Study on Soil Freeze–Thaw and Surface Deformation Patterns in the Qilian Mountains Alpine Permafrost Region Using SBAS-InSAR Technique

Xue, Zelong; Zhao, Shangmin; Zhang, Bin

doi:10.3390/rs16234595

Open AccessArticle

Study on Soil Freeze–Thaw and Surface Deformation Patterns in the Qilian Mountains Alpine Permafrost Region Using SBAS-InSAR Technique

by

Zelong Xue

,

Shangmin Zhao

^*

and

Bin Zhang

College of Geological and Surveying Engineering, Taiyuan University of Technology, Taiyuan 030024, China

^*

Author to whom correspondence should be addressed.

Remote Sens. 2024, 16(23), 4595; https://doi.org/10.3390/rs16234595

Submission received: 9 October 2024 / Revised: 27 November 2024 / Accepted: 4 December 2024 / Published: 6 December 2024

(This article belongs to the Special Issue Monitoring and Modelling of Geological Disasters Based on InSAR Observations: 3rd Edition)

Download

Browse Figures

Versions Notes

Abstract

:

The Qilian Mountains, located on the northeastern edge of the Qinghai–Tibet Plateau, are characterized by unique high-altitude and cold-climate terrain, where permafrost and seasonally frozen ground are extensively distributed. In recent years, with global warming and increasing precipitation on the Qinghai–Tibet Plateau, permafrost degradation has become severe, further exacerbating the fragility of the ecological environment. Therefore, timely research on surface deformation and the freeze–thaw patterns of alpine permafrost in the Qilian Mountains is imperative. This study employs Sentinel-1A SAR data and the SBAS-InSAR technique to monitor surface deformation in the alpine permafrost regions of the Qilian Mountains from 2017 to 2023. A method for spatiotemporal interpolation of ascending and descending orbit results is proposed to calculate two-dimensional surface deformation fields further. Moreover, by constructing a dynamic periodic deformation model, the study more accurately summarizes the regular changes in permafrost freeze–thaw and the trends in seasonal deformation amplitudes. The results indicate that the surface deformation time series in both vertical and east–west directions obtained using this method show significant improvements in accuracy over the initial data, allowing for a more precise reflection of the dynamic processes of surface deformation in the study area. Subsidence is predominant in permafrost areas, while uplift mainly occurs in seasonally frozen ground areas near lakes and streams. The average vertical deformation rate is 1.56 mm/a, with seasonal amplitudes reaching 35 mm. Topographical (elevation; slope gradient; aspect) and climatic factors (temperature; soil moisture; precipitation) play key roles in deformation patterns. The deformation of permafrost follows five distinct phases: summer thawing; warm-season stability; frost heave; winter cooling; and spring thawing. This study enhances our understanding of permafrost deformation characteristics in high-latitude and high-altitude regions, providing a reference for preventing geological disasters in the Qinghai–Tibet Plateau area and offering theoretical guidance for regional ecological environmental protection and infrastructure safety.

Keywords:

permafrost; surface deformation; SBAS-InSAR; spatiotemporal interpolation; Qinghai–Tibet Plateau; Qilian Mountains; freeze–thaw cycles

1. Introduction

Permafrost, defined as soil or rock that remains at or below 0 °C for at least two consecutive years, is highly sensitive to temperature changes and vulnerable to degradation under global warming [1,2,3,4]. Its degradation accelerates soil erosion and desertification, posing risks to global ecosystems and making permafrost an important “climate change indicator”.

The Qinghai–Tibet Plateau (QTP), home to extensive permafrost and cryogenic landforms, is the world’s largest permafrost area in a mid- or low-latitude region, covering around 1.06 × 10⁶ km² [5,6,7]. Over the past 50 years, the warming rate of the QTP has been approximately twice the global average, leading to predictions that around 80 km³ of ground ice could melt per decade [8,9,10,11,12].

The eastern QTP experiences substantial snow and ice fluctuations, making it highly sensitive to climate change [13,14,15]. The Qilian Mountains, located on the northeastern edge of the QTP (38.0°N, 98.8°E) [16], stretch approximately 850 km in length and 200–300 km in width [17]. With annual precipitation ranging from 400 to 700 mm [18], this region is among those most profoundly affected by climate change and human activities on the QTP [19]. However, permafrost monitoring in this region remains limited compared to other areas of the QTP, necessitating a detailed analysis of surface deformation patterns and the seasonal and interannual characteristics of permafrost thawing.

In the Qilian Mountains, permafrost degradation is driven by temperature increases that accelerate thawing, seasonal freeze–thaw cycles causing frost heave and subsidence, and precipitation variations that impact soil moisture and enhance thaw subsidence. The region’s diverse topography further influences permafrost stability. Understanding these factors and their interactions is essential for accurately characterizing surface deformation in this sensitive high-altitude area.

Recent advancements in remote sensing have made Interferometric Synthetic Aperture Radar (InSAR) a powerful tool for high-resolution surface deformation monitoring, though challenges in separating topographic phase contributions and eliminating atmospheric effects [20,21,22,23,24,25]. Differential Interferometric Synthetic Aperture Radar (D-InSAR) provides more precise high-temporal-resolution surface deformation data [21,26], but remains limited by decorrelation and noise, which constrain its accuracy [27,28].

To enhance InSAR’s effectiveness, multi-temporal methods like Multi-Temporal Interferometric Synthetic Aperture Radar (MT-InSAR), including Permanent Scatterers (PS-InSAR) and Small Baseline Subset (SBAS-InSAR), have been developed. These methods enable high-precision time series measurements of deformation, with SBAS-InSAR, providing greater flexibility in monitoring complex terrains and mitigating atmospheric effects [29,30,31,32,33,34,35,36,37,38,39].

However, these methods primarily measure deformation along the radar line of sight (LOS), limiting representation of actual surface deformation. To overcome this, researchers have combined ascending and descending orbit data to create two-dimensional fields focusing on vertical and east–west components, though this is hindered by inconsistent observation times [40,41,42,43,44]. Some researchers have explored short-term three-dimensional surface deformation by merging ascending and descending orbit SAR data with Offset-Tracking observations, InSAR results, and Multiple-aperture SAR Interferometry (MAI) results, but these suffer from poor accuracy and are only suitable for observing large-scale surface deformations [45,46]. Other scholars have tried to obtain long-term three-dimensional deformation rates by integrating InSAR results with GPS horizontal measurements; however, due to limited GPS coverage, this method lacks broad applicability [47,48]. Considering the strengths and weaknesses of these methods, this study opts for a fusion of ascending and descending orbit data to calculate two-dimensional surface deformations and proposes a method based on spatiotemporal interpolation to obtain more accurate two-dimensional surface deformation information.

Furthermore, surface deformation in permafrost regions can be modeled using time-series InSAR approaches like linear models, cyclic models, or those incorporating temperatures and climate factors. Linear models capture long-term patterns but may miss finer deformations, leading to errors [49,50]. Liu et al. used a simplified Stefan equation to model seasonal deformations based on freeze–thaw cycles, though limited to SAR images from June to December [51,52]. Wang et al. applied degree–day models for freezing–thawing cycles, but they often overestimate seasonal amplitudes and necessitate high-resolution temperature data [6,53]. Daout et al. used cyclic models to simulate seasonal deformation, which showed seasonal changes but missed long-term subsidence trends from global warming [54,55]. Therefore, this study seeks to improve upon the cyclic model and further optimize it using the least squares method, aiming to more accurately represent the patterns of surface deformation in permafrost regions.

In this study, we primarily discuss surface deformation patterns and influencing factors in the Qilian Mountains’ alpine permafrost region from March 2017 to March 2023. We established an SBAS-InSAR processing workflow, applied a spatiotemporal interpolation method to integrate ascending and descending orbit data, and employed a dynamic periodic model to analyze freeze–thaw processes. This study provides essential data for regional policy-making and contributes to understanding permafrost deformation patterns and natural disaster prevention amid climate change.

2. Study Area and Datasets

2.1. Study Area

The study area is located in the central and eastern parts of the Qilian Mountains on the QTP, as depicted in Figure 1. It extends from the Tuolai Mountains in the north to Qinghai Lake in the south, spanning geographical coordinates from 37.20°N to 38.30°N latitude and from 98.86°E to 100.27°E longitude. The area covers approximately 1.52 × 10⁴ km², with an average elevation of about 3853 m. The predominant surface coverage is grassland [56], which plays a crucial role in moisture conservation [57,58], and the vegetation exhibits distinctive vertical zonation characteristics [59]. Moreover, the region boasts significant permafrost resources [60] and is characterized by a typical high-altitude (sub-alpine) semi-arid climate [61,62].

The northern part of the mountains is rugged with deep valleys and towering ranges, located deep within the Eurasian continent, which results in a dry climate, sparse vegetation, and a fragile ecological environment. In contrast, the southern mountains are comparatively gentle with less pronounced relief and lower elevations. Due to the proximity to Qinghai Lake and the influence of monsoons, the climate here is more humid, the vegetation richer, and the ecological environment relatively stable, though still vulnerable to climate change. The annual average temperature of the Qilian Mountains is about −4.8 °C, showing a decreasing trend with increasing latitude, longitude, and elevation. The average annual precipitation is about 449 mm, mainly concentrated from May to September, and decreases from east to west [63]. These geographical characteristics significantly influence the development patterns of permafrost within the region. The main body of permafrost in the area is situated in the north, transitioning to seasonally frozen ground as latitude decreases. Continuous permafrost covers approximately 9.66 × 10³ km², representing about 63.5% of the study area; discontinuous permafrost (transition regions) covers about 1.56 × 10³ km², or 10.3% of the area; and seasonally frozen ground spans about 3.72 × 10³ km², accounting for 24.5% of the area.

Previous studies have indicated the sensitivity of permafrost in the Qilian Mountains to climate change. Cheng and Wu identified ongoing permafrost degradation in this region, driven by rising temperatures and increasing active layer thickness [11]. Zhao et al. examined the impacts of freeze–thaw cycles, highlighting their influence on altering hydrological and thermal conditions, particularly in transitional zones between continuous and seasonally frozen ground [64]. In addition, Geo et al. highlighted the role of topography, with valleys experiencing higher subsidence rates due to increased soil moisture retention compared to steeper slopes [12,65].

These findings underscore the need for refined, high-resolution monitoring techniques to better capture the complex dynamics of permafrost deformation in the Qilian Mountains. Building on this foundation, this study utilizes SBAS-InSAR methods to investigate the surface deformation characteristics and freeze–thaw processes of permafrost in the region, providing a more detailed spatial and temporal understanding of these phenomena.

2.2. Data

2.2.1. SAR Data

This study utilized Sentinel-1 SAR Single Look Complex (SLC) C-band remote sensing data downloaded from the Alaska Satellite Facility (ASF) for the monitoring of surface deformation in the study area. The basic information regarding the Sentinel-1 SAR data used is presented in Table 1, which includes average values for the azimuth angle and incidence angle specific to the study region.

The selected data from the ascending orbit covers the period from March 2017 to March 2023, while the descending orbit data spans from February 2017 to April 2023. The ascending orbit data includes 163 images, resulting in 724 interferometric pairs after manual selection and quality filtering to ensure proper connectivity of the pairs. Similarly, the descending orbit data includes 182 images, with 853 interferometric pairs retained following a similar manual selection and filtering process. Additionally, Precise Orbit Ephemerides (POD) published by ESA were used to correct orbit information.

2.2.2. GLO-30 DEM

The DEM data selected for this study is the Copernicus DEM GLO-30 (CopDEM GLO-30), with a resolution of 30 m. This product is considered the best DEM of its class available today [66], derived from global radar satellite data acquired during the TanDEM-X mission (2010–2015) and generated through interferometric processing, among other steps. Given its recent collection period, it is highly suitable for use in the SBAS-InSAR workflow employed in this study for slope calculation, elimination of topographic phase, and geocoding, and plays a crucial role in the analysis and determination of permafrost freeze–thaw cycles over several years.

2.2.3. Underlying Surface Data and Permafrost-Related Data

To better understand the seasonal deformation patterns of high-altitude permafrost in the Qilian Mountains and their influencing factors, this study selected 30 types of underlying surface data, including elevation and permafrost-related data, to explore their correlation with freeze–thaw cycles, detailed in Table 2.

2.2.4. GNSS Data

This study employed raw surface deformation data from the GNSS DATA PRODUCTS OF CHINA EARTHQUAKE ADMINISTRATION (CEA GNSS Data), which were processed to verify the accuracy of the spatiotemporal alignment of the two-dimensional surface deformation fields produced in this study, thus ensuring the theoretical feasibility of the research [67,68]. The study area includes two reference points: the Qinghai Gangcha site [QHGC] (37.33°N, 100.17°E) and the Qinghai Qilian site [QHQL] (38.19°N, 100.24°E). The QHQL site, located in a forested area, showed poor coherence in InSAR results. Hence, the vertical and east–west GNSS data from QHGC was used to verify the outcomes of this study.

3. Method

Figure 2 illustrates the workflow for monitoring surface deformation in permafrost regions, comprising three specific steps. In the SBAS-InSAR processing workflow, the first step involves selecting Sentinel-1 SAR images that comprehensively cover the study area and generating differential interferograms. Next, the interferograms are processed for time series analysis to produce LOS surface deformation results. The second step establishes a two-dimensional field of surface deformation by interpolating the LOS deformation results from ascending and descending SAR images to align them temporally and then decomposing them into vertical and east–west components of surface deformation. Finally, in the third step of the surface deformation characteristic analysis phase, the vertical deformation time series is decomposed into annual deformation rates and seasonal deformation amounts, with comprehensive interpretation and analysis conducted across the entire region and at specific typical points.

3.1. SBAS-InSAR Processing Workflow

Initially, the required Sentinel-1 SAR data are retrieved through the ASF data platform. Subsequently, by setting the temporal and perpendicular baseline parameters, the necessary interferogram pairs are selected for further interferometric processing. Optimal settings for the temporal and perpendicular baselines ensure good coherence of the results [69,70]. In this study, the threshold for the temporal baseline is set at 60 days, and for the perpendicular baseline, at 200 m.

Additionally, we employed the Goldstein–Werner Adaptive Phase Filter (ADF) to enhance the visibility of fringes in the interferograms, further reducing phase noise. This approach allowed for increased smoothing in areas of high coherence within the study region, while applying less smoothing in areas of lower coherence to avoid the removal of signals that may represent actual surface deformation [71]. In this study, we set the adaptive filter parameter, which controls the degree of smoothing applied, to 0.6. This parameter, which typically ranges from 0 to 1, controls the degree of smoothing based on the coherence of the data. A value closer to 1 applies stronger smoothing, effectively reducing noise but with a risk of oversmoothing signals in low-coherence areas. Conversely, a lower value preserves more fine details but may result in less noise reduction. Setting the parameter to 0.6 provided a balance between effective noise reduction and the preservation of deformation signals, particularly in regions with moderate to high coherence. Subsequently, we selected an appropriate number of looks to determine the resulting pixel spacing and resolution. For this study, we opted for a 10 × 2 configuration of looks, resulting in a pixel spacing of 40 m. The study area includes parts of rivers and lakes. Since water bodies have almost zero coherence, we applied masking to them during phase unwrapping to exclude their influence.

Following the completion of the above processes, geocoding was performed to transform the data from radar coordinates to geographic coordinates, with the output formatted as GeoTIFF files for time-series processing. This step ensured a regular distribution of the final vector results, ensuring uniformity between the geospatial results of ascending and descending data.

It is noteworthy that during the processing of interferometric pairs, some SAR images had acquisition intervals exceeding 60 days, leading to discontinuities in the interferometric pair connection baselines, requiring manual connection of interferometric pairs, while also excluding interferometric pairs with poor coherence or significant phase noise.

After interferometric processing, time series analysis is performed using the interferometric results. The coherence threshold for the time series is set to 0.7. Atmospheric errors generated during the SBAS-InSAR process are corrected using the Generic Atmospheric Correction Online Service for InSAR (GACOS) [72]. Ultimately, this process yields the line of sight (LOS) measurements of surface deformation for both ascending and descending orbits within the study area.

3.2. Establishment of a Two-Dimensional Surface Deformation Field

3.2.1. Time Series Registration of Ascending and Descending Results Based on Cubic Spline Interpolation

Monitoring surface deformation using InSAR technology can be challenging due to limitations like revisit intervals, making it difficult to obtain continuous time series for cumulative deformation. Additionally, when calculating the two-dimensional deformation field, it is necessary to spatially and temporally align the results from ascending and descending orbits. This alignment is challenging because the data acquisition times for ascending and descending data do not coincide. Therefore, interpolation techniques are employed to process the time series of cumulative deformations from ascending and descending results, estimating deformation values for unified or missing moments [73].

Cubic spline interpolation is a complex interpolation method that uses multiple piecewise cubic polynomials to interpolate between each known data point. These polynomials must be continuous at the data points and have the same first and second derivatives to ensure the smoothness of the overall curve. The general form of the cubic spline interpolation function is:

S_{i} (x) = a_{i} + b_{i} (x - x_{i}) + c_{i} {(x - x_{i})}^{2} + d_{i} {(x - x_{i})}^{3}

(1)

Here,

S_{i} (x)

represents the cubic spline polynomial for the

i

-th interval, and the coefficients

a_{i}

,

b_{i}

,

c_{i}

,

d_{i}

are computed by solving a system of equations that ensure continuity and smoothness across adjacent intervals, including matching the first and second derivatives at the data points.

This method is particularly suitable for handling multiple data points, generating a smooth curve while avoiding numerical oscillation issues [74]. This study employs cubic spline interpolation for temporal interpolation of descending InSAR results.

For ascending and descending SAR image data acquired at times

T_{1}

,

T_{2}

…

T_{n}

and

T_{1}^{'}

,

T_{2}^{'}

…

T_{n}^{'}

, respectively, the cumulative deformation values obtained through SBAS-InSAR technology are

D_{T}

for ascending and

D_{T^{'}}^{'}

for descending orbits. The ascending and descending SAR data capture surface deformations

D_{T_{1} T_{2}}

and

D_{T_{1}^{'} T_{2}^{'}}^{'}

during the periods

T_{1} T_{2}

and

T_{1}^{'} T_{2}^{'}

, respectively. However, due to the non-coincidence of acquisition times—

T_{1} \neq T_{1}^{'}

,

T_{2} \neq T_{2}^{'}

—it is not possible to determine the surface deformation

D_{T_{1} T_{2}}^{'}

captured by descending SAR images during the

T_{1} T_{2}

period. Therefore, using

T_{1}

and

T_{2}

as reference times, the cubic spline interpolation method is applied to the descending results to obtain the corresponding deformation values

D_{T_{1}}^{'}

and

D_{T_{2}}^{'}

at these moments.

3.2.2. Calculation of Two-Dimensional Surface Deformation Fields

After spatial resampling (geocoding) and temporal interpolation (using cubic spline interpolation) of the observations from ascending and descending orbits,

D_{T_{1} T_{2}}

and

D_{T_{1} T_{2}}^{'}

represent the LOS deformation values for the same location during the same time interval from ascending and descending orbits, respectively. The relationship between the LOS direction and the surface location is illustrated in Figure 3. The OA direction indicates the projection of the LOS direction on the ground, and the OB direction represents the projection of the satellite’s flight direction on the ground. θ is the satellite’s incidence angle. α is the azimuth angle of the satellite’s flight direction, i.e., the angle between the north direction and the satellite’s flight direction (measured clockwise). β is the angle between the OA direction and north (measured clockwise).

From the geometric relationships, we derive:

β = α - \frac{3}{2} π

(2)

From Figure 3b,c, it can be observed that the LOS deformation in the ascending SAR image is:

\{\begin{matrix} D_{T 1 T 2} = D_{c o s 1} + D_{c o s 2} \\ D_{c o s 1} = d_{h} \sin θ_{A} \\ D_{c o s 2} = d_{u p} \cos θ_{A} \end{matrix}

(3)

where

d_{h}

represents the horizontal deformation vector of the ground point, i.e., the projection of the LOS deformation onto the horizontal plane,

d_{u p}

represents the vertical deformation vector of the ground point,

θ_{A}

is the radar incidence angle in the ascending orbit,

D_{c o s 1}

denotes the projection of the horizontal deformation vector onto the satellite’s line of sight, and

D_{c o s 2}

represents the projection of the vertical deformation vector onto the satellite’s line of sight.

It is important to note that the relationship between horizontal deformation, vertical deformation, and LOS deformation is not a simple trigonometric function, nor does it exhibit a fixed magnitude relationship. In processing these measurements, a three-dimensional perspective must be adopted, considering not only the magnitudes but also the directions of both the horizontal and vertical deformation vectors. This approach ensures that the contributions of both components are properly accounted for, leading to accurate results [45].

From the geometric relationship shown in Figure 3c, we can derive:

d_{h} = d_{E s t} \times (- \sin β_{A}) + d_{N t h} \times (- \cos β_{A})

(4)

where

d_{E s t}

represents the east–west deformation vector of the surface,

d_{N t h}

represents the north–south deformation vector, and

β_{A}

is the angle between the north direction and the ground range direction (measured clockwise) in the ascending orbit.

Due to the Sentinel-1 satellite’s operation in a near-polar, sun-synchronous orbit, it exhibits limited sensitivity to deformations in the north–south direction when monitoring surface deformations. Therefore, when processing LOS deformations, north–south components are excluded, and results from both ascending and descending orbits are integrated to compute the vertical and east–west components of the two-dimensional deformation field [75,76]. This is achieved by constructing the following system of equations:

\{\begin{matrix} D_{T_{1} T_{2}} = d_{u p} \cos θ_{A} - d_{E s t} \sin θ_{A} \sin β_{A} \\ D_{T_{1} T_{2}}^{'} = d_{u p} \cos θ_{D} - d_{E s t} \sin θ_{D} \sin β_{D} \end{matrix}

(5)

It is important to note that

D_{T_{1} T_{2}}^{'}

in the equation set represents the LOS deformation value from the descending orbit results after cubic spline interpolation, aligned in space and time with the ascending orbit observations. Additionally,

θ_{D}

is the radar incidence angle in the descending orbit, and

β_{D}

is the angle between the north direction and the ground range direction (measured clockwise) in the descending orbit.

3.3. GNSS Processing

The original GNSS data for the QHGC site in both vertical and east–west directions showed linear trend errors and outliers. To address these issues, a two-step process of detrending and wavelet-based denoising was applied using MATLAB R2022b, effectively isolating noise from the true surface deformation signals. The flat and dry surroundings of the QHGC site, free from landslides and human interference, ensure that the original deformation information remains intact for validation of the InSAR results.

Figure 4 displays the original GNSS data for the QHGC site in both the vertical and east–west directions. Due to the presence of linear trend errors and outliers in the data, this study employs a two-step process of detrending and denoising to restore the true surface deformation values at the QHGC site. These restored values are then used to validate the accuracy of the InSAR results. Additionally, the area surrounding this site is flat and dry, devoid of landslides and human interference, ensuring that the original deformation information is preserved.

3.3.1. Detrending GNSS Data

Using MATLAB R2022b, the original GNSS data were processed to remove linear trends in both the vertical and east–west directions (Figure 5) using a high-pass filter with a cutoff frequency of 1/365 Hz. This effectively eliminated the low-frequency components associated with annual trends. However, it is evident that numerous outliers remain, which could potentially affect the precision of subsequent validation efforts.

3.3.2. Denoising GNSS Data

Studies have shown that GNSS observations and noise possess different frequency characteristics in wavelet multi-scale spaces, with noise typically exhibiting high-frequency features and GNSS observations displaying low-frequency characteristics. Hence, this study designs an appropriate wavelet filter using the db6 wavelet basis to effectively isolate and remove noise from GNSS data (Figure 6). This treatment substantially restores the true measures of surface deformation.

3.4. Development of a Permafrost Dynamic Periodic Deformation Model

After acquiring the time series of surface vertical deformation for the study area, the study further differentiates between the seasonal and interannual deformations of permafrost. Additionally, it enhances the accuracy of representing the annual variability of seasonal deformation amplitudes. This study improves upon traditional cyclical deformation models by employing a linear model to express the interannual surface deformation trend and fitting seasonal deformations with a dynamic sine model, mathematically abstracted as follows:

B = A t + α

(6)

d_{u p}^{i} = v \times t + B \cdot \sin (\frac{2 π}{T} \times t + φ) + ε

(7)

Here,

d_{u p}^{i}

represents the deformation on the

i

-th day of the year relative to the first SAR image; B denotes the dynamic cyclical deformation amplitude; T is the period, here being 365 days;

φ

is the initial phase;

ε

represents the residual phase.

Consequently, the seasonal surface deformation S can be expressed by Equation (8):

S = 2 \times |B^{'}|

(8)

In the above formula, B is a dynamic amplitude value. Therefore, when determining the specific values of seasonal deformation, the mean value of B, denoted as

B^{'}

, is used to replace the amplitude value.

3.5. Application of Q-Statistics to Analyze Interannual Deformation Rate and Seasonal Deformation Influencing Factors

In this study, we employed the optimal parameters-based geographical detector (Geodetector) model, which incorporates the Q-Statistic to assess spatial heterogeneity and identify factors influencing surface deformation in the Qilian Mountains’ permafrost region. The Q-Statistic, proposed by Song and colleagues [77], quantifies the relationship between explanatory variables and the spatial distribution of a phenomenon. The Q-Statistic is defined as:

Q = 1 - \frac{\sum_{f = 1}^{M} N_{f} σ_{f}^{2}}{N σ^{2}}

(9)

Here,

N

and

σ^{2}

represent the total number of observations and the population variance of the response variable across the entire study area, and

N_{f}

and

σ_{f}^{2}

represent the number of observations and the population variance of the response variable within the

f

th sub-region (

f = 1, \dots, M

) of the explanatory variable.

A higher Q value indicates a stronger association between the explanatory variable and the spatial distribution of the phenomenon.

4. Results

4.1. SBAS-InSAR Results

Figure 7 displays the perpendicular and temporal baselines for the ascending and descending SAR data used in this study, respectively. The added interferometric pairs are shown as blue lines in Figure 7, while the excluded interferometric pairs are displayed as black lines. Since the study area is located in alpine permafrost regions with a significant proportion of grasslands, bare rocks in flat terrains are relatively rare, and urban areas frequently undergo construction. Therefore, the reference points for InSAR processing were chosen in the southern part of the study area where the seasonally frozen ground is flat, dry, and has a coherence close to one.

Through SBAS-InSAR processing, a coherence map of the study area was generated (Figure 8). The transition from black to white represents gradually increasing coherence. It is evident from the map that the overall coherence in the perennially frozen soil areas is lower than in the seasonally frozen ground areas. This is likely due to the presence of snow cover in the permafrost regions, which introduces additional scattering and attenuates the radar signal, resulting in lower coherence values. Additionally, areas with high vegetation cover (such as the elliptical areas) show severe decorrelation, while non-vegetated areas containing buildings, rocks, and bare soil (such as the triangular areas) display stronger coherence. Water bodies (such as the rectangular areas) have the lowest coherence.

After SBAS-InSAR processing, LOS deformation rate maps were generated for both ascending and descending orbits, as seen in Figure 9. The underlying relief map of the study area is included. Cool colors indicate movement towards the sensor’s LOS, while warm colors indicate movement away from the sensor’s LOS. Blank areas represent data discarded due to coherence falling below the threshold (0.7). The results from both ascending and descending orbits align well, showing significant subsidence signals in the northeastern part of the study area and notable surface uplift signals in the southwestern part. Moreover, the subsidence rates in the permafrost regions are higher than those in the seasonally frozen ground areas, with surface uplift trends observed around lakes.

4.2. Establishment of the Two-Dimensional Surface Deformation Field and Cross-Validation with GNSS Data

To verify the feasibility of this study, SBAS-InSAR results from both ascending and descending orbits were extracted within a 25 m radius buffer around the QHGC site. Using inverse distance weighting interpolation, the deformation values at the QHGC site were obtained. These values were then used to generate two-dimensional surface deformation time series (vertical and east–west directions), which were compared with the processed GNSS data.

It is important to note that the QHGC site is located in a region with flat terrain and high coherence, where ascending and descending orbit results are output as regularly gridded data. These conditions provide a high degree of accuracy for the derived results. However, in regions with complex terrain, or when irregular point data are directly output without geocoding, challenges such as data gaps may arise. This can ultimately result in difficulties or inaccuracies in generating effective two-dimensional surface deformation fields.

Initially, the results from the ascending orbit and the interpolated results from the descending orbit at the QHGC site were input into Equation (5) to calculate the two-dimensional deformation field at the site, as shown in Figure 10. Upon inspection of the results, it is evident that there are deviations between the surface vertical deformation and the east–west displacement derived from the method used in this study and the initial data from the ascending and descending orbits. These discrepancies arise from the inherent nature of InSAR measurements, which provide line of sight (LOS) displacement data. The LOS displacement includes both vertical and horizontal components, and their relationship is determined by the satellite’s viewing geometry, with a specific angle of projection between them [78].

Subsequently, the derived vertical and east–west InSAR results were compared with the processed GNSS data. The comparison indicates a high degree of consistency between the surface deformation fields obtained in this study and the GNSS data, as illustrated in Figure 11.

Figure 12 demonstrates that, compared to GNSS data, the R² and RMSE values for the vertical ascending InSAR results (converted from LOS to vertical based on the ascending incidence angle) are 0.887 and 4.875 mm, respectively; for the interpolated vertical descending InSAR results (converted from LOS to vertical based on the descending incidence angle), the values are 0.814 and 4.088 mm; for the vertical InSAR results obtained using the methods described in this study, the values are 0.985 and 0.669 mm; and for the east–west InSAR results, they are 0.983 and 0.705 mm, respectively. These results strongly suggest that the SBAS-InSAR results, after spatiotemporal interpolation, can more accurately reflect surface deformations, providing a solid data foundation for studying the freeze–thaw cycles of permafrost. Additionally, it was observed that in the study area, the ascending InSAR results tend to be higher, whereas the descending InSAR results tend to be lower.

4.3. Analysis of Vertical Surface Deformation Time Series in Typical Regions

To ensure a comprehensive understanding of surface deformation patterns in the permafrost regions of the Qilian Mountains, vertical surface deformation time series data were analyzed for six representative points within the study area (Figure 13). These points, strategically distributed across areas covered by permafrost, reveal distinct, regular deformation patterns.

DS1 and DS2 are situated in the seasonally frozen ground region. The vertical surface deformation time series for DS1 and DS2 are shown in Figure 13h and Figure 13i, respectively. Both locations exhibit notable stability. DS1, located in a flat, arid area, as depicted in Figure 13b, demonstrates a highly stable deformation time series with minimal fluctuations. In contrast, DS2, positioned in a flat area near a stream as shown in Figure 13c, displays more pronounced regular fluctuations. This is due to its proximity to a water source, which significantly influences soil moisture levels. Annually, the surface tends to bulge from September to March of the following year and subsides from March to September. This pattern is driven by the freezing and thawing of the soil’s water content, which in turn causes surface deformations due to the volumetric changes in water and ice.

DS3, DS4, DS5, and DS6 are located in the permafrost region, where the analyzed vertical surface deformation time series clearly reflect the freeze–thaw cycles typical of permafrost. DS3, found in a flat, dry area (see Figure 13d), shows the highest subsidence rate among the six selected representative points, at 9.25 mm/a, with an average seasonal deformation amplitude of approximately 31.43 mm. A gradual increase in seasonal amplitude over time indicates significant ice loss and a thickening active layer. DS4, located in a low-lying damp area (see Figure 13e), with an uplift rate of 1.54 mm/a and an average seasonal deformation amplitude of approximately 44.93 mm, exhibits significant seasonal deformations with a consistent seasonal amplitude over the years but demonstrates a trend of surface uplift. This uplift may be due to the nearby stream providing a steady water supply, which aids in the upward freezing of the active layer, thus causing surface uplift. DS5 and DS6, positioned on slopes with the same gradient but different aspects (see Figure 13f,g), exhibit smaller annual deformation rates and seasonal changes compared to flat areas. Specifically, DS5 has a subsidence rate of 3.44 mm/a with an average seasonal deformation amplitude of approximately 29.19 mm, while DS6 shows a subsidence rate of 1.06 mm/a and an average seasonal deformation amplitude of approximately 26.45 mm. The comparative analysis of Figure 13l,m reveals that the subsidence rate and the magnitude of seasonal deformation on the shaded slope are greater than those on the sunny slope, with a slightly larger increase in seasonal deformation amplitude as well. This disparity is attributed to the reduced sunlight exposure and weaker solar radiation received by the shaded slope, resulting in lower temperatures and less evaporation of soil moisture. Consequently, the higher moisture content contributes to a greater annual and seasonal deformation rate on the shaded slope compared to the sunny slope.

4.4. Spatial Distribution of Annual Vertical Surface Deformation Rates

Figure 14a displays the spatial distribution of long-term vertical deformation rates across the study area. The spatial distribution of annual deformation rates effectively reflects the characteristics of permafrost distribution within the region. Areas with continuous permafrost exhibit a faster subsidence trend, while regions with seasonally frozen ground show smaller subsidence trends. Additionally, due to the proximity to lakes, these areas have higher soil moisture content, which is associated with more frequent surface uplift signals. These patterns are consistent with the LOS surface deformation rate results. The range of annual vertical deformation rates in the study area varies from −63.93 mm/yr to 60.83 mm/yr, primarily concentrated between −20 mm/yr and 20 mm/yr. Red areas indicate surface uplift, while blue areas denote subsidence. Points in the Figure with temporal coherence less than 0.7 are excluded from the results as they are considered unreliable. Lower coherence (or decoherence) may be due to obstructions from surface cover, and reflections from snow and lakes, preventing satellite signals from penetrating the surface. Observations also indicate that between Datong Mountain and Tuolai Mountain, in the valley areas, there are significant surface uplift rates, likely influenced by factors such as landslide deposits and tectonic movements.

4.5. Spatial Distribution of Seasonal Deformation in Permafrost

Figure 14b shows the spatial distribution of seasonal deformation. This pattern also effectively reflects the permafrost distribution trends. The figure indicates an average seasonal deformation amplitude of 35 mm. In areas with continuous permafrost, the seasonal deformation values are significant and more concentrated; in patchy permafrost and seasonally frozen ground regions near lakes, due to higher soil moisture content, the seasonal deformation amplitude near the lakes is also higher and decreases with increasing distance from the lakes, becoming more dispersed. Furthermore, a correlation exists between seasonal deformation and annual deformation rates; areas with significant subsidence often exhibit larger seasonal deformation amplitudes, and regions with intense seasonal deformation typically experience significant subsidence. This is because seasonal deformations are influenced by the water content in the active layer, with freeze–thaw cycles driven by temperature changes affecting the ice-water content in the active layer, thereby causing surface deformation.

4.6. Analysis of Interannual Deformation Rate and the Influencing Factors of Seasonal Deformation

4.6.1. Factors Influencing Interannual Deformation Rate

As depicted in Figure 15a, apart from PZI and HAA, all factors explored in this study passed the significance test, each explaining the interannual deformation rate of permafrost to varying degrees. Specifically, ET, SD, Land Cover, SFD, SMsurf, and LST explain more than 50% of the variations in the permafrost deformation rate within the study area. Furthermore, MAP, BGT, EVI, GPP, SMroot, NPP, NDVI, Slope, Aspect, MAATMN, Altitude, MAATMX, and MAAT account for more than 40% of the explanatory power. FVC, SWDR, ALT, LAI, LAI, PGC23, SWE, MTSFG, and PET contribute over 20%. PGC50, PGC35, PZI, and HAA show less than 20% explanatory power.

4.6.2. Factors Influencing the Amplitude of Seasonal Deformation

According to Figure 15b, except for HAA, all explored factors passed the significance test. Among these, SMroot, SMsurf, SFD, and LST explain more than 50% of the seasonal deformation amplitude. SD, BGT, SWE, MAP, MAAT, MAATMN, MAATMX, ALT, and Altitude explain more than 40%. EVI, NPP, LAI, ET, NDVI, GPP, PET, Aspect, Slope, FVC, and PZI contribute more than 20%. PGC35, SWDR, MTSFG, PGC23, PGC50, Land Cover, and HAA have less than 20% explanatory power, with Land Cover at 11.8%—primarily because the study area is predominantly grassland, showing minimal change; HAA is only 8.1%, suggesting minimal impact from human activities on the seasonal deformation of permafrost.

Analysis reveals that factors with significant explanatory power for the surface vertical interannual deformation rates and seasonal deformation amplitude include altitude, slope, and aspect, which describe the local topographical features and surface processes. MAAT, MAATMN, MAATMX, and LST illustrate local air and surface temperature variations. FVC, EVI, NDVI, GPP, NPP, and LAI primarily reflect vegetation growth and coverage. Soil moisture directly affects vegetation conditions, where sufficient soil moisture supports photosynthesis and growth. Thus, these factors also indirectly indicate local soil moisture conditions. Conversely, ET, PET, SMsurf, SMroot, PGC, SWE, SD, and MAP directly represent local soil moisture conditions, with variations in MAP significantly influencing soil moisture dynamics.

Therefore, the factors with strong explanatory power for interannual deformation rates and seasonal deformation can be categorized into climate factors (temperature, soil moisture, precipitation) and topographical factors (altitude, slope, aspect), which directly impact the local permafrost environment, posing significant challenges to human habitation and infrastructure development in the region.

4.7. Relationship Between Permafrost Freeze–Thaw-Induced Surface Deformation and Topographical Factors

Based on Figure 15, there is a clear correlation between vertical surface deformation of permafrost and local topographical factors. This section focuses on three main aspects: altitude, slope, and aspect. To mitigate the influence of lakes and streams on the surface deformation monitoring results, the transect line AB at location Figure 16b and the boxed area in Figure 16c were selected for the analysis of topographical factors.

4.7.1. The Impact of Altitude on Permafrost Deformation

Altitude directly affects the surface temperature range and the duration of the warm period, thereby influencing the process of frost heave and thaw settlement in permafrost. Figure 17 illustrates the relationship between the seasonal deformation amplitude, the vertical interannual deformation rate of permafrost, and altitude. Analysis indicates that in higher altitude regions (altitude greater than 4500 m), the seasonal deformation amplitude of permafrost is smaller, and the vertical interannual deformation rate is lower. In contrast, in lower altitude regions (altitude less than 4100 m), both the seasonal deformation amplitude and the vertical interannual deformation rate of permafrost increase. Therefore, a negative correlation can be summarized between altitude and both the seasonal deformation amplitude and the vertical interannual deformation rate of permafrost: the higher the altitude, the lower the vertical interannual deformation rate, and the smaller the seasonal deformation amplitude. Specifically, the Pearson correlation coefficient between altitude and the seasonal deformation amplitude is −0.256 with a p-value < 0.005, while the Pearson correlation coefficient between altitude and the vertical interannual deformation rate is −0.371 with a p-value < 0.005. Moreover, from Figure 17, a weak negative correlation is also observed between the time of maximum subsidence and altitude, which may be due to the shorter warm periods at higher altitudes, leading to earlier maximum subsidence times in permafrost.

4.7.2. The Impact of Slope on Permafrost Deformation

Among topographical factors, slope significantly affects the direction and concentration of soil moisture transport, which in turn influences the development of permafrost. Figure 17 illustrates the relationship between slope and surface deformation due to frost heave and thaw settlement in permafrost along transect line AB. The figure reveals that areas with substantial vertical deformation rates and large seasonal deformation amplitudes are primarily found in regions with slopes of 0–10°, especially on flat grounds and near mountain bases. Statistical analysis shows a Pearson correlation coefficient of −0.41 (p-value < 0.005) between slope and the seasonal deformation amplitude, and a Pearson correlation coefficient of −0.279 (p-value < 0.005) between slope and the vertical interannual deformation rate. In steeper mountainous areas (slopes greater than 15°), the lower temperatures and the inconsistent and insufficient water supply result in a relatively thinner active layer of permafrost. Additionally, snow cover serves as an insulator in these areas, stabilizing the internal structure of the permafrost and leading to smaller vertical deformation rates and seasonal changes compared to flatter regions. This pattern is particularly noticeable in the 0–12 km section of the transect line. As the distance along the transect line increases, the slope gradually decreases, and the distribution of swamps and ponds, along with sedimentation, thickens. Consequently, soil moisture rises gradually, and the development of permafrost improves, leading to a decrease in both the vertical deformation rate and the seasonal deformation amplitude, showing a negative correlation. This pattern aligns with typical permafrost deformation patterns illustrated in Figure 13.

4.7.3. The Impact of Aspect on Permafrost Deformation

Aspect affects the amount of solar radiation a surface receives, thereby indirectly impacting local temperature variations. The study area, as shown in Figure 16c, is categorized into sunny and shaded slopes based on topography and sunlight exposure. Figure 18 displays histograms of the vertical surface deformation rates and seasonal deformation amplitudes of permafrost for these locations.

The average vertical deformation rates for sunny and shaded slopes are 0.75 mm/a and −2.02 mm/a, respectively, with average seasonal deformation amplitudes of 25.32 mm and 30.13 mm. Shaded slopes exhibit both greater annual subsidence rates and larger seasonal deformation amplitudes compared to sunny slopes.

4.8. The Relationship Between Permafrost Freeze–Thaw-Induced Surface Deformation and Climatic Factors

Apart from topographical factors, permafrost is highly sensitive to temperature and soil moisture. Temperature changes trigger the “frost heave and thaw settlement” process, leading to surface deformation. Higher soil moisture enhances the soil’s heat capacity and thermal conductivity, thereby accelerating the melting of the active layer. Precipitation is a primary determinant of soil moisture; more frequent and abundant rainfall typically increases soil moisture levels. This section discusses the response of permafrost surface deformation to air temperature, soil moisture, and precipitation, further clarifying the correlation between permafrost and environmental factors.

4.8.1. Response of Permafrost Deformation to Temperature

This section utilizes average temperature data from the Gangcha meteorological station within the study area to analyze the temperature response patterns at locations DS3 and DS4 (Figure 19). The results indicate a clear negative correlation between temperature and surface deformation: as temperatures rise, the thawing of the active layer leads to surface subsidence; conversely, as temperatures fall, the freezing of the active layer results in surface uplift. Notably, DS3 is situated in an alpine meadow area near a stream, while DS4 is located in a drier alpine barren land area.

Observations at both sites reveal that the annual maximum surface subsidence (melting point) consistently occurs later than the peak temperatures (time differences Δ_t1 ≈ 117 and Δ_t5 ≈ 90, respectively), and also later than the temperatures falling to 0 °C. Similarly, the annual maximum surface uplift (frost heave point) occurs after the lowest temperatures of the year (time differences Δ_t2 ≈ 105 and Δ_t6 ≈ 97, respectively), and also later than the temperatures rising back to 0 °C. The DS3 area, with its higher soil moisture, exhibits more pronounced seasonal deformations and thus shows greater time differences, Δ_t3 ≈ 22 and Δ_t4 ≈ 23, between the deformation extremes and the corresponding 0 °C timings. For DS4, the time differences between the deformation extremes and the corresponding 0 °C timings are Δ_t7 ≈ 2 and Δ_t8 ≈ 13. This indicates a certain lag in surface deformation response to temperature changes, a phenomenon influenced by factors such as solar radiation, soil properties, and vegetation cover, which cause discrepancies between soil temperature, surface temperature, and air temperature. Additionally, the water-rich active layer of permafrost influences the timing of freeze–thaw cycles due to the heat absorption which occurs during ice melting and heat release during water freezing. However, a large-scale statistical analysis across the entire study area is not feasible due to the limitations of the available data. The long-term temperature data for the Qilian Mountains region has a low resolution, which prevents accurate statistical analysis and reduces the generalizability of the results.

Additionally, analysis from Section 4.7.3 on deformation patterns across different slope aspects also demonstrates that under similar conditions, weaker solar radiation leads to less evaporation of soil moisture, thereby increasing the magnitude of seasonal deformations.

4.8.2. Response of Permafrost Deformation to Soil Moisture and Precipitation

The freeze–thaw cycles of permafrost are closely related to variations in soil moisture. Soil moisture significantly influences the dynamics of permafrost freezing and thawing through mechanisms that affect heat capacity, thermal conductivity, the latent heat of phase changes, and moisture migration. These factors, in turn, impact the stability and distribution of permafrost, thereby affecting surface deformation. In the context of global climate change, the role of soil moisture in permafrost dynamics is particularly critical, holding significant implications for human efforts to predict and manage permafrost degradation.

Precipitation is the primary source of soil moisture. Figure 20a illustrates the temporal relationship between local precipitation and surface deformation at locations DS3 and DS4. It is evident that precipitation primarily occurs during the thawing stage of permafrost, directly increasing the moisture content in the soil. Moist soil, with its higher thermal conductivity and heat capacity, conducts and stores heat more efficiently. This leads to a thickening of the active layer, making the soil more susceptible to seasonal temperature variations, and subsequently increasing the amplitude of seasonal deformation. Moreover, precipitation impacts the freeze–thaw processes of permafrost differently across seasons. Spring and summer precipitation rapidly increases surface moisture, accelerating permafrost thawing. At the onset of the rainy season each year, the rate of vertical subsidence of the surface begins to accelerate, reaching its maximum in August when precipitation peaks. Conversely, winter snow acts as an insulator for permafrost, with surface snow helping to prevent thermal exchange between the surface and the atmosphere, thereby slowing the freezing process.

Data on soil moisture at locations DS3 and DS4 was analyzed to assess its impact on the freeze–thaw state of permafrost, as shown in Figure 20b. The soil moisture at DS4 is notably higher than at DS3, which means that during summer thawing, the thaw depth at DS4 is greater; similarly, during winter freezing, the freeze depth is also greater, resulting in markedly larger seasonal deformations at DS4. This is because soil with higher moisture content possesses greater heat capacity and thermal conductivity, enabling it to store and transfer more heat. During the permafrost freeze–thaw cycle, moist soil responds more quickly to temperature changes, leading to more pronounced freeze–thaw changes and more intense seasonal surface deformations. Conversely, dry soils tend to be more stable, with relatively smaller seasonal surface deformations. Over longer time scales, as global warming progresses and precipitation patterns shift, increasing soil moisture will further accelerate permafrost degradation, significantly affecting the freeze–thaw processes of permafrost.

5. Discussion

5.1. Monitoring Surface Deformation Using SBAS-InSAR Technology

Figure 14 showcases the annual vertical deformation rates and seasonal deformation amplitudes across the study area, revealing the dynamic characteristics of permafrost. The annual vertical deformation rates mainly range from −20 mm/a to 20 mm/a, with an average of approximately 1.56 mm/a, indicating a general trend of subsidence in the permafrost region. This highlights the significant impact of the freeze–thaw cycles on surface deformation [79]. Studies have found that the spatial distribution of annual vertical deformation rates is closely linked to the distribution of permafrost. In areas with continuous permafrost, a clear trend of surface subsidence is evident. In contrast, areas with seasonally frozen ground, particularly those with higher soil moisture, show less subsidence and sometimes even localized uplift. This suggests that subsidence in permafrost regions is significantly connected to soil moisture content and the freeze–thaw processes of the active layer. These findings are consistent with those of Liu et al. [79,80]. Additionally, in mountainous regions, particularly in valleys, there is a common occurrence of higher surface uplift rates, illustrating that topographic and geological factors also play a crucial role in the surface deformation of permafrost regions [79]. Regarding the seasonal deformation amplitude of permafrost, the average seasonal deformation amplitude is about 35 mm, with wet areas (such as near rivers and lakes) exhibiting larger seasonal deformation amplitudes. The larger seasonal deformation amplitudes in high-moisture areas are usually accompanied by significant surface subsidence, indicating a correlation between annual deformation rates and seasonal deformation amplitudes. This phenomenon can be attributed to the ice–water phase change processes within the active layer, with temperature fluctuations intensifying the freeze–thaw cycles and thus exacerbating the degree of surface deformation [79].

In addition to surface deformation, the thawing of permafrost due to repeated freeze–thaw cycles can lead to significant environmental concerns. As permafrost thaws, it releases trapped greenhouse gasses such as methane and carbon dioxide, which contribute to global warming. Furthermore, the destabilization of permafrost can affect local hydrology, leading to changes in water flow and soil erosion, which may further exacerbate the degradation of ecosystems and infrastructure in vulnerable areas [11].

5.2. Factors Affecting Permafrost Deformation

Permafrost surface deformation is influenced by both topographic and climatic factors, and understanding these factors is critical for assessing the stability of permafrost in a changing climate. Monitoring permafrost evolution is strategically important because thawing permafrost has broad implications for environmental stability and climate change.

Elevation is a key factor influencing surface deformation in permafrost regions. As elevation increases, temperatures typically drop, leading to deeper freeze depths and a thinner active layer. This results in reduced subsidence and smaller seasonal deformation amplitudes. Conversely, at lower elevations, particularly in valleys, warmer temperatures lead to less extensive freezing and greater surface subsidence. Similar results were observed by Daout et al. [55]. The critical point here is that in lowland areas, where permafrost is more susceptible to thawing due to higher temperatures, the risk of accelerated degradation increases.

Slope and aspect further influence subsidence patterns. Flat areas, especially near mountain bases, show larger seasonal deformation amplitudes, while steep slopes tend to have thinner active layers and reduced deformation due to moisture drainage; this supports the conclusions of Li et al. [81]. On shaded slopes, where solar radiation is limited, moisture retention leads to increased subsidence, similar results were observed by Deng et al. [82]. This phenomenon is likely due to the lower temperatures, reduced sunlight duration, and weaker solar radiation on shaded slopes, which make soil moisture less likely to evaporate, resulting in higher water content. This contributes to greater subsidence rates and deformation amplitudes. During the summer, when the sun’s elevation is higher, both slopes are exposed to solar radiation, minimizing the temperature difference and leading to the melting of the permafrost on both slopes. However, in winter, with the sun at a lower elevation, the shaded slope receives little to no solar radiation, whereas the sunny slope continues to be exposed to sunlight for extended periods. This difference creates significant temperature variations between the two locations, resulting in smaller freeze expansion in the permafrost on the sunny slope and greater expansion on the shaded slope. This highlights another critical aspect: areas with greater moisture retention are more vulnerable to thawing, which could lead to significant environmental and infrastructural impacts [83].

Climatic factors like temperature, precipitation, and soil moisture content also drive deformation. Temperature is perhaps the most direct and influential factor affecting the freeze–thaw cycles in permafrost regions. As temperatures increase, the active layer of the permafrost—where freeze–thaw cycles occur—becomes deeper, accelerating the thawing process. This results in increased surface subsidence due to the melting of ice within the soil, supporting the conclusions of Zhang et al. [57]. Moreover, this temperature-induced thawing creates a feedback loop, where thawing further reduces permafrost stability, leading to additional surface subsidence and amplification of climate change effects [64].

Soil moisture and precipitation are closely linked factors that significantly influence permafrost dynamics. Precipitation, in the form of rainfall or snowmelt, directly affects the moisture content of the soil, which in turn has a critical role in the thawing of permafrost. When precipitation levels are high, whether during rainfall events or through snowmelt in the spring, the soil becomes more saturated. This increased moisture content enhances the soil’s thermal conductivity, allowing heat to penetrate more deeply into the permafrost. This supports the conclusions of Li et al. [84]. If these critical climatic thresholds are surpassed, the permafrost system may undergo irreversible changes, posing risks to local ecosystems and infrastructure [57].

However, the study’s limitation is that temperature data are available only from a single meteorological station in the study area, restricting the spatial scope of the analysis. While this limits the generalization of the findings across the entire region, it allows for accurate analysis at the localized station site. Future studies with more stations could provide a broader understanding of temperature variations and their impact on permafrost.

5.3. The Five Stages of Seasonal Deformation in Permafrost

Analyzing the time series of vertical surface deformation across various regions and understanding the dynamic relationship between temperature and surface deformation, we can divide permafrost seasonal deformation into five stages: the warming and melting phase (including the summer melting process and the warm season stabilization process) and the cooling and freezing phase (including the freezing uplift process, the winter cooling process, and the spring warming process) [85,86,87,88,89], as seen in Figure 21.

5.3.1. Warming and Melting Phase

The warming and melting phase consists of two main processes:

Summer Melting Process (P1): This process occurs from mid to late May through mid to late August. During this period, the active layer begins to melt from the surface down to its maximum thaw depth, creating a negative temperature gradient. As temperatures gradually increase, the surface ice and snow start to melt, precipitation increases, and the atmospheric water cycle becomes more active. The active layer absorbs heat and transfers it downward, causing internal temperatures to rise and the thawing front to descend. Free water near the thawing front moves downward due to the combined effects of gravity and temperature gradients, resulting in the gradual thawing of the active layer from top to bottom. As temperatures continue to rise, the upper parts of the active layer absorb more heat, leading to gradual surface subsidence [89].

Warm Season Stabilization Process (P2): This process occurs typically from mid to late August through mid to late September, during which the active layer has thawed to its maximum depth, and surface deformation gradually decreases. After reaching the highest surface temperatures of the year, the temperatures begin to decrease. The surface layer of the active layer absorbs heat during the day and releases it at night, initiating a diurnal freeze–thaw cycle that results in minor surface deformations [88].

5.3.2. Cooling and Freezing Phase

The cooling and freezing phase is divided into three main processes:

Freezing Uplift Process (P3): Typically occurring from early October to early November, this process involves the active layer freezing unidirectionally from the bottom up, as well as bidirectionally from both the top down and the bottom up, entering a state known as the “zero-curtain” phase [88]. Generally, as latitude and elevation increase, the mean annual ground surface temperature decreases, enhancing the proportion of bottom-up freezing. Once freezing of the active layer is complete, its internal temperature distribution shows colder temperatures at the top and warmer at the bottom, initiating the winter cooling process.

Winter Cooling Process (P4): This phase typically lasts from the end of the active layer’s freezing until early February of the following year. During this time, the surface temperature continues to drop, and the positive temperature gradient in the active layer gradually increases. The ice–water phase change process almost comes to a halt, resulting in minimal surface deformation.

Spring Warming Process (P5): Typically occurring from late February to mid-May [88], this process begins as surface temperatures start to warm. The positive temperature gradient in the active layer decreases, evaporation intensifies, and the surface of the active layer undergoes a daily freeze–thaw cycle. These changes lead to slight surface deformations and a decrease in surface stability [85].

5.4. Applications of Two-Dimensional Surface Deformation Fields Under Spatiotemporal Interpolation

This study achieves a unified spatiotemporal dimension for ascending and descending orbit InSAR results through geographic and cubic spline interpolation methods, addressing inconsistencies in space and time. Subsequently, based on these results and considering the geometric relationship between the satellite trajectory and the ground, we calculate two-dimensional surface deformation fields in both vertical and east–west directions. We verified the accuracy of the vertical surface deformations and east–west displacements by comparing them with GNSS data from the study area, demonstrating that this method is suitable for long-term and large-area surface deformation monitoring.

However, this method does have some limitations related to error accumulation. During the spatiotemporal interpolation process, fluctuations in surface deformation caused by temperature, precipitation, and snowfall are somewhat overlooked, potentially resulting in the loss of detailed information in the two-dimensional surface deformation fields. Moreover, the Sentinel-1A data are limited by its revisit cycle; although it can reflect the freeze–thaw patterns of permafrost over long periods, it cannot provide daily, all-weather deformation information. Future approaches could involve using multi-source high-resolution satellite fusion to shorten monitoring intervals and enhance the precision of surface deformation monitoring.

5.5. Applications of Dynamic Periodic Deformation Model

In processing the SBAS-InSAR results of this study, we used a dynamic periodic deformation model to more precisely calculate the vertical time series of the study area’s surface. By calculating the annual deformation rate and seasonal deformation amount through parameter estimation, this approach not only significantly reduced the number of unknowns but also stabilized the model, avoiding numerical fluctuations in the time series [54,72,90]. Additionally, it allowed for a more detailed reflection of the interannual variability of seasonal deformation amplitudes, aiding in further understanding the freeze–thaw cycles of permafrost and the deformation patterns caused by global warming.

However, setting up models to understand the patterns of surface deformation and permafrost freeze–thaw cycles cannot fully reflect the complex factors causing deformations in areas of permafrost and seasonally frozen ground. In permafrost regions, the factors influencing surface deformation are complex, involving not only regional temperatures and precipitation but also topography, soil moisture, soil porosity, groundwater, and thermal conductivity, among other factors. Therefore, in some regions, the periodic deformation model may not accurately reflect the actual surface deformations, leading to discrepancies between the fitted model and the actual deformation values. Additionally, because the periodic deformation model emphasizes regularity, it may not fully capture deformations caused by temperature fluctuations, precipitation, and winter snow accumulation.

In addition, another limitation of this study is the limited consideration of human activities and infrastructure in areas of severe subsidence, particularly near traffic roads or construction zones. While this study primarily focuses on the impact of natural factors on permafrost surface deformation, the influence of man-made structures on permafrost stability is an important aspect that warrants further investigation. Future studies could integrate remote sensing techniques, such as SBAS-InSAR, with ground-based observations to assess the long-term effects of infrastructure development on permafrost dynamics.

6. Conclusions

This study utilized Sentinel-1A SAR data from 2017 to 2023, applying the SBAS-InSAR technique with spatiotemporal interpolation to monitor the alpine permafrost regions of the Qilian Mountains. It yielded the following conclusions:

Spatiotemporal Interpolation: This study innovatively integrated ascending and descending orbit SAR data, generating a two-dimensional surface deformation field. The vertical deformation ranged from −20 mm/a to 20 mm/a, with an average rate of 1.56 mm/a. Subsidence dominated in permafrost regions, while areas near Qinghai Lake showed surface uplift due to high soil moisture.
Dynamic Periodic Deformation Model: The first use of this model to analyze vertical time-series deformation provided insights into seasonal amplitude changes. The average seasonal deformation amplitude was 35 mm, influenced by soil moisture and temperature, with a growing trend towards permafrost instability and degradation.
Factors Influencing Permafrost Deformation: Permafrost freeze–thaw cycles are influenced by both topographical factors (elevation, slope gradient, aspect) and climatic factors (temperature, soil moisture, precipitation). As elevation increases, both the annual and seasonal deformation rates generally decrease. Steeper slopes, which lose moisture more quickly, experience lower deformation rates, while gentler slopes with higher moisture accumulation show greater deformation. Northern slopes, receiving less solar radiation and having more temperature variation, tend to experience higher subsidence rates than southern slopes. Permafrost is highly sensitive to temperature and moisture. Surface deformation is inversely related to temperature changes, with some delay. As temperatures rise, the active layer melts, causing surface subsidence, while freezing temperatures cause the layer to freeze, leading to surface uplift. Moist soil accelerates permafrost melting due to its higher heat capacity and thermal conductivity, increasing the amplitude of seasonal deformation. Additionally, increased precipitation contributes to permafrost instability by raising soil moisture and enhancing percolation.
Deformation Phases: The paper delineated the surface deformation process of permafrost into a warming and melting phase and cooling and freezing phase. The warming and melting phase included the summer melting process and the warm season stability process, primarily characterized by subsidence due to the melting of the active layer. The cooling and freezing phase encompassed the frost heave process, the winter cooling process, and the spring warming process. These phases mainly displayed gradual surface uplift and minor deformations triggered by temperature fluctuations. These processes, influenced by temperature changes and the dynamic freeze–thaw of the active layer, affected the stability and deformation characteristics of the surface.

Author Contributions

Conceptualization, Z.X. and S.Z.; methodology, Z.X.; software, Z.X.; validation, Z.X.; formal analysis, Z.X.; investigation, Z.X.; resources, S.Z.; data curation, Z.X.; writing—original draft preparation, Z.X.; writing—review and editing, S.Z. and B.Z.; visualization, Z.X.; supervision, S.Z. and B.Z.; project administration, S.Z.; funding acquisition, S.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Natural Science Foundation of China (42271432, 42171424), the Natural Science Foundation of Shanxi Province (202403021211007), the ESA project Dragon 6 (95177), and the Cooperation and Exchange Program by the Science and Technology Department of Shanxi Province (202304041101051).

Data Availability Statement

The original contributions presented in the study are included in the article data, further inquiries can be directed to the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

References

Ding, J.Z.; Chen, L.Y.; Ji, C.J.; Hugelius, G.; Li, Y.N.; Liu, L.; Qin, S.Q.; Zhang, B.B.; Yang, G.B.; Li, F.; et al. Decadal Soil Carbon Accumulation across Tibetan Permafrost Regions. Nat. Geosci. 2017, 10, 420–424. [Google Scholar] [CrossRef]
Webb, E.E.; Liljedahl, A.K. Diminishing Lake Area across the Northern Permafrost Zone. Nat. Geosci. 2023, 16, 202–209. [Google Scholar] [CrossRef]
Li, C.Y.; Zhao, L.; Wang, L.X.; Liu, S.B.; Zhou, H.Y.; Li, Z.; Liu, G.Y.; Du, E.; Zou, D.F.; Hou, Y.X. Ground Deformation and Permafrost Degradation in the Source Region of the Yellow River, in the Northeast of the Qinghai-Tibet Plateau. Remote Sens. 2023, 15, 3153. [Google Scholar] [CrossRef]
Schuur, E.A.G.; McGuire, A.D.; Schädel, C.; Grosse, G.; Harden, J.W.; Hayes, D.J.; Hugelius, G.; Koven, C.D.; Kuhry, P.; Lawrence, D.M.; et al. Climate Change and the Permafrost Carbon Feedback. Nature 2015, 520, 171–179. [Google Scholar] [CrossRef]
Zou, D.F.; Zhao, L.; Sheng, Y.; Chen, J.; Hu, G.J.; Wu, T.H.; Wu, J.C.; Xie, C.W.; Wu, X.D.; Pang, Q.Q.; et al. A New Map of Permafrost Distribution on the Tibetan Plateau. Cryosphere 2017, 11, 2527–2542. [Google Scholar] [CrossRef]
Wang, L.X.; Zhao, L.; Zhou, H.Y.; Liu, S.B.; Du, E.; Zou, D.F.; Liu, G.Y.; Wang, C.; Li, Y. Permafrost Ground Ice Melting and Deformation Time Series Revealed by Sentinel-1 InSAR in the Tanggula Mountain Region on the Tibetan Plateau. Remote Sens. 2022, 14, 811. [Google Scholar] [CrossRef]
Li, R.; Wu, Q.B.; Li, X.; Sheng, Y.; Hu, G.; Cheng, G.D.; Zhao, L.; Jin, H.J.; Zou, D.F.; Wu, X.D. Characteristic, Changes and Impacts of Permafrost on Qinghai-Tibet Plateau. Chin. Sci. Bull. 2019, 64, 2783–2795. [Google Scholar] [CrossRef]
Kang, S.C.; Xu, Y.W.; You, Q.L.; Flügel, W.-A.; Pepin, N.; Yao, T.D. Review of Climate and Cryospheric Change in the Tibetan Plateau. Environ. Res. Lett. 2010, 5, 015101. [Google Scholar] [CrossRef]
Wang, Q.X.; Fan, X.H.; Wang, M.B. Recent Warming Amplification over High Elevation Regions across the Globe. Clim. Dyn. 2014, 43, 87–101. [Google Scholar] [CrossRef]
Zhang, G.Q.; Yao, T.D.; Piao, S.L.; Bolch, T.; Xie, H.J.; Chen, D.L.; Gao, Y.H.; O’Reilly, C.M.; Shum, C.K.; Yang, K.; et al. Extensive and Drastically Different Alpine Lake Changes on Asia’s High Plateaus during the Past Four Decades. Geophys. Res. Lett. 2017, 44, 252–260. [Google Scholar] [CrossRef]
Cheng, G.D.; Wu, T.H. Responses of Permafrost to Climate Change and Their Environmental Significance, Qinghai-Tibet Plateau. J. Geophys. Res. Earth Surf. 2007, 112, 2006JF000631. [Google Scholar] [CrossRef]
Yang, M.; Nelson, F.E.; Shiklomanov, N.I.; Guo, D.; Wan, G. Permafrost Degradation and Its Environmental Effects on the Tibetan Plateau: A Review of Recent Research. Earth Sci. Rev. 2010, 103, 31–44. [Google Scholar] [CrossRef]
He, Z.J.; Xu, X.C.; Zhong, Z.T.; Li, X.; Liu, X.P. Spatial-Temporal Variations Analysis of Snow Cover in China from 1992–2010. Chin. Sci. Bull. 2018, 63, 2641–2654. [Google Scholar] [CrossRef]
Li, D.-L. Spatial-Temporal Variations of Snow Cover Days and the Maximum Depth of Snow Cover in China during Recent 50 Years. J. Glaciol. Geocryol. 2012, 2, 247–256. [Google Scholar] [CrossRef]
Ma, Q.Q.; Keyimu, M.; Li, X.Y.; Wu, S.X.; Zeng, F.J.; Lin, L.S. Climate and Elevation Control Snow Depth and Snow Phenology on the Tibetan Plateau. J. Hydrol. 2023, 617, 128938. [Google Scholar] [CrossRef]
Tian, H.Z.; Yang, T.B.; Liu, Q.P. Climate Change and Glacier Area Shrinkage in the Qilian Mountains, China, from 1956 to 2010. Ann. Glaciol. 2014, 55, 187–197. [Google Scholar] [CrossRef]
Sun, M.P.; Liu, S.Y.; Yao, X.J.; Guo, W.Q.; Xu, J.L. Glacier Changes in the Qilian Mountains in the Past Half-Century: Based on the Revised First and Second Chinese Glacier Inventory. J. Geogr. Sci. 2018, 28, 206–220. [Google Scholar] [CrossRef]
Chen, R.S.; Han, C.T.; Wang, L.; Zhao, Y.N.; Song, Y.X.; Yang, Y.; Liu, J.F.; Liu, Z.W.; Wang, X.Q.; Guo, S.H.; et al. Impact of the Alpine Precipitation Measurements on the Precipitation in 2019 and 2020 in the Qilian Mountains. J. Glaciol. Geocryol. 2023, 45, 676–687. [Google Scholar] [CrossRef]
Zhang, L.; Zhang, Q.; Feng, J.Y.; Bai, H.Z.; Zhao, J.H.; Xu, X.H. A Study of Atmospheric Water Cycle over the Qilian Mountains(I):Variation of Annual Water Vapor Transport. J. Glaciol. Geocryol. 2014, 36, 1079–1091. [Google Scholar] [CrossRef]
Massonnet, D.; Rossi, M.; Carmona, C.; Adragna, F.; Peltzer, G.; Feigl, K.; Rabaute, T. The Displacement Field of the Landers Earthquake Mapped by Radar Interferometry. Nature 1993, 364, 138–142. [Google Scholar] [CrossRef]
Du, Q.S.; Li, G.Y.; Chen, D.; Zhou, Y.; Qi, S.S.; Wu, G.; Chai, M.T.; Tang, L.Y.; Jia, H.L.; Peng, W.L. SBAS-InSAR-Based Analysis of Surface Deformation in the Eastern Tianshan Mountains, China. Front. Earth Sci. 2021, 9, 729454. [Google Scholar] [CrossRef]
Massonnet, D.; Feigl, K.L. Radar Interferometry and Its Application to Changes in the Earth’s Surface. Rev. Geophys. 1998, 36, 441–500. [Google Scholar] [CrossRef]
Hanssen, R.F.; Weckwerth, T.M.; Zebker, H.A.; Klees, R. High-Resolution Water Vapor Mapping from Interferometric Radar Measurements. Science 1999, 283, 1297–1299. [Google Scholar] [CrossRef]
Ferretti, A.; Prati, C.; Rocca, F. Nonlinear Subsidence Rate Estimation Using Permanent Scatterers in Differential SAR Interferometry. IEEE Trans. Geosci. Remote Sens. 2000, 38, 2202–2212. [Google Scholar] [CrossRef]
Berardino, P.; Fornaro, G.; Lanari, R.; Sansosti, E. A New Algorithm for Surface Deformation Monitoring Based on Small Baseline Differential SAR Interferograms. IEEE Trans. Geosci. Remote Sens. 2002, 40, 2375–2383. [Google Scholar] [CrossRef]
Yang, H.L.; Jiang, Q.; Han, J.F.; Kang, K.-Y.; Peng, J.H. InSAR Measurements of Surface Deformation over Permafrost on Fenghuoshan Mountains Section, Qinghai-Tibet Plateau. J. Syst. Eng. Electron. 2021, 32, 1284–1303. [Google Scholar] [CrossRef]
Hooper, A.; Bekaert, D.; Spaans, K.; Arıkan, M. Recent Advances in SAR Interferometry Time Series Analysis for Measuring Crustal Deformation. Tectonophysics 2012, 514–517, 1–13. [Google Scholar] [CrossRef]
Wang, S.; Xu, B.; Shan, W.; Shi, J.C.; Li, Z.W.; Feng, G.C. Monitoring the Degradation of Island Permafrost Using Time-Series InSAR Technique: A Case Study of Heihe, China. Sensors 2019, 19, 1364. [Google Scholar] [CrossRef]
Pepe, A.; Calò, F. A Review of Interferometric Synthetic Aperture RADAR (InSAR) Multi-Track Approaches for the Retrieval of Earth’s Surface Displacements. Appl. Sci. 2017, 7, 1264. [Google Scholar] [CrossRef]
Zhu, J.; Yang, Z.; Li, Z.W. Recent Progress in Retrieving and Predicting Mining-Induced 3D Displace-Ments Using InSAR. Acta Geod. Cartogr. Sin. 2019, 48, 135–144. [Google Scholar] [CrossRef]
Gatsios, T.; Cigna, F.; Tapete, D.; Sakkas, V.; Pavlou, K.; Parcharidis, I. Copernicus Sentinel-1 MT-InSAR, GNSS and Seismic Monitoring of Deformation Patterns and Trends at the Methana Volcano, Greece. Appl. Sci. 2020, 10, 6445. [Google Scholar] [CrossRef]
Shahzad, N.; Ding, X.L.; Wu, S.B.; Liang, H.Y. Ground Deformation and Its Causes in Abbottabad City, Pakistan from Sentinel-1A Data and MT-InSAR. Remote Sens. 2020, 12, 3442. [Google Scholar] [CrossRef]
Ferretti, A.; Prati, C.; Rocca, F. Permanent Scatterers in SAR Interferometry. IEEE Trans. Geosci. Remote Sens. 2001, 39, 8–20. [Google Scholar] [CrossRef]
Colesanti, C.; Ferretti, A.; Prati, C.; Rocca, F. Monitoring Landslides and Tectonic Motions with the Permanent Scatterers Technique. Eng. Geol. 2003, 68, 3–14. [Google Scholar] [CrossRef]
Li, Q.; Zhou, C.X.; Zheng, L.; Liu, T.T.; Yang, X.T. Monitoring Evolution of Melt Ponds on First-Year and Multiyear Sea Ice in the Canadian Arctic Archipelago with Optical Satellite Data. Ann. Glaciol. 2020, 61, 154–163. [Google Scholar] [CrossRef]
Ge, W.L.; Li, Y.J.; Wang, Z.C.; Zhang, C.M.; Yang, H.L. Spatial-Temporal Ground Deformation Study of Baotou Based on the PS-InSAR Method. Acta Geol. Sin. 2021, 95, 674–683. [Google Scholar] [CrossRef]
Du, Q.S.; Li, G.Y.; Zhou, Y.; Chai, M.T.; Chen, D.; Qi, S.S.; Wu, G. Deformation Monitoring in an Alpine Mining Area in the Tianshan Mountains Based on SBAS-InSAR Technology. Adv. Mater. Sci. Eng. 2021, 2021, 9988017. [Google Scholar] [CrossRef]
Han, Y.; Zhao, Y.; Zhang, Y.W.; Wang, X.B.; Wu, L.M.; Ding, P.P.; Jin, L.H. Monitoring and Analysis of Land Subsidence in Modern Yellow River Delta Using SBAS-InSAR Technology. IOP Conf. Ser. Earth Environ. Sci. 2021, 643, 012166. [Google Scholar] [CrossRef]
Hu, J.; Ge, Q.Q.; Liu, J.H.; Yang, W.Y.; Du, Z.G.; He, L.H. Constructing Adaptive Deformation Models for Estimating DEM Error in SBAS-InSAR Based on Hypothesis Testing. Remote Sens. 2021, 13, 2006. [Google Scholar] [CrossRef]
Gray, L. Using Multiple RADARSAT InSAR Pairs to Estimate a Full Three-Dimensional Solution for Glacial Ice Movement: Multiple Interferograms for 3-d Motion. Geophys. Res. Lett. 2011, 38, 46484. [Google Scholar] [CrossRef]
Samsonov, S.; d’Oreye, N. Multidimensional Time-Series Analysis of Ground Deformation from Multiple InSAR Data Sets Applied to Virunga Volcanic Province. Geophys. J. Int. 2012, 191, 1095–1108. [Google Scholar] [CrossRef]
Sousa, J.J.; Ruiz, A.M.; Hanssen, R.F.; Bastos, L.; Gil, A.J.; Galindo-Zaldívar, J.; Sanz De Galdeano, C. PS-InSAR Processing Methodologies in the Detection of Field Surface Deformation—Study of the Granada Basin (Central Betic Cordilleras, Southern Spain). J. Geodyn. 2010, 49, 181–189. [Google Scholar] [CrossRef]
Hou, J.X.; Xu, B.; Li, Z.W.; Zhu, Y.; Feng, G.C. Block PS-InSAR Ground Deformation Estimation for Large-Scale Areas Based on Network Adjustment. J. Geod. 2021, 95, 111. [Google Scholar] [CrossRef]
Liu, Y.Y.; Zhao, C.Y.; Zhang, Q.; Yang, C.S. Complex Surface Deformation Monitoring and Mechanism Inversion over Qingxu-Jiaocheng, China with Multi-Sensor SAR Images. J. Geodyn. 2018, 114, 41–52. [Google Scholar] [CrossRef]
Fialko, Y.; Simons, M.; Agnew, D. The Complete (3-D) Surface Displacement Field in the Epicentral Area of the 1999 M_W7.1 Hector Mine Earthquake, California, from Space Geodetic Observations. Geophys. Res. Lett. 2001, 28, 3063–3066. [Google Scholar] [CrossRef]
Jung, H.-S.; Lu, Z.; Won, J.-S.; Poland, M.; Miklius, A. Mapping Three-Dimensional Surface Deformation by Combining Multiple-Aperture Interferometry and Conventional Interferometry: Application to the June 2007 Eruption of Kilauea Volcano, Hawaii. IEEE Geosci. Remote Sens. Lett. 2011, 8, 34–38. [Google Scholar] [CrossRef]
Catalao, J.; Nico, G.; Hanssen, R.; Catita, C. Merging GPS and Atmospherically Corrected InSAR Data to Map 3-D Terrain Displacement Velocity. IEEE Trans. Geosci. Remote Sens. 2011, 49, 2354–2360. [Google Scholar] [CrossRef]
Guglielmino, F.; Nunnari, G.; Puglisi, G.; Spata, A. Simultaneous and Integrated Strain Tensor Estimation from Geodetic and Satellite Deformation Measurements to Obtain Three-Dimensional Displacement Maps. IEEE Trans. Geosci. Remote Sens. 2011, 49, 1815–1826. [Google Scholar] [CrossRef]
Short, N.; Brisco, B.; Couture, N.; Pollard, W.; Murnaghan, K.; Budkewitsch, P. A Comparison of TerraSAR-X, RADARSAT-2 and ALOS-PALSAR Interferometry for Monitoring Permafrost Environments, Case Study from Herschel Island, Canada. Remote Sens. Environ. 2011, 115, 3491–3506. [Google Scholar] [CrossRef]
Zhang, Y.H.; Wu, H.A.; Sun, G.T. Deformation Model of Time Series Interferometric SAR Techniques. Acta Geod. Cartogr. Sin. 2012, 41, 864–869. [Google Scholar] [CrossRef]
Liu, L.; Zhang, T.J.; Wahr, J. InSAR Measurements of Surface Deformation over Permafrost on the North Slope of Alaska. J. Geophys. Res. Earth Surf. 2010, 115, 1547. [Google Scholar] [CrossRef]
Liu, L.; Schaefer, K.; Zhang, T.J.; Wahr, J. Estimating 1992–2000 Average Active Layer Thickness on the Alaskan North Slope from Remotely Sensed Surface Subsidence. J. Geophys. Res. Earth Surf. 2012, 117, 2011JF002041. [Google Scholar] [CrossRef]
Chang, T.; Han, J.; Li, Z.; Wen, Y.; Hao, T.; Lu, P.; Li, R. Active Layer Thickness Retrieval over the Qinghai-Tibet Plateau from 2000 to 2020 Based on Insar-Measured Subsidence and Multi-Layer Soil Moisture. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2021, XLIII-B3-2021, 437–442. [Google Scholar] [CrossRef]
Li, Z.W.; Zhao, R.; Hu, J.; Wen, L.X.; Feng, G.C.; Zhang, Z.Y.; Wang, Q.J. InSAR Analysis of Surface Deformation over Permafrost to Estimate Active Layer Thickness Based on One-Dimensional Heat Transfer Model of Soils. Sci. Rep. 2015, 5, 15542. [Google Scholar] [CrossRef]
Daout, S.; Doin, M.-P.; Peltzer, G.; Socquet, A.; Lasserre, C. Large-Scale InSAR Monitoring of Permafrost Freeze-Thaw Cycles on the Tibetan Plateau. Geophys. Res. Lett. 2017, 44, 901–909. [Google Scholar] [CrossRef]
Geng, L.Y.; Che, T.; Wang, X.F.; Wang, H.B. Detecting Spatiotemporal Changes in Vegetation with the BFAST Model in the Qilian Mountain Region during 2000–2017. Remote Sens. 2019, 11, 103. [Google Scholar] [CrossRef]
Zhang, L.F.; Yan, H.W.; Qiu, L.S.; Cao, S.P.; He, Y.; Pang, G.J. Spatial and Temporal Analyses of Vegetation Changes at Multiple Time Scales in the Qilian Mountains. Remote Sens. 2021, 13, 5046. [Google Scholar] [CrossRef]
Ding, M.J.; Zhang, Y.L.; Liu, L.S.; Zhang, W.; Wang, Z.F.; Bai, W.Q. The Relationship between NDVI and Precipitation on the Tibetan Plateau. J. Geogr. Sci. 2007, 17, 259–268. [Google Scholar] [CrossRef]
Deng, S.F.; Yang, T.B.; Zeng, B.; Zhu, X.F.; Xu, H.J. Vegetation Cover Variation in the Qilian Mountains and Its Response to Climate Change in 2000–2011. J. Mt. Sci. 2013, 10, 1050–1062. [Google Scholar] [CrossRef]
Qin, Y.; Lei, H.M.; Yang, D.W.; Gao, B.; Wang, Y.H.; Cong, Z.T.; Fan, W.J. Long-Term Change in the Depth of Seasonally Frozen Ground and Its Ecohydrological Impacts in the Qilian Mountains, Northeastern Tibetan Plateau. J. Hydrol. 2016, 542, 204–221. [Google Scholar] [CrossRef]
Lin, P.F.; He, Z.B.; Du, J.; Chen, L.F.; Zhu, X.; Li, J. Recent Changes in Daily Climate Extremes in an Arid Mountain Region, a Case Study in Northwestern China’s Qilian Mountains. Sci. Rep. 2017, 7, 2245. [Google Scholar] [CrossRef] [PubMed]
Li, Z.X.; He, Y.Q.; Theakstone, W.H.; Wang, X.F.; Zhang, W.; Cao, W.H.; Du, J.K.; Xin, H.J.; Chang, L. Altitude Dependency of Trends of Daily Climate Extremes in Southwestern China, 1961–2008. J. Geogr. Sci. 2012, 22, 416–430. [Google Scholar] [CrossRef]
Sun, F.X.; Lyu, Y.H.; Fu, B.J.; Hu, J. Hydrological Services by Mountain Ecosystems in Qilian Mountain of China: A Review. Chin. Geogr. Sci. 2016, 26, 174–187. [Google Scholar] [CrossRef]
Zhao, L.; Wu, Q.B.; Marchenko, S.S.; Sharkhuu, N. Thermal State of Permafrost and Active Layer in Central Asia during the International Polar Year. Permafr. Periglac. Process. 2010, 21, 198–207. [Google Scholar] [CrossRef]
Gao, H.; Wang, J.; Yang, Y.; Pan, X.; Ding, Y.; Duan, Z. Permafrost Hydrology of the Qinghai-Tibet Plateau: A Review of Processes and Modeling. Front. Earth Sci. 2021, 8, 576838. [Google Scholar] [CrossRef]
Li, Y.Y.; Li, L.Y.; Chen, C.F.; Liu, Y. Correction of Global Digital Elevation Models in Forested Areas Using an Artificial Neural Network-Based Method with the Consideration of Spatial Autocorrelation. Int. J. Digit. Earth 2023, 16, 1568–1588. [Google Scholar] [CrossRef]
Zhao, B.; Huang, Y.; Zhang, C.H.; Wang, W.; Tan, K.; Du, R.L. Crustal Deformation on the Chinese Mainland during 1998–2014 Based on GPS Data. Geod. Geodyn. 2015, 6, 7–15. [Google Scholar] [CrossRef]
Yu, J.S.; Tan, K.; Zhang, C.H.; Zhao, B.; Wang, D.Z.; Li, Q. Present-Day Crustal Movement of the Chinese Mainland Based on Global Navigation Satellite System Data from 1998 to 2018. Adv. Space Res. 2019, 63, 840–856. [Google Scholar] [CrossRef]
Yi, Y.N.; Xu, X.W.; Xu, G.Y.; Gao, H.R. Rapid Mapping of Slow-Moving Landslides Using an Automated SAR Processing Platform (HyP3) and Stacking-InSAR Method. Remote Sens. 2023, 15, 1611. [Google Scholar] [CrossRef]
Zhang, B.; Chang, L.; Stein, A. Spatio-Temporal Linking of Multiple SAR Satellite Data from Medium and High Resolution Radarsat-2 Images. ISPRS J. Photogramm. Remote Sens. 2021, 176, 222–236. [Google Scholar] [CrossRef]
Goldstein, R.M.; Werner, C.L. Radar Interferogram Filtering for Geophysical Applications. Geophys. Res. Lett. 1998, 25, 4035–4038. [Google Scholar] [CrossRef]
Zhang, Y.J.; Fattahi, H.; Amelung, F. Small Baseline InSAR Time Series Analysis: Unwrapping Error Correction and Noise Reduction. Comput. Geosci. 2019, 133, 104331. [Google Scholar] [CrossRef]
Karamvasis, K.; Karathanassi, V. Performance Analysis of Open Source Time Series InSAR Methods for Deformation Monitoring over a Broader Mining Region. Remote Sens. 2020, 12, 1380. [Google Scholar] [CrossRef]
Su, C.W.; Ding, K.L.; Zhou, M.D.; Liu, M. Subway Surface Subsidence Monitoring and Prediction Model Combined with Cubic Spline Interpolation Function. Bull. Surv. Map. 2015, 102, 160–162. [Google Scholar] [CrossRef]
Ren, T.H.; Gong, W.P.; Gao, L.; Zhao, F.M.; Cheng, Z. An Interpretation Approach of Ascending–Descending SAR Data for Landslide Identification. Remote Sens. 2022, 14, 1299. [Google Scholar] [CrossRef]
Yang, Z.F.; Li, Z.W.; Zhu, J.J.; Preusse, A.; Hu, J.; Feng, G.C.; Papst, M. Time-Series 3-D Mining-Induced Large Displacement Modeling and Robust Estimation from a Single-Geometry SAR Amplitude Data Set. IEEE Trans. Geosci. Remote Sens. 2018, 56, 3600–3610. [Google Scholar] [CrossRef]
Song, Y.Z.; Wang, J.F.; Ge, Y.; Xu, C.D. An Optimal Parameters-Based Geographical Detector Model Enhances Geographic Characteristics of Explanatory Variables for Spatial Heterogeneity Analysis: Cases with Different Types of Spatial Data. GIScience Remote Sens. 2020, 57, 593–610. [Google Scholar] [CrossRef]
Zhang, B.; Chang, L.; Stein, A. A Model-Backfeed Deformation Estimation Method for Revealing 20-Year Surface Dynamics of the Groningen Gas Field Using Multi-Platform SAR Imagery. Int. J. Appl. Earth Obs. Geoinf. 2022, 111, 102847. [Google Scholar] [CrossRef]
Liu, X.Y.; Peng, X.Q.; Zhang, Y.Y.; Frauenfeld, O.W.; Wei, G.; Chen, G.Q.; Huang, Y.; Mu, C.C.; Du, J. Observed Retrogressive Thaw Slump Evolution in the Qilian Mountains. Remote Sens. 2024, 16, 2490. [Google Scholar] [CrossRef]
Zhang, Z.J.; Wang, M.M.; Wu, Z.J.; Liu, X.G. Permafrost Deformation Monitoring Along the Qinghai-Tibet Plateau Engineering Corridor Using InSAR Observations with Multi-Sensor SAR Datasets from 1997–2018. Sensors 2019, 19, 5306. [Google Scholar] [CrossRef]
Li, R.X.; Li, Z.S.; Han, J.P.; Lu, P.; Qiao, G.; Meng, X.L.; Hao, T.; Zhou, F.J. Monitoring Surface Deformation of Permafrost in Wudaoliang Region, Qinghai–Tibet Plateau with ENVISAT ASAR Data. Int. J. Appl. Earth Obs. Geoinf. 2021, 104, 102527. [Google Scholar] [CrossRef]
Deng, H.M.; Zhang, Z.J.; Wu, Y. Accelerated Permafrost Degradation in Thermokarst Landforms in Qilian Mountains from 2007 to 2020 Observed by SBAS-InSAR. Ecol. Indic. 2024, 159, 111724. [Google Scholar] [CrossRef]
Miner, K.R.; Turetsky, M.R.; Malina, E.; Bartsch, A.; Tamminen, J.; McGuire, A.D.; Fix, A.; Sweeney, C.; Elder, C.D.; Miller, C.E. Permafrost Carbon Emissions in a Changing Arctic. Nat. Rev. Earth Environ. 2022, 3, 55–67. [Google Scholar] [CrossRef]
Li, D.S.; Wen, Z.; Cheng, Q.G.; Xing, A.G.; Zhang, M.L.; Li, A.Y. Thermal Dynamics of the Permafrost Active Layer under Increased Precipitation at the Qinghai-Tibet Plateau. J. Mt. Sci. 2019, 16, 309–322. [Google Scholar] [CrossRef]
Wang, Q.F.; Jin, H.J.; Zhang, T.J.; Wu, Q.B.; Cao, B.; Peng, X.Q.; Wang, K.; Li, L.L. Active layer seasonal freeze-thaw processes and influencing factors in the alpine permafrost regions in the upper reaches of the Heihe River in Qilian Mountains. Chin. Sci. Bull. 2016, 61, 2742–2756. [Google Scholar] [CrossRef]
Hinkel, K.M.; Nicholas, J.R.J. Active Layer Thaw Rate at a Boreal Forest Site in Central Alaska, U.S.A. Arct. Alp. Res. 1995, 27, 72–80. [Google Scholar] [CrossRef]
Romanovsky, V.E.; Osterkamp, T.E. Thawing of the Active Layer on the Coastal Plain of the Alaskan Arctic. Permafrost Periglac. Process. 1997, 8, 1–22. [Google Scholar] [CrossRef]
Zhao, L.; Cheng, G.D.; Li, S.X.; Wang, X.M.; Wang, Z.L. Thawing and frezzing processes of the active layer in Wudaoliang region of Tibetan Plateau. Chin. Sci. Bull. 2000, 45, 1205–1211. [Google Scholar] [CrossRef]
Liu, S.B.; Zhao, L.; Wang, L.X.; Zhou, H.; Zou, D.F.; Sun, Z.; Xie, C.W.; Qiao, Y.P. Intra-Annual Ground Surface Deformation Detected by Site Observation, Simulation and InSAR Monitoring in Permafrost Site of Xidatan, Qinghai-Tibet Plateau. Geophys. Res. Lett. 2022, 49, e2021GL095029. [Google Scholar] [CrossRef]
Daout, S.; Dini, B.; Haeberli, W.; Doin, M.-P.; Parsons, B. Ice Loss in the Northeastern Tibetan Plateau Permafrost as Seen by 16 Yr of ESA SAR Missions. Earth Planet. Sci. Lett. 2020, 545, 116404. [Google Scholar] [CrossRef]

Figure 1. Study area: (a) location of the Qilian Mountains within the QTP, with the range’s map base showing permafrost classification; (b) DEM of the Qilian Mountains; (c) surface cover classification map of the Qilian Mountains; (d) permafrost classification map of the study area; (e) slope map of the study area; (f) terrain classification map of the study area.

Figure 2. The method flow chart used in this study.

Figure 3. Schematic of radar imaging, using ascending SAR imagery as an example. (a) Spatial relationship between LOS deformation and true surface deformation; (b) geometric relationship of LOS deformation in the horizontal plane (all arrows indicate the positive direction); (c) geometric relationship of LOS deformation in the vertical plane (all arrows indicate the positive direction); (d) geometric relationship of LOS deformation in the horizontal plane.

Figure 4. Original GNSS data. (a) Vertical direction original GNSS data (QHGC); (b) east–west direction original GNSS data (QHGC).

Figure 5. GNSS data after detrending. (a) Vertical direction GNSS data after detrending (QHGC); (b) east–west direction GNSS data after detrending (QHGC).

Figure 6. GNSS data after denoising. (a) Vertical direction GNSS data after denoising (QHGC); (b) east–west direction GNSS data after denoising (QHGC).

Figure 7. Baseline diagrams for ascending SAR data. (a) The perpendicular baseline of ascending SAR data; (b) the perpendicular baseline of descending SAR data.

Figure 8. InSAR coherence map. (a) Ascending InSAR coherence results; (b) descending InSAR coherence results. Red ellipses indicate examples of areas with high vegetation coverage, red triangles denote examples of areas with no vegetation, and red rectangles represent examples of water body areas.

Figure 9. LOS deformation rate maps. (a) Ascending SAR data LOS deformation rate results; (b) descending SAR data LOS deformation rate results.

Figure 10. Two-dimensional SBAS-InSAR results at the QHGC site. (a) Vertical SBAS-InSAR results; (b) east–west SBAS-InSAR results.

Figure 11. Comparison of two-dimensional InSAR results with GNSS data at the QHGC site. (a) Comparison of vertical InSAR and GNSS data; (b) comparison of east–west InSAR and GNSS data.

Figure 12. Accuracy verification of two-dimensional surface deformation using the methods described in this study: (a) mutual verification results of vertical ascending InSAR and GNSS data; (b) mutual verification results of interpolated vertical descending InSAR and GNSS data; (c) mutual verification results of vertical InSAR and GNSS data; (d) mutual verification results of east–west InSAR and GNSS data.

Figure 13. Vertical surface deformation time series in typical areas. (a) Geographic location map of typical feature points; (b–g) enlarged views of each feature point location; (h–m) vertical surface deformation time series for DS1-DS6, with the dynamic periodic deformation model for each feature point marked in the figure.

Figure 14. Two-dimensional deformation field. (a) Map of annual vertical deformation rates, with a histogram of annual vertical deformation rates displayed in the lower right inset; (b) seasonal deformation amplitude, with a histogram of seasonal deformation amplitudes shown in the lower right inset.

Figure 15. Q-statistic of the Geodetector. (a) Q-statistic of the interannual deformation rates; (b) Q-statistic of the seasonal deformation amplitudes.

Figure 16. Selected analysis area locations. (a) Transect line and local geographical map of the analysis area; (b,c) DEM maps of the analysis area.

Figure 17. Relationship between seasonal deformation, annual deformation rate, slope, and elevation (Transect line AB).

Figure 18. Histograms of annual vertical deformation rates and seasonal deformation amplitudes by slope aspect. (a) Histograms of annual vertical deformation rates at sunny and shaded slopes within the study area; (b) histograms of seasonal deformation amplitudes of permafrost at sunny and shaded slopes within the study area.

Figure 19. Time series of vertical surface deformation in relation to temperature changes at different locations. (a) Relationship between vertical surface deformation time series and temperature changes at point DS3; (b) relationship between vertical surface deformation time series and temperature changes at point DS4.

Figure 20. Relationship between vertical surface deformation time series and precipitation at different locations. (a) Relationship between vertical surface deformation time series and precipitation at points DS3 and DS4; (b) relationship between vertical surface deformation time series and local soil moisture at points DS3 and DS4.

Figure 21. The Five Stages of Seasonal Deformation in Permafrost at DS6 (P1: Summer Melting Process; P2: Warm Season Stabilization Process; P3: Freezing Uplift Process; P4: Winter Cooling Process; P5: Spring Warming Process).

Table 1. Information on Sentinel-1 SAR remote sensing data used for surface monitoring.

Orbit Type	Path	Polarization	Data Mode	Azimuth Angle (°)	Incidence Angle (°)
Ascending	26	VV	IW	100.2	38
Descending	106	VV	IW	−100.4	36

Table 2. Basic information on influencing factors.

Data	Grid Size (m)	Time (Year)	Data Source
Fractional Vegetation Cover (FVC)	250	2017–2023	https://data.tpdc.ac.cn (accessed on 15 March 2024)
Mean Annual Precipitation (MAP)	1000	2017–2023	CRU & WorldClim
Mean Annual Atmospheric Temperature (MAAT)	1000	2017–2023	CRU & WorldClim
Mean Annual Atmospheric Temperature_MIN (MAATMN)	1000	2017–2023	CRU & WorldClim
Mean Annual Atmospheric Temperature_MAX (MAATMX)	1000	2017–2023	CRU & WorldClim
Land Surface Temperature (LST)	1000	2017–2023	MODIS
Evapotranspiration (ET)	500	2017–2023	MODIS
Potential Evapotranspiration (PET)	500	2017–2023	MODIS
Enhanced Vegetation Index (EVI)	1000	2017–2023	MODIS
Normalized Difference Vegetation Index (NDVI)	1000	2017–2023	MODIS
Gross Primary Production (GPP)	500	2017–2023	MODIS
Net Primary Production (NPP)	500	2017–2023	MODIS
Leaf Area Index (LAI)	500	2017–2023	MODIS
Active Lager Thickness (ALT)	1000	2017–2020	MODIS
Maximum Thickness of Seasonally Frozen Ground (MTSFG)	1000	2017–2020	https://data.tpdc.ac.cn (accessed on 15 March 2024)
Surface Soil Moisture (SMsurf)	6000	2017–2020	GLEAM
Rootzone Soil Moisture (SMroot)	6000	2017–2020	GLEAM
Permafrost Zoning Index (PZI)	250	The last 50 years	https://data.tpdc.ac.cn (accessed on 15 March 2024)
Human Activity Affection (HAA)	30	2021	https://data.tpdc.ac.cn (accessed on 15 March 2024)
Permafrost Ground-ice content [2 m–3 m] (PGC23)	1000	The last 28 years	https://data.tpdc.ac.cn (accessed on 15 March 2024)
Permafrost Ground-ice content [3 m–5 m] (PGC35)	1000	The last 28 years	https://data.tpdc.ac.cn (accessed on 15 March 2024)
Permafrost Ground-ice content [5 m–10 m] (PGC510)	1000	The last 28 years	https://data.tpdc.ac.cn (accessed on 15 March 2024)
Short Water Equivalent (SWE)	3655	2017–2023	https://data.tpdc.ac.cn (accessed on 15 March 2024)
Snow Depth (SD)	3655	2017–2023	https://data.tpdc.ac.cn (accessed on 15 March 2024)
Soil Freezing Duration (SFD)	9000	2017–2022	ERA5-LAND
Land Cover	30	2020	GlobeLand30
Basic Geomorphologic type	1000	2009	https://data.tpdc.ac.cn (accessed on 15 March 2024)
Slope	30	2022	DEM
Aspect	30	2022	DEM

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

Share and Cite

MDPI and ACS Style

Xue, Z.; Zhao, S.; Zhang, B. Study on Soil Freeze–Thaw and Surface Deformation Patterns in the Qilian Mountains Alpine Permafrost Region Using SBAS-InSAR Technique. Remote Sens. 2024, 16, 4595. https://doi.org/10.3390/rs16234595

AMA Style

Xue Z, Zhao S, Zhang B. Study on Soil Freeze–Thaw and Surface Deformation Patterns in the Qilian Mountains Alpine Permafrost Region Using SBAS-InSAR Technique. Remote Sensing. 2024; 16(23):4595. https://doi.org/10.3390/rs16234595

Chicago/Turabian Style

Xue, Zelong, Shangmin Zhao, and Bin Zhang. 2024. "Study on Soil Freeze–Thaw and Surface Deformation Patterns in the Qilian Mountains Alpine Permafrost Region Using SBAS-InSAR Technique" Remote Sensing 16, no. 23: 4595. https://doi.org/10.3390/rs16234595

APA Style

Xue, Z., Zhao, S., & Zhang, B. (2024). Study on Soil Freeze–Thaw and Surface Deformation Patterns in the Qilian Mountains Alpine Permafrost Region Using SBAS-InSAR Technique. Remote Sensing, 16(23), 4595. https://doi.org/10.3390/rs16234595

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Menu

Study on Soil Freeze–Thaw and Surface Deformation Patterns in the Qilian Mountains Alpine Permafrost Region Using SBAS-InSAR Technique

Abstract

1. Introduction

2. Study Area and Datasets

2.1. Study Area

2.2. Data

2.2.1. SAR Data

2.2.2. GLO-30 DEM

2.2.3. Underlying Surface Data and Permafrost-Related Data

2.2.4. GNSS Data

3. Method

3.1. SBAS-InSAR Processing Workflow

3.2. Establishment of a Two-Dimensional Surface Deformation Field

3.2.1. Time Series Registration of Ascending and Descending Results Based on Cubic Spline Interpolation

3.2.2. Calculation of Two-Dimensional Surface Deformation Fields

3.3. GNSS Processing

3.3.1. Detrending GNSS Data

3.3.2. Denoising GNSS Data

3.4. Development of a Permafrost Dynamic Periodic Deformation Model

3.5. Application of Q-Statistics to Analyze Interannual Deformation Rate and Seasonal Deformation Influencing Factors

4. Results

4.1. SBAS-InSAR Results

4.2. Establishment of the Two-Dimensional Surface Deformation Field and Cross-Validation with GNSS Data

4.3. Analysis of Vertical Surface Deformation Time Series in Typical Regions

4.4. Spatial Distribution of Annual Vertical Surface Deformation Rates

4.5. Spatial Distribution of Seasonal Deformation in Permafrost

4.6. Analysis of Interannual Deformation Rate and the Influencing Factors of Seasonal Deformation

4.6.1. Factors Influencing Interannual Deformation Rate

4.6.2. Factors Influencing the Amplitude of Seasonal Deformation

4.7. Relationship Between Permafrost Freeze–Thaw-Induced Surface Deformation and Topographical Factors

4.7.1. The Impact of Altitude on Permafrost Deformation

4.7.2. The Impact of Slope on Permafrost Deformation

4.7.3. The Impact of Aspect on Permafrost Deformation

4.8. The Relationship Between Permafrost Freeze–Thaw-Induced Surface Deformation and Climatic Factors

4.8.1. Response of Permafrost Deformation to Temperature

4.8.2. Response of Permafrost Deformation to Soil Moisture and Precipitation

5. Discussion

5.1. Monitoring Surface Deformation Using SBAS-InSAR Technology

5.2. Factors Affecting Permafrost Deformation

5.3. The Five Stages of Seasonal Deformation in Permafrost

5.3.1. Warming and Melting Phase

5.3.2. Cooling and Freezing Phase

5.4. Applications of Two-Dimensional Surface Deformation Fields Under Spatiotemporal Interpolation

5.5. Applications of Dynamic Periodic Deformation Model

6. Conclusions

Author Contributions

Funding

Data Availability Statement

Conflicts of Interest

References

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.