. Taking Gongga Mountain as the study area, this study utilizes summer Landsat imagery from 1986 to 2024 and constructs a remote sensing method based on NDVI and NDSI indices using the Otsu thresholding algorithm on the Google Earth Engine platform to automatically extract the positions of the upper limit of vegetation and the snowline. Results show that over the past four decades, the APWZ in Gongga Mountain has exhibited a continuous upward shift, with the mean elevation rising from 4101 m to 4575 m. The upper limit of vegetation advanced at an average rate of 17.43 m/a, significantly faster than the snowline shift (3.9 m/a). The APWZ also experienced substantial areal shrinkage, with an average annual reduction of approximately 13.84 km², highlighting the differential responses of various surface cover types to warming. Spatially, the most pronounced changes occurred in high-elevation zones (4200–4700 m), moderate slopes (25–33°), and sun-facing aspects (east, southeast, and south slopes), reflecting a typical climate–topography coupled driving mechanism. In the upper APWZ, glacier retreat has intensified weathering and increased debris accumulation, while the newly formed vegetation zone in the lower APWZ remains structurally fragile and unstable. Under extreme climatic disturbances, this setting is prone to triggering chain-type hazards such as landslides and debris flows. These findings enhance our capacity to monitor alpine ecological boundary changes and identify associated disaster risks, providing scientific support for managing climate-sensitive mountainous regions.

1. Introduction

The Alpine Periglacial Weathering Zone (APWZ) refers to the transitional zone situated between glacial areas and alpine vegetation belts, typically characterized by the development of periglacial landforms such as stone stripes, sorted circles, and alpine solifluction slopes. This zone exhibits intense periglacial weathering activity and strong climatic sensitivity [1,2]. Functioning as a critical transitional boundary between the snowline and the upper limit of vegetation [3], the APWZ links the cryosphere, hydrosphere, biosphere, and lithosphere. Its structural evolution not only reflects the response of landforms and ecosystems to climate warming but also exerts a profound influence on regional disaster occurrence patterns [4,5]. Under the influence of global warming, glaciers across the Qinghai–Tibet Plateau have been undergoing continuous retreat. Comparative analyses between the First and Second Chinese Glacier Inventories show that the total glacier area within China has decreased by approximately 17.7% since the end of the last century, with glacier volume declining by about 20% [6,7]. This retreat has led to the large-scale transformation of glacial areas into APWZs, fundamentally altering surface cover configurations. Concurrently, the expansion of APWZs has intensified permafrost degradation, rock mass unloading, and the fragmentation of loose material sources, significantly increasing the frequency and magnitude of hazards such as landslides and debris flows [8,9,10,11]. These phenomena are further compounded in alpine environments by extended disaster chains and cascading amplification effects, reflecting a distinct pattern of compound risk [12,13].

In recent years, frontier research has focused extensively on glacial retreat, permafrost degradation, the dynamics of alpine vegetation lines (including treelines and grasslines), and alpine vegetation succession, offering important insights into the mechanisms underlying APWZ change. Regarding glacial and permafrost processes, studies have shown that with ongoing warming in mountainous regions, the permanent snowline (i.e., the lower boundary of glaciers) has experienced significant upward retreat [14,15], while the active layer of permafrost has thickened and the overall extent of permafrost has decreased. These changes have reshaped hydrological regimes and ecosystem patterns [16,17], providing new habitats for the expansion or migration of adjacent plant communities [18] and enabling vegetation to encroach into higher altitudinal zones [19]. Simultaneously, the thickening of the active permafrost layer has increasingly destabilized slopes and heightened the risk of mass movements [20]. With respect to the alpine vegetation line and high-altitude vegetation succession, many studies have confirmed a marked upward migration trend of the treeline and grassline under global warming conditions [21,22,23]. Additionally, the rate and pattern of vegetation succession on newly deglaciated terrain are closely related to slope aspect and hydrothermal conditions [24,25]. However, most existing research has focused solely on either the snowline [5,26] or the upper limit of vegetation [27,28,29], with limited attention given to the integrated recognition and analysis of the transitional zone between glaciers and alpine vegetation as a unified system [8,9,30].

In terms of remote sensing extraction methods, advances in remote sensing technologies—particularly the use of the Landsat satellite series in combination with the Google Earth Engine (GEE) platform—have significantly enhanced capabilities for dynamic monitoring and time-series analysis, offering new avenues for investigating surface processes in cold regions [31,32]. However, machine learning-based approaches typically require large amounts of training data for model calibration [33]. Since the upper limit of vegetation and the snowline are generally located in extremely high-altitude mountainous areas, where harsh environmental conditions and limited accessibility prevail, traditional field surveys are often impractical to conduct over large extents in such alpine zones. In contrast, threshold-based methods are known for their simplicity and efficiency [34] and have been widely applied in delineating vegetation lines and glacier boundaries [35,36]. Nevertheless, conventional fixed-threshold approaches are often inadequate for addressing the uncertainties introduced by complex terrain and the temporal variability inherent in multi-temporal remote sensing imagery [37,38].

In this context, this study selects Mount Gongga, located on the eastern margin of the Qinghai–Tibet Plateau and characterized by a typical altitudinal zonation spectrum, as the study area. We define the upper and lower boundaries of the Alpine Periglacial Weathering Zone (APWZ) using the “snowline” (glacier–bare land interface) and “upper limit of vegetation” (bare land–vegetation interface), and we propose a remote sensing extraction method that integrates NDVI and NDSI dual indices with the adaptive Otsu thresholding algorithm. This method is implemented on the GEE platform to enable long-term dynamic identification. The objective is to develop a remote sensing extraction framework for the APWZ based on the coupled recognition of its upper and lower boundaries and to systematically analyze its spatial evolution over the past four decades, as well as its responses to climatic change, topographic factors, and geological hazards. This research not only provides fundamental data support for evaluating the ecological effects of climate change in this region but also contributes to a deeper understanding of how climate change impacts mountain altitudinal zonation. Furthermore, it enriches the body of case studies concerning changes in the snowline and upper limit of vegetation in alpine and extremely high-mountain environments.

2. Materials and Methods

2.1. Study Area Overview

Mount Gongga is located on the southeastern margin of the Qinghai–Tibet Plateau, in the eastern section of the Hengduan Mountains, at 29°35′44″N and 101°52′44″E. It is the main peak of the Daxue Mountains and reaches an elevation of 7556 m, making it the highest peak in Sichuan Province. The region exhibits distinct geomorphological characteristics: the eastern slope is relatively gentle, while the western slope is steep and rugged. Overall, the terrain is highly variable, with a complex assemblage of landform types. The climate of Mount Gongga is diverse, shaped by the combined influences of high altitude, westerlies, and both the southwestern and southeastern monsoons. This results in pronounced altitudinal zonation of natural vegetation and climate. Precipitation is primarily concentrated between May and October, with summer rainfall accounting for approximately 75–90% of the annual total [39]. The region is currently experiencing a warming and moistening climatic trend. Within the altitudinal zonation system, the Alpine Periglacial Weathering Zone (APWZ) serves as a transitional zone between the glacier zone and the alpine meadow zone. It is characterized by typical periglacial landforms such as stone stripes, solifluction lobes, and blockfields. This zone is marked by intense periglacial weathering and abundant clastic material, making it a critical geomorphic unit. The upper and lower boundaries of the APWZ are primarily controlled by the extent of glaciers and the distribution limits of shrubs/meadows, respectively. Hence, glacier retreat and the upward expansion of shrubland/meadow belts directly drive changes in the spatial boundaries of the APWZ. In recent years, driven by climatic warming and increased humidity, this transitional zone has become an important source region for disaster development, such as landslides, collapses, and debris flows [9]. Frequent occurrences of hazards such as landslides, ice avalanches, and glacial lake outburst floods (GLOFs) within the region underscore the central role of the APWZ in high-mountain environmental transformation and hazard processes. These dynamics have also drawn growing attention to the mechanisms underlying its spatial and temporal evolution [12].

2.2. Data Sources and Research Methodology

Data Sources

The data used in this study are summarized in

Table 1. Overview of data sources.

Data Types	Data Source Unit	Data Name	Spatial Resolution	Time Range	Remark
Remote sensing images *	United States Geological Survey (USGS)	Landsat 5 TM/Landsat 8 OLI	30 m	1986–2024	Summer, cloud cover <10%, 5-year interval
	Copernicus Open Access Hub	Sentinel-2	10 m	2 August 2021	Cloud cover <5%
Climate data *	National Ecological Science Data Center	Monthly mean temperature and precipitation	Point Data	1998–2018	Gongga Mountain area, summer
Digital Elevation Model	NASA	ASTER GDEM V3	30 m	-	Corrected version, higher precision
Glacier data	National Glacier, Frozen and Desert Data Center	The Second National Glacier Resources Survey	-	2006–2011	Gongga Glacier
Disaster data	Data Center for Resources and Environmental Sciences of the Chinese Academy of Sciences and Field Collection	Geological Hazards Dataset	-	1985–2024	-
Water body data	EC JRC/Google	JRC Monthly Water Recurrence, v1.4	-	1984–2021	-

Note: “-“ indicates no data. “*” indicates missing image data for 2011 and 2012; missing meteorological data for 2005, 2015 and 2017.

. Remote sensing imagery: A total of 653 scenes of Landsat imagery, including Landsat 5 TM and Landsat 8 OLI, were employed, covering the period from 1986 to 2024. Images were selected for each year from June to September, with cloud coverage below 10%. To ensure temporal consistency, images were composited in five-year intervals to represent typical years. In addition, a Sentinel-2 image dated 2 August 2021 with less than 5% cloud cover was selected for manual visual interpretation of the upper and lower boundaries of the Alpine Periglacial Weathering Zone (APWZ) in Mount Gongga, which served as a reference for accuracy assessment. All imagery underwent systematic radiometric correction and precise geometric correction using ground control points (GCPs). Terrain correction and cloud removal were performed based on a digital elevation model (DEM). Climate data: Considering the sparse distribution and limited temporal continuity of meteorological stations in the high mountainous areas of western China, this study utilized observational records from two representative national meteorological stations (Figure 1c) located on the eastern slope of Mount Gongga. These records included monthly average air temperature and precipitation, from which summer climate indicators—mean monthly temperature and precipitation during June–September—were derived. Digital Elevation Model (DEM): A significantly revised version of the GDEM dataset was used. It provides high spatial and vertical accuracy, suitable for extracting topographic variables and conducting geomorphic analyses in mountainous terrain. Glacier data: Data from the Second National Glacier Inventory of China were used for indirect validation of glacier boundary extraction. Disaster data: Geological hazard data for the Mount Gongga region were compiled and used to analyze the APWZ’s disaster response characteristics. Water body data: The JRC Monthly Water Recurrence v1.4 dataset was adopted to minimize water-related errors in boundary extraction. Study area delineation: The spatial extent of the study area was primarily determined based on the hydrographic network surrounding Mount Gongga.

2.3. Methodology

This study was structured into three main steps, with the overall workflow illustrated in Figure 2. The first step (Figure 2a) involved the identification of Alpine Periglacial Weathering Zone (APWZ) boundaries and the extraction of their spatiotemporal changes. Landsat series imagery from 1986 to 2024 was accessed via the Google Earth Engine (GEE) platform. The Normalized Difference Snow Index (NDSI) and the Normalized Difference Vegetation Index (NDVI) were calculated for each image. The Otsu adaptive thresholding method was then applied to automatically extract the glacier–bare land boundary (upper limit) and the bare land–vegetation boundary (lower limit), thereby delineating the upper and lower boundaries of the APWZ. To improve extraction accuracy, results were further refined using visual interpretation of Sentinel-2 imagery and cross-validation with the Second National Glacier Inventory. Multi-temporal spatial distribution datasets of the APWZ were ultimately generated. The second step (Figure 2b) focused on analyzing the area variations of the APWZ under different topographic conditions. Elevation, slope, and aspect data were derived from the ASTER GDEM V3 digital elevation model. These topographic factors were used to statistically assess the distribution and dynamics of the APWZ across varying geomorphic settings, aiming to identify topographic control patterns. The third step (Figure 2c) investigated the driving mechanisms of APWZ variation and its linkage to geological hazards. Climate variables such as temperature and precipitation, along with topographic data and historical records of geological hazards, were collected. Trend analysis and statistical regression were employed to explore the relationships between APWZ changes and climatic factors. Furthermore, the potential impacts of APWZ shifts on hazards such as landslides and rockfalls were analyzed, highlighting its critical role in the regional disaster chain.

2.3.1. Algorithm for Extraction of the Alpine Periglacial Weathering Zone

(1)

Image Preprocessing

To accurately extract the boundaries of the Alpine Periglacial Weathering Zone (APWZ), including the snowline and the upper limit of vegetation, this study utilized a time series of Landsat TM/OLI satellite images from 1986 to 2024. The images were selected within the alpine vegetation growing season (June to September) each year. In consideration of the influence of snow cover on snowline identification and the contribution of vegetation growth to the detection of the upper limit of vegetation [40,41], only surface reflectance images with less than 10% cloud cover and atmospheric correction were used as the base data.

Given that Landsat imagery is often affected by cloud cover, the specific acquisition month of usable images within the growing season may vary across different periods. To reduce the uncertainty introduced by such temporal phase differences, a five-year compositing strategy was adopted: for each five-year period (with 2021–2024 treated as the final period), all available images from June to September with cloud cover ≤10% were collected. High-quality composite images were then generated based on the temporal stability of NDVI and NDSI metrics. This approach not only ensured comparability of vegetation and snow cover information across time, but also effectively minimized the influence of inter-month or inter-annual variability on the boundary extraction of the APWZ. In addition, the Google Earth Engine (GEE) platform’s built-in automatic cloud masking algorithm was applied, utilizing the QA_PIXEL band to mask opaque and cirrus clouds. Topographic correction was also performed following Carrasco (2019) [42]. Through this preprocessing workflow, the selected remote sensing imagery achieved a high degree of spatiotemporal consistency and reliability, providing a stable foundation for the subsequent automatic extraction of APWZ boundaries.

(2)

Index Calculation

Snowline extraction was based on the Normalized Difference Snow Index (NDSI), which effectively differentiates snow and ice from other land surface features. The calculation is shown in Equation (1). For the upper limit of vegetation (i.e., grassline) extraction, the Normalized Difference Vegetation Index (NDVI) was applied, as shown in Equation (2). Since mountainous terrain often produces significant shadow effects, resulting in abnormally low NDVI values, a simulated terrain-shadow correction model was applied to adjust NDVI values in shaded areas, thereby improving vegetation boundary identification accuracy [43].

$$NDSI={\displaystyle \frac{\mathsf{\rho }\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}-\mathsf{\rho }\mathrm{S}\mathrm{W}\mathrm{I}\mathrm{R}}{\mathsf{\rho }\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}+\mathsf{\rho }\mathrm{S}\mathrm{W}\mathrm{I}\mathrm{R}}}$$

(1)

where ρGreen denotes the green band reflectance and ρSWIR denotes the short-wave infrared band reflectance.

$$NDVI={\displaystyle \frac{(\rho NIR-\rho Red)}{(\rho NIR+\rho Red)}}$$

(2)

where ρNIR is the reflectance in the near-infrared band and ρRed is the reflectance in the red band, and NDVI reflects the degree of vegetation cover, with larger values indicating denser vegetation.

(3)

Adaptive Threshold Segmentation—Otsu Method

To enhance the accuracy and efficiency of snowline and vegetation limit extraction, the Otsu algorithm was implemented on the GEE platform to automatically determine segmentation thresholds for the feature indices. The Otsu method [44] is a widely used adaptive thresholding algorithm for image segmentation. It assumes that an image comprises two pixel classes (foreground and background) and selects an optimal threshold $t^{*}$ that maximizes the inter-class variance ${\sigma }_b^2\left(t\right)$ between the two classes. Let the gray level range be L, the probability of gray level i be pi, the foreground probability be ${\omega }_1\left(t\right)$, the background probability be ${\omega }_0\left(t\right)$, and the class means be ${\mu }_{0\left(t\right)}$ and ${\mu }_{1\left(t\right)}$, then the goal is to find $t^{*}$ that maximizes${\sigma }_b^2\left(t\right)$ (Equations (3) and (4)) [45].

$${\sigma }_b^2\left(t\right)={\omega }_0\left(t\right){\omega }_1\left(t\right){[{\mu }_{0\left(t\right)}-{\mu }_{1\left(t\right)}]}^2$$

(3)

$$t^{*}=arg {max}_t{[\sigma }_b^2\left(t\right)]$$

(4)

The Otsu algorithm has been widely applied in water body extraction [46,47], vegetation mapping [48,49], and impervious surface detection [50]. In this study, the Otsu method was used to automatically identify inflection points in NDVI and NDSI histograms, enabling delineation of the upper and lower APWZ boundaries (i.e., snowline and vegetation limit). This algorithm offers globally optimal thresholding without manual input and exhibits strong robustness, making it particularly suitable for the ambiguous transitions and indistinct boundaries in alpine environments. Leveraging the batch processing capabilities of the GEE platform, this method efficiently handles multi-temporal image sets, producing a long-term series of boundary evolution with enhanced objectivity and reproducibility. Manual adjustment of the initial Otsu threshold was further applied to optimize classification accuracy and ensure alignment with actual surface patterns.

(4)

Post-processing

To further refine the extracted boundaries, post-processing steps involving multi-source data were conducted. First, the JRC Monthly Water Recurrence v1.4 dataset on GEE was used to extract water body extents during the warmest month of each year [51], minimizing misclassification of rivers and lakes as part of the APWZ. Second, based on previous studies and ASTER GDEM V3 data, a slope threshold of 5° was applied to exclude low-slope non-glacial areas, enabling more precise control over the snowline extent [52]. The combined water and slope masks were overlaid with initial boundary results to constrain anomalous classifications.

In addition, to address the patchy noise in remote sensing imagery caused by moraines, snow cover, or shadows [7,53], GEE’s built-in morphological dilation and erosion algorithms were applied for gap filling and edge smoothing. A connectivity-based small patch removal algorithm was also employed to eliminate discrete misclassified regions [54], thereby enhancing the overall accuracy and spatial stability of boundary delineation. Finally, the post-processed raster results were converted into vector format, yielding the final delineations of the upper and lower boundaries of the APWZ—namely, the snowline and the upper limit of vegetation. This comprehensive workflow significantly improved the accuracy of automated extraction.

2.3.2. Accuracy Indicators

To evaluate the accuracy of the extracted upper and lower boundaries of the Alpine Periglacial Weathering Zone (APWZ) from remote sensing imagery, this study employed a combination of indirect and direct validation approaches. The former involved comparative analysis of total glacier area, while the latter focused on the spatial positional accuracy of the extracted upper and lower boundaries.

(1)

Indirect Accuracy Verification

Given that the Second National Glacier Inventory of China delineated glacier boundaries using Landsat TM imagery from 2006 to 2010, the glacier boundaries extracted for the year 2010 in this study were compared with the inventory data. The relative error (Er) between the two datasets was calculated to indirectly assess the accuracy of the glacier extraction, using Equation (5):

$$E_r={\displaystyle \frac{\left|A_{ext}-A_{ref}\right|}{A_{ref}}}\times 100\%$$

(5)

where A_ext is the glacier area extracted in this paper and A_ref is the glacier area in the census data.

(2)

Direct Accuracy Verification

To evaluate the spatial positional accuracy of the upper and lower boundaries of the Alpine Periglacial Weathering Zone, high-resolution Sentinel-2 imagery (10 m) was used for manual visual interpretation. The visually interpreted boundaries were taken as reference lines. To ensure consistency with Landsat-derived results, the reference lines were resampled to a 30 m resolution using the Max Resampling method. For each boundary pixel extracted from Landsat imagery, the nearest reference boundary pixel within a 2 km radius was identified [55], and their elevation values were used for quantitative accuracy assessment. The mean absolute error (MAE) and coefficient of determination (R²) were adopted as primary metrics to evaluate the elevation deviation between extracted and reference boundaries, as shown in Equations (6) and (7). Additionally, a linear regression model was fitted between the elevations of extracted and reference boundaries. The regression slope was used to further indicate the presence of systematic bias and the degree of trend consistency in the extracted results.

$$MAE={\displaystyle \frac{1}{n}}\sum _{i=1}^n\left|H_i^{ext}-H_i^{ref}\right|$$

(6)

$$R²=1-{\displaystyle \frac{\sum _{i=1}^n{{(H}_i^{ref}-H_i^{ext})}^2}{\sum _{i=1}^n{(H_i^{ref}-{\overline{H}}^{ref})}^2}}$$

(7)

where $H_i^{ext}$ and $H_i^{ref}$ are the elevation values of the extracted boundary and the reference boundary of the ith matched pixel, respectively, n is the total number of matched pixels, and ${\overline{H}}^{ref}$ denotes the average value of the reference elevation.

2.3.3. Data Extraction and Statistical Analysis

This study extracted key topographic factors—namely, elevation, slope, and aspect—based on ASTER GDEM V3 data, and overlaid them with the spatial distribution of the APWZ for each time period. This enabled an investigation of the spatial heterogeneity in the APWZ’s evolution under different geomorphic conditions.

Linear regression models were used to evaluate the temporal trends of both the APWZ area and the climatic variables (temperature and precipitation) [56]. Subsequently, the linear trends of APWZ dynamics were compared with those of temperature and precipitation to explore the underlying climatic drivers of its evolution. It is worth noting that, since the extraction of APWZ boundaries relies on summer imagery—when vegetation is at its peak growth and snow cover interference is minimal—only summer temperature and precipitation data were used in the climate correlation analysis to better isolate the direct driving factors of APWZ change.

In addition, this study compiled and analyzed regional geological hazard data, including landslides and debris flows, to examine their occurrence frequency, temporal distribution (by month), and hazard types. This was done to explore the relationship between geological hazards and the spatial evolution of the APWZ.

3. Results

3.1. Extraction Results for the Alpine Periglacial Weathering Zone Based on Otsu Algorithm

Based on the constructed extraction algorithm for the Alpine Periglacial Weathering Zone (APWZ) boundaries and leveraging the computational capabilities of the Google Earth Engine (GEE) platform, this study extracted the spatial distribution of the APWZ in the Gongga Mountain region from 1986 to 2024. A total of eight time-series remote sensing products were generated (Figure 3). Over the past four decades, the APWZ in Gongga Mountain has exhibited a gradual shrinking trend (Figure 4), with an average annual reduction rate of approximately 13.84 km²/year. The variation range (Figure 5) was primarily concentrated between elevations of 3623 m and 5249 m, with the mean elevation shifting from 4101 m to 4574 m. According to the temporal characteristics of area and elevation changes, the evolution of the APWZ can be divided into the following three stages:

(1)

1986–2000: Rapid Shrinkage Phase.

During this period, the weathering zone experienced a pronounced decline in area, with an average annual shrinkage rate of 23.727 km²/year. Concurrently, the mean elevation increased from 4101 m to 4423 m. The 15th percentile elevation rose from 3281 m to 3941 m, while the 90th percentile elevation increased from 4941 m to 5009 m, indicating an overall upward shift of the zone. The expansion rate of the lower boundary (vegetation upper limit) exceeded the retreat rate of the upper boundary (snowline), suggesting a more sensitive ecological boundary response.

(2)

2000–2005: Short-Term Expansion Phase.

This stage saw a temporary increase in area, with an average expansion rate of 35.967 km²/year, representing the most significant expansion event during the study period. The mean elevation decreased to 4263 m, while the 15th and 90th percentiles dropped to 3613 m and 4978 m, respectively, reflecting a downward shift of the zone concurrent with its expansion. In contrast to the previous stage, the retreat rate of the lower boundary was faster than the expansion rate of the upper boundary, likely influenced by short-term climate fluctuations.

(3)

2005–2024: Slow Shrinkage Phase.

The weathering zone entered another contraction phase, but with a notably reduced rate compared to the first stage, averaging 19.14 km²/year. The mean elevation increased from 4263 m to 4573 m, with the 15th percentile rising from 3613 m to 4112 m and the 90th percentile from 4978 m to 5110 m, suggesting continued vertical upward migration despite area loss. The expansion rate of the lower boundary remained higher than the retreat rate of the upper boundary.

In summary, the Alpine Periglacial Weathering Zone in the Gongga Mountain region experienced a “rapid shrinkage–short-term expansion–slow shrinkage” evolutionary sequence from 1986 to 2024, characterized by continuous area reduction and overall upward elevation migration. Over the same period, glaciers showed a general retreat at an average rate of approximately 3.9 m/year, while the vegetation upper limit advanced at about 17.43 m/year. Spatially, the most significant contraction occurred during the 1990s–2000s, with a shrinkage rate higher than that observed between 2005 and 2024, although a clear expansion event was recorded between 2000 and 2005. These changes reflect the sensitive and nonlinear response of the Alpine Periglacial Weathering Zone to climate change.

3.2. Accuracy Assessment

For the indirect validation, results from the Second Chinese Glacier Inventory were used as the reference dataset. This inventory was derived from Landsat TM and ETM+ imagery acquired between 2006 and 2011, and is widely recognized for its authority and representativeness. In the Gongga Mountain region, the glacier area reported in the inventory was 274.454 km², whereas the area extracted in this study for the year 2010 was 263.328 km². The calculated relative error was 6.415%, corresponding to an extraction accuracy of 93.585%. This indicates that the glacier identification method combining Landsat imagery with the Otsu thresholding algorithm can reliably reflect the actual distribution of glaciers in the study area. The overall extraction accuracy is within a reasonable range, demonstrating methodological feasibility.

For the direct validation, the upper and lower boundaries (i.e., the snowline and vegetation upper limit) of the Alpine Periglacial Weathering Zone extracted from 2024 Landsat imagery were compared against reference boundaries manually delineated from a high-resolution Sentinel-2 image (10 m) acquired on 2 August 2023. The upper boundary (snowline) derived using the Otsu method from Landsat data exhibited high agreement with the visually interpreted boundary from Sentinel-2 imagery. Based on the minimum-distance matching approach (Figure 6), the mean absolute error (MAE) between the two datasets was 45.37 m, with a coefficient of determination (R²) of 0.96, and a regression slope of 0.96 for the snowline elevation (Figure 6a). Similarly, the extraction of the vegetation upper limit also showed high accuracy (Figure 6b), with a mean absolute error of 48 m, R² of 0.96, and a regression slope of 0.98. These results confirm the stability and reliability of the boundary extraction method. A visual comparison between the Landsat-derived boundaries and the manually interpreted boundaries further supports the accuracy of the delineation approach (Figure 7).

3.3. Area Variations of the Alpine Periglacial Weathering Zone Under Different Topographic Conditions

To investigate the spatial variation characteristics of the Alpine Periglacial Weathering Zone (APWZ) under varying terrain conditions, we compared the distributional changes across three evolutionary stages. The analysis focused on the elevation, slope, and aspect dimensions of the APWZ boundaries, aiming to reveal the relationship between boundary shifts and topographic factors.

3.3.1. Variations Across Elevation Bands

The spatial distribution of boundary changes exhibited marked concentration and gradient features along the elevation dimension (Figure 8), primarily occurring within the 3000–4500 m range. Area changes followed a symmetrical decreasing trend around the mean, presenting an approximately normal distribution pattern. During the two shrinkage phases (1986–2000 and 2005–2024), the retreat predominantly occurred between 2500 and 4500 m, with average change elevations of 3354 m and 3818 m, respectively, indicating an overall upward migration of the APWZ. In contrast, the short-term expansion phase (2000–2005) mainly occurred between 3000 and 4000 m, yet its expansion magnitude was significantly lower than the shrinkage observed during the adjacent periods. This trend suggests that under ongoing global warming, the rate of glacier retreat is outpaced by vegetation advancement, leading to a more pronounced upward shift of the upper boundary compared to the lower boundary contraction, highlighting the heightened climatic sensitivity of ecological thresholds.

3.3.2. Variations Across Slope Gradients

Slope exerts significant control on the spatial dynamics of the APWZ (Figure 9). Over the past four decades, changes were primarily concentrated within the 20–50° slope interval, with an average change slope of approximately 33°. Overall, a distribution pattern of “maximum sensitivity on moderate slopes” was observed. During the two shrinkage phases, area loss within this slope range greatly exceeded the gain observed in the expansion phase, implying that moderate slopes—with their favorable hydrothermal conditions and energy availability—serve as both key expansion pathways for the vegetation upper limit and sensitive retreat zones for glacier margins.

3.3.3. Variations Across Aspect Classes

The influence of aspect on APWZ variations (Figure 10) is primarily driven by differences in solar radiation and illumination. Significant changes were concentrated on east-, southeast-, and south-facing slopes. East-facing slopes dominated the area changes during both shrinkage and expansion phases, with shrinkage magnitude notably greater. In general, sunny slopes experienced stronger boundary dynamics than shaded slopes, manifesting an intensifying change pattern with increasing solar exposure from shaded to sunny aspects.

In summary, the spatial evolution of the Alpine Periglacial Weathering Zone is governed by the combined effects of elevation, slope, and aspect, and is closely linked to regional climate warming. The observed changes demonstrate pronounced spatial heterogeneity and strong geographical process characteristics, providing an essential foundation for future modeling of regional environmental change and ecological adaptation.

3.4. Analysis of the Response of Alpine Periglacial Weathering Zone (APWZ) Dynamics to Climate Change and Topographic Factors

3.4.1. Response of APWZ Dynamics to Climate Change

Based on observational data from two meteorological stations in the Gongga Mountain region, the relationship between the spatial boundary changes of the Alpine Periglacial Weathering Zone (APWZ) and climatic factors was examined (Figure 11). The findings indicate that climate change—particularly variations in temperature and precipitation—is a key driver of APWZ dynamics [57,58,59]. Rising temperatures can lead to the upward shift of the snowline and the expansion of the upper limit of vegetation, directly affecting the spatial extent of the APWZ [60]. Simultaneously, glacier retreat creates new space for the APWZ to expand, while precipitation plays an indirect regulatory role by influencing vegetation growth and moisture availability [56].

Meteorological records from the Gongga Mountain region between 1998 and 2018 show a fluctuating upward trend in annual mean temperature, with a linear warming rate of 0.3 °C per decade. During this period, the overall area of the APWZ shrank, especially between 2006 and 2009 when a pronounced rise in temperature coincided with accelerated shrinkage of the APWZ. During these years, the snowline and the upper limit of vegetation both shifted upward, resulting in a rapid contraction of the APWZ, which reflects the direct impact of rising temperatures on its dynamics.

The mechanisms by which temperature influences the APWZ include two main aspects: first, warming breaks the low-temperature constraints, enabling alpine meadows to expand into higher elevations, thereby reducing the extent of bare ground and suppressing periglacial weathering processes [61,62,63]; second, rising temperatures accelerate glacier retreat, transforming deglaciated zones into newly formed APWZ areas with increasingly complex geomorphic and ecological structures. Additionally, glacier meltwater facilitates vegetation colonization and enhances soil weathering.

In contrast, the influence of precipitation on the APWZ is relatively indirect [4,64]. Although summer precipitation exhibited a decreasing trend from 1998 to 2018, the annual minimum remained above 793 mm, which did not fall below the threshold required for vegetation growth. Thus, changes in precipitation did not exert a significant direct influence on the APWZ. Its primary role lies in regulating vegetation status, thereby indirectly influencing the distribution of snowline and bare land and subsequently affecting periglacial weathering processes.

Overall, the APWZ in Gongga Mountain has shown a shrinking trend from 1986 to 2024 (Figure 3 and Figure 4), with a particularly rapid rate of contraction between 1986 and 2000, followed by a deceleration in the rate of change after 2000. This process is predominantly driven by sustained warming, which has caused the upper and lower boundaries of the APWZ to migrate upward in elevation (Figure 5), further increasing the vulnerability of regional ecosystems.

In conclusion, temperature change—especially warming—is the dominant factor driving the contraction of the APWZ, as it directly regulates its spatial configuration by controlling the positions of the snowline and the upper limit of vegetation. In comparison, precipitation plays a secondary role, primarily influencing the APWZ distribution indirectly through its effect on vegetation growth. Under future warming scenarios, the APWZ is likely to continue shrinking, potentially triggering more complex ecological succession processes and increasing geological disaster risks.

3.4.2. Response of APWZ Dynamics to Topographic Factors

As a transitional belt linking alpine vegetation and glaciers in high mountain ecosystems, the spatial distribution of the Alpine Periglacial Weathering Zone (APWZ) reflects the coupled response of ecological boundaries and cryospheric limits under climate warming [60]. Topographic factors, as the surface carriers of climatic processes, play a critical role in modulating the redistribution of radiation, moisture, and heat on the land surface, thereby significantly influencing APWZ dynamics [65].

Elevation Control: Dominant Factor in Boundary Migration. In the Gongga Mountain region, the APWZ is primarily distributed between 3000 and 5200 m, with the upper limit of vegetation typically located at 3200–4200 m and the snowline stabilized at 4900–5200 m. Driven by rising temperatures, the average elevation of the APWZ has increased from 4101 m in 1986 to 4573 m in 2024, exhibiting a clear upward trend. Elevation regulates air temperature, atmospheric pressure, humidity, and soil formation conditions, ultimately determining the distribution limits of alpine plants [66,67]. In lower elevation areas, vegetation expands rapidly as temperatures more easily surpass critical thresholds; by contrast, glaciers at higher elevations respond more slowly to warming, resulting in limited snowline migration. This leads to an asymmetric pattern characterized by “greater variability in the upper limit of vegetation versus relative stability of the snowline” [68,69].

Slope Gradient Regulation: Topographic Threshold for Expansion Potential. Slope gradient affects soil moisture retention, slope stability, and surface runoff, thereby indirectly controlling the potential for alpine meadow expansion. APWZ changes in the Gongga Mountain region are mainly concentrated in areas with slopes ranging from 20° to 50°, with the average occurring at approximately 33°. Moderate slopes (around 30°) provide favorable drainage and heat accumulation conditions, facilitating upward vegetation expansion. In contrast, steep slopes with poor soils and limited moisture hinder vegetation growth. Additionally, glaciers on steep slopes tend to melt more easily, with snowline retreat further expanding the spatial range of the APWZ [70].

Aspect Variability: Driver of Spatial Heterogeneity. Slope aspect controls the input of solar radiation and thermal energy and serves as a key spatial determinant of APWZ evolution [71]. Changes in the APWZ are most prominent on the eastern and southeastern slopes of Gongga Mountain. Sun-facing slopes receive more solar radiation and accumulate more heat, promoting rapid expansion of the upper limit of vegetation and intensified glacier melt, which leads to significant APWZ shrinkage. In contrast, the colder, shaded slopes respond more slowly to warming [72]. This pattern reflects a typical terrain–ecology coupling mechanism of “active response on sunny slopes, lagged response on shaded slopes” [73].

Synergistic Effects of Topographic Factors: Foundation of Spatial Differentiation Patterns. The combined influence of elevation, slope, and aspect determines the spatial response trajectory of the APWZ under different topographic configurations. Transitional zones at mid-to-high elevations with gentle slopes and sunny aspects show the most pronounced changes, indicating a coordinated shift between ecological and cryospheric boundaries. These topographic factors not only define local hydrothermal conditions but also regulate the coupled dynamics of vegetation expansion and glacier retreat. Their interactive effects shape the spatial heterogeneity of the APWZ [74].

In summary, the dynamic changes of the APWZ result from the synergistic effects of climate change and topographic structure. Elevation governs the migration of its upper and lower boundaries; slope gradient sets thresholds for its expansion; and slope aspect regulates the directionality of its spatial distribution. Understanding the collaborative mechanisms of these topographic factors is essential for revealing the response pathways of alpine ecosystems to climate warming and provides a scientific basis for ecological restoration and adaptive management in sensitive high-mountain environments.

3.5. Hazard Responses to the Dynamics of the Alpine Periglacial Weathering Zone in the Gongga Mountain Region

Under the context of global warming, the Alpine Periglacial Weathering Zone (APWZ), as a transitional zone between the cryosphere and alpine ecosystems, exhibits spatial dynamics that not only reflect the response of ecological boundaries to climate change but also profoundly influence the material sources, triggering mechanisms, and cascading evolution of mountainous geological disasters. In the Gongga Mountain region, the APWZ has shown a trend of continuous elevation increase and fluctuating spatial extent, which, under the joint influence of topographic factors, has significantly reshaped regional slope stability and ecosystem vulnerability, forming a critical environmental background for disaster occurrence.

To further verify the internal relationship between changes in the APWZ and geological disaster responses, this study extracted the mean elevation series of the APWZ from 1986 to 2024 (Figure 5), and compiled the frequency of geological disaster events during corresponding periods (

Table 2. Statistics of geological disasters in Gongga Mountain area from 1985 to 2024.

Year	Month	Type of Incident	Specific Location	Description
1985	July	Landslide	North slope of Gongga Mountain	Landslide caused by heavy rainfall blocked the climbing route and affected the travel of mountaineers.
1988	June	Landslide	South slope of Gongga Mountain	Spring snowmelt led to landslides, blocking part of the climbing route to Gongga Mountain.
1991	August	Mudslide	Luding County (near Luding Bridge)	Heavy rainfall triggered a mudslide, affecting traffic near Luding Bridge and causing casualties.
……	……	……	……	……
2019	August	Mudslides	Hailougou Scenic Area	Heavy rainfall triggered mudslides, which damaged much infrastructure in the scenic area and threatened the safety of tourists.
2021	June	Landslides	Eastern slope of Gongga Mountain (near Yanzigou)	Landslide near Yanzigou led to the evacuation of tourists and some roads in the scenic area needed to be repaired.
2022	September	Earthquake and secondary landslides	Luding County and surrounding areas	Secondary landslides and mudslides triggered by a 6.8-magnitude earthquake destroyed many roads in Luding and Shimian and disrupted communications.
2023	June	Landslides	Gongga Mountain Scenic Area (near Base Camp)	Heavy rainfall led to landslides, mountain climbing routes were blocked and tourists were forced to evacuate.
2024	May	Landslide	South slope of Gongga Mountain	Spring snowmelt combined with rainfall triggered landslides, affecting climbing routes and some roads in the area.
2024	July	Mudslide	Eastern part of Kangding City, near Erlang Mountain	Heavy rainfall triggered mudslides, resulting in severe damage to the road from Kangding to Luding.

Note: The table shows information for only some of the disaster sites in the study area, “……” indicates that this table only shows information on some of the disaster sites and some are omitted.

, Figure 12). A preliminary correlation analysis was conducted, and the results revealed a significant positive correlation between the average elevation of the APWZ and the frequency of geological disasters (R = 0.81, p < 0.05) [56]. This suggests that the ongoing upward shift of the APWZ substantially increases the exposure of ecologically vulnerable zones, enhances the accumulation of weathered materials, and significantly elevates the risk of geological hazards. These findings quantitatively support the hypothesis proposed in this study that “the upward shift of the Alpine Periglacial Weathering Zone drives the expansion of disaster risk,” reinforcing the indicative value of APWZ boundary dynamics as a triggering mechanism within the mountain hazard cascade system.

Upper Boundary Rise: Release Mechanism of High-Altitude Weathered Material Sources. With glacier retreat, the upper boundary of the APWZ has significantly migrated upward, exposing high-altitude bare rock areas that were previously long covered by snow and ice to intense weathering and gravitational processes. Under climatic conditions characterized by large diurnal temperature ranges and mean annual temperatures fluctuating around 0 °C in the plateau region, freeze–thaw cycles are frequent, leading to the development of rock fissures and intensified erosion [75]. This process continuously accumulates loose debris, providing essential material sources for gravity-induced hazards such as debris flows and landslides [8]. Simultaneously, glacier retreat causes rock mass unloading, further weakening slope stability—especially in steep slope areas with dense vertical joints and frequent tectonic activity—thereby exacerbating the hazard-inducing environment [9].

Lower Boundary Rise: Risk Exposure in Ecologically Fragile Zones. Rising temperatures and the upward expansion of vegetation have driven the continuous upward shift of the lower boundary of the APWZ. The newly formed vegetation boundary zone remains in an early ecological stage, characterized by shallow root systems, thin soils, and loose structure, with low resistance to disturbance [76,77,78]. When faced with extreme precipitation, sudden warming, or frequent freeze–thaw events, surface disintegration and shallow landslides are highly likely. The coupling of exposed bare rock at the upper boundary and ecological fragility at the lower boundary renders the intermediate zone a typical high-risk composite hazard-prone belt.

Hazard Chain Response: Spatial Coupling of High-Altitude Triggering and Low-Altitude Impact (Figure 13). Due to the pronounced altitudinal gradient of Gongga Mountain, hazard processes exhibit a typical chain evolution pattern characterized by “high-altitude material source—low-altitude impact” [79,80]. Once loose materials from high elevations are mobilized, they rapidly gain kinetic energy, traveling downslope along valley corridors and potentially triggering multi-hazard cascading chains such as landslides, damming events, and debris flows [5]. Historical events—such as the landslides of 1985 and 1993 and the debris flow of 2019—occurred under such geomorphological and environmental backgrounds (Table 2), demonstrating the typical driving role of weathered materials from the APWZ. The increasing frequency of geohazards in Gongga Mountain in recent decades (Figure 12) further confirms the intrinsic relationship between APWZ dynamics and regional disaster risk.

Eastern Slope Hotspot: Overlapping of Disaster Risk and Anthropogenic Disturbance. The eastern slope of Gongga Mountain features broad terrain, moderate slope gradients, high human activity intensity, and dense settlements and infrastructure, making it the most sensitive area for the interaction between APWZ changes and disaster response [56,81]. When disasters occur, they not only affect broad areas but also impact concentrated exposed elements, resulting in more severe consequences [82,83]. Therefore, the upward migration of the APWZ not only serves as an ecological response indicator of cryospheric change to climate warming but also constitutes a geomorphological triggering threshold for the expansion and intensification of mountain disaster risk.

In conclusion, the changes in the APWZ essentially represent a restructuring of the coupled ecological–hazard system driven by both climate warming and topographic structure. Glacier retreat at the upper boundary induces the accumulation of weathered material and slope instability [84], while the upward shift of vegetation at the lower boundary leads to increased ecological vulnerability [85]. The coupling of these factors forms typical initiation zones for mountain disaster chains. Strengthening dynamic monitoring and mechanistic research on this transitional zone is of great scientific significance and practical value for enhancing disaster prevention capabilities and adaptive management in sensitive alpine environments.

4. Discussion

This study demonstrates that the Alpine Periglacial Weathering Zone (APWZ) in the Gongga Mountain region exhibited a continuous shrinking trend and an overall upward migration from 1986 to 2024, with the rate of upward movement of the upper vegetation limit significantly exceeding that of the snowline retreat. This “boundary compression effect” is consistent with the findings of Pan et al. (2011) [86] and Zhou et al. (2022) [56], who reported a decrease in glacier area in Gongga Mountain since 1966, as well as with the observed upward shift of the alpine treeline and increasing vegetation cover noted by He et al. (2020) [87] and Zhu Wanze et al. (2017) [88], thereby confirming the reliability of the present study’s results.

Further analysis revealed that the APWZ in Gongga Mountain is highly sensitive to summer temperature variations, in agreement with the findings of Kosiński et al. (2019) [89] and Tuft et al. (2016) [90]. Rising temperatures simultaneously drive the upward shift of both the snowline and the vegetation limit, compressing the vertical extent of the weathering zone, which constitutes the primary mechanism of its shrinkage [60,61]. Although precipitation is not identified as the dominant factor, it still indirectly influences the interannual variability of the APWZ by regulating vegetation moisture conditions. Future studies should explore its ecological thresholds and lagged response mechanisms [56]. Topographic factors, in addition to modulating climate effects, also shape the spatial structure of weathering zone evolution. This study confirms that the APWZ responds consistently to elevation gradients, slope thresholds, and aspect patterns, with the most significant changes occurring at mid-to-high elevations (4200–4700 m), moderate slopes (25–35°), and sunny aspects. This finding aligns with the “spatial regulation mechanism” proposed by Cornacchia (2018) [91], further validating the dominant role of surface heat redistribution in ecological boundary evolution under complex terrain conditions [68,73]. In terms of disaster mechanisms, this study further clarifies the APWZ as a key zone in alpine hazard-prone systems. In the upper part of the weathering zone, glacier retreat and rock mass unloading expose bare bedrock, intensifying frost weathering and debris accumulation; meanwhile, the lower part, where nascent vegetation communities are still in early successional stages, remains structurally fragile and highly susceptible to disturbance. Together, these zones constitute a source area for geological disaster chains. The Gongga Mountain region exhibits a typical cascade mechanism of “glacier retreat—rock weathering—sediment accumulation—hazard triggering,” which corresponds closely with the mountain hazard chain models proposed by Eichel et al. (2018) [8] and Wang et al. (2021) [79], as well as with recent studies by Wei (2024) [92] and Mourey (2022) [93]. Therefore, future efforts should integrate the dynamic evolution of the APWZ into disaster risk monitoring and early warning systems, and develop simulation frameworks for hazard chains centered on weathering processes.

Methodologically, this study utilized the Otsu algorithm to construct a long-term time series recognition system for the APWZ from 1986 to 2024 on the GEE platform, enabling automated extraction of the upper and lower boundaries of the weathering zone. The Otsu method determines optimal threshold values by maximizing inter-class variance, thereby minimizing human interference and enabling robust application across multi-temporal remote sensing images and complex mountainous terrain. It offers high efficiency, consistency, and automation advantages [40,49,94,95,96]. This approach not only improves the precision and efficiency of boundary delineation but also mitigates regional data incomparability, demonstrating strong applicability and transferability [97], and provides a scalable technical pathway for large-scale dynamic monitoring of the APWZ.

Despite these findings, several limitations remain in this study: (1) Limited resolution and temporal consistency of remote sensing data. The long-term analysis was based on Landsat imagery, which, due to its spatial resolution, is insufficient for identifying fine-scale APWZ units. Although a five-year compositing strategy was applied to reduce intra-seasonal temporal phase differences, minor discrepancies in image acquisition time may still introduce uncertainties in the extracted results; (2) Susceptibility of boundary extraction to complex surface conditions. In alpine regions, the delineation of the snowline and the upper limit of vegetation may be disturbed by surface debris cover and local geological hazards, potentially leading to localized misidentifications. These disturbances limit the spatial stability and consistency of the APWZ boundary; (3) Insufficient disaster data precision and limited causal mechanism modeling. While this study statistically analyzed the frequency of geological disasters and conducted a preliminary correlation analysis, revealing a significant relationship between the rising mean elevation of the APWZ and increasing disaster occurrences (R = 0.81, p < 0.05), the limited spatiotemporal resolution of the disaster records restricts the development of a more detailed and direct quantitative causal model linking the dynamic changes in the upper and lower boundaries of the APWZ to specific hazard events. This hampers deeper modeling and prediction of hazard chain triggering mechanisms. Future research should aim to do the following: (1) Incorporate high-resolution remote sensing data (e.g., Gaofen, Sentinel) and advanced boundary identification algorithms (e.g., deep convolutional neural networks) to enhance the accuracy and consistency of APWZ boundary extraction; (2) Integrate UAV photogrammetry and field transect monitoring to establish a multi-source validation system; (3) Strengthen field investigations and remote sensing-based hazard identification to build a refined and spatially explicit geological disaster event database for improved disaster modeling and response analysis; (4) Develop a coupled simulation framework that integrates APWZ dynamics, ecological boundary shifts, and hazard responses, enabling comprehensive analysis and dynamic prediction of cryospheric surface system evolution. Such a framework would provide critical scientific support for alpine ecosystem restoration and disaster risk management in climate-sensitive mountain environments.

5. Conclusions

Based on summer Landsat remote sensing imagery from 1986 to 2024, this study integrated NDVI/NDSI indices with the Otsu automatic segmentation algorithm on the Google Earth Engine (GEE) platform to extract the temporal boundaries of the Alpine Periglacial Weathering Zone (APWZ) in the Gongga Mountain region across eight time periods. The extraction results were validated using high-resolution satellite imagery and in situ data. A comprehensive analysis of the APWZ’s spatiotemporal evolution, dominant driving factors, and disaster response mechanisms led to the following main conclusions:

(1)

Methodological framework and validation: The proposed APWZ boundary extraction method—based on the fusion of NDVI/NDSI indices and Otsu algorithm—achieved high automation and timeliness on the GEE platform, demonstrating strong applicability and transferability. Comparison with the Second National Glacier Inventory showed an extraction error of 6.42% for glacier extent, corresponding to an overall accuracy of 93.58%. The snowline and vegetation upper limit boundaries were also validated against Sentinel-2 visual interpretations, yielding mean absolute errors of 45.37 m and 48 m, respectively, and R² values of 0.96 for both. These results confirm the robustness and reliability of the method in complex alpine terrain.

(2)

Change characteristics: From 1986 to 2024, the APWZ in Gongga Mountain exhibited a persistent shrinking trend and overall upward migration. The mean elevation of the APWZ boundaries rose from 4101 m to 4575 m, with an average annual area loss of 13.84 km². The change process can be divided into three phases: rapid shrinkage (1986–2000), brief expansion (2000–2005), and gradual shrinkage (2005–2024). The average annual uplift rates of the snowline and vegetation upper limit were 3.9 m and 17.43 m, respectively. The latter had a greater magnitude and played a dominant role in the reduction of APWZ area, highlighting the sensitivity of alpine vegetation to climate warming.

(3)

Dominant driving mechanisms: The evolution of the APWZ is co-driven by climatic and topographic factors. Temperature increase is the primary driver, promoting simultaneous upward movement of the snowline and vegetation limit, thereby compressing the vertical space of the weathering zone. Precipitation serves a regulatory role, indirectly influencing boundary fluctuations by affecting vegetation moisture conditions. Topographic factors shape the spatial heterogeneity of the response: elevation controls the migration trajectories of upper and lower boundaries; slope determines expansion potential; aspect regulates solar radiation and thermal accumulation. The most significant APWZ changes occurred at mid-high elevations (4200–4700 m), moderate slopes (25–35°), and sunny aspects, reflecting its spatial sensitivity to climate warming.

(4)

Disaster response mechanisms: The upward shift of the APWZ has substantially restructured the regional disaster-prone environment. Glacier retreat at the upper boundary exposes high-elevation bedrock, intensifying debris accumulation and slope instability. The lower boundary, characterized by nascent and fragile vegetation communities with shallow root systems, is vulnerable to disturbances. These conditions collectively form a typical chain-type hazard-prone pattern of “high-elevation sediment source–mid-slope deformation–low-elevation impact.” In recent years, the frequency of landslides and debris flows in the Gongga Mountain region has increased significantly, confirming the amplifying effect of APWZ changes on mountain disaster risks. The spatial dynamics of the APWZ have become a critical indicator for assessing mountain hazard chain initiation zones and triggering thresholds.

In summary, the transformation of the Alpine Periglacial Weathering Zone in Gongga Mountain not only reflects the integrated response of cryosphere–ecosystem boundaries to climate change but also constitutes a fundamental mechanism for the escalating risks of mountain geological disaster systems. The findings of this study provide essential scientific insights into the coupled evolution of alpine surface systems under climate warming and offer valuable support for improving ecological security assessments and disaster prevention strategies in cold regions.