Response of Alpine Timberline to Permafrost Degradation on Changbai Mountain

Abstract

In the permafrost zone, the vegetation growth condition is closely related to the permafrost occurrence state. Changbai Mountain is the highest peak in Northeast China and is also a permafrost distribution area, where the vegetation shows an obvious vertical distribution pattern, and the alpine timberline has a clear boundary. The intersecting zone of alpine timberline is an ecologically fragile area that can be used as an external signal amplifier and is an important site for monitoring climate change. In this study, the surface frost number and alpine timberline in the Changbai Mountain area were analyzed through combining ground and remote-sensing data, using the freezing number model and support vector machine method. The results show that the distribution characteristics of alpine timberline and permafrost at elevation are correlated, there is a response mechanism of alpine timberline to the degradation of permafrost, and the upward migration rate of alpine timberline has increased in the last 20 years. The continuous degradation of permafrost will change the environment of vegetation growth, which, in turn, will affect the global carbon cycle process. Focusing on the state of permafrost will help us to understand climate change in depth, and we can respond to a series of impacts caused by ecological changes in cold regions in advance.

1. Introduction

As per the Sixth Assessment Report released by the Intergovernmental Panel on Climate Change (IPCC), the global surface temperature is projected to increase by a minimum of 1 °C above the 1850–1900 average during the period of 2011–2020 [1], with a more pronounced temperature increase expected at higher altitudes and latitudes. Over the past approximately three decades, the mean annual temperature in Northeast China has exhibited a warming trend of 0.6 °C per decade [2]. Climate change and anthropogenic activities constitute significant factors driving permafrost alterations, which, in turn, significantly influence energy conversion within ecosystems, as well as hydrological and geomorphological transformations [3].

There is an interaction between vegetation growth and permafrost changes [4]. On one hand, the development of vegetation helps to intercept solar radiation, thus hindering the melting of permafrost, and the demise of vegetation will also lead to permafrost degradation [5]; on the other hand, permafrost also affects vegetation as, in the short term, the degradation of permafrost will affect the hydrothermal environment of the soil layer in the alpine zone, and the water and organic matter released by permafrost thawing will promote the development of vegetation [6]. However, in the long run, the large-scale degradation of permafrost will trigger a series of primary or secondary successions of vegetation, leading to vegetation degradation [7]. The formation of vegetation is the result of long-term climate change and the integrated effect of multiple factors; as an important part of terrestrial ecosystems, vegetation regulates the global energy balance and water cycle [8]. In the study of global climate change, vegetation plays the role of “indicator” [9,10,11].

As the upper boundary of tree and forest growth, the alpine timberline is the transition zone from forest to alpine meadow [12,13], and it is also an ecologically fragile zone [14,15]. The ecological transition zone is sensitive to climate change, and it can be used as an amplifier of external signals and an early warning zone for global climate change, so the alpine timberline has a very important role in the detection of climate change.

The mountain effect is the main cause of permafrost development [16], and permafrost is widely distributed in the Changbai Mountain region due to topographic and climatic factors. There is already a complex connection between vegetation and permafrost, and there is very limited research on the relationship between vegetation and permafrost. In addition, the distribution of permafrost is also affected by spatial heterogeneity, and the relationship between permafrost and vegetation growth cannot be generalized. Although there have been some research results on the changes in alpine timberlines [17,18,19], basically, the effect of permafrost on the changes in alpine timberlines has not been considered. The mechanism of the effect of permafrost on vegetation remains a challenge for ecologists in cold regions.

In this study, the research was focused on Changbai Mountain, a region characterized by permafrost, situated in Northeast China. Against the backdrop of global warming, the study aimed to analyze the changes in the state of permafrost and the trends in the response of the alpine timberline in Changbai Mountain over the past 20 years by decoding both ground-based and multi-source remote-sensing data. Compared with traditional technical means, remote-sensing technology can quickly and accurately obtain results over long time series and a large study area.

2. Data and Methods

2.1. Study Area

Changbai Mountain is the highest mountain in Northeast China (Figure 1), and permafrost is mainly distributed in the high-altitude area dominated by the Changbai Mountain range. The highest elevation of Changbai Mountain reaches 2749 m, which can effectively block oceanic airflow and is a natural barrier in Northeast China [20]. Changbai Mountain is a dormant ring volcano, with the center of Tianchi (a lake in the center of a crater volcano) at longitude 128°3′18″ and latitude 42°0′28″, and the region belongs to the temperate continental monsoon alpine climate, with a long spring and winter, and a relatively short summer and fall. The temperature decreases rapidly with the increase in elevation and, at the tops of the mountains at the higher elevations, the average annual temperatures range from −7° to 3°. The precipitation is mostly concentrated in the months from May to September, which account for more than 75%. The distribution of vegetation in the Changbai Mountain region is jointly influenced by a variety of factors such as topography, climate, soil components, and permafrost distribution, and the dominant tree species are Betula ermanii forests and Changbai Larch [21,22].

2.2. Data Description and Data Sources

The digital elevation model (DEM) was selected from NASA’s (National Aeronautics and Space Administration) SRTM (Shuttle Radar Topography Mission) with a spatial resolution of 30 m (data source: https://cmr.earthdata.nasa.gov/). Topographic factors such as the slope, aspect, contour and summit point of the study area were extracted using ArcGIS 10.8 software for analyzing timberlines.

The normalized difference vegetation index (NDVI) was selected from NASA’s MOD13Q1 dataset for 2000–2021 (data source: https://ladsweb.modaps.eosdis.nasa.gov/). The spatial resolution was 250 m, and the temporal resolution was 16 d. The downloaded MODIS NDVI data were spliced and transposed using the MRT tool, and then the maximum value compositing (MVC) method was used in ArcGIS 10.8 to obtain monthly NDVI data. The average NDVI was calculated for the growing season (June–September) each year [23].

The land surface temperature (LST) was selected from NASA’s MOD11A1 dataset, which is available from the International Scientific Data Service Platform (http://datamirror.csdb.cn) with a spatial resolution of 1 km and a temporal resolution of 1 d [24,25].

Air temperature and precipitation data released by the China Meteorological Science Data Center (http://data.cma.cn/) were used as meteorological data. The images were selected from June to September, considering that snow accumulation at temperatures below zero would cause great interference in the identification of timberlines.

In this paper, Landsat series remote-sensing images (Path 116, Row 31) were selected to extract alpine timberlines, using the Landsat dataset from USGS (United States Geological Survey) (https://earthexplorer.usgs.gov/). Cloudless, snow-covered 30 m resolution images were selected for June-October 2000–2021, and the image types and dates are shown in

Table 1. Information on selected remote-sensing image data.

Landsat Sensor Type	Dates
Landsat5 TM	12 September 2000
Landsat7 ETM	4 October 2005
Landsat7 ETM	3 September 2011
Landsat7 ETM	14 September 2015
Landsat8 OLI/TIRS	18 June 2021

. Since the landsat7 satellite lost the image stripes due to SLC damage after 31 May 2003, the landsat gapfill tool in ENVI5.6 was used to repair the stripes on the landsat7 image before processing, and then the data were uniformly subjected to operations such as radiometric calibration and atmospheric correction in order to obtain a complete and applicable image. Finally, cropping was performed to obtain the distribution data from remote-sensing images in the study area.

2.3. Analysis Methods

2.3.1. Significance Analysis of Vegetation Trends

Trend significance tests were performed using the Mann–Kendall (MK) test and the Theil–Sen median method (Sen’s slope estimation) [26,27]. The MK test is a nonparametric test that does not require a priori assumptions about the distribution of the data [28,29,30], and the data need not follow a normal distribution. The combination of the MK test and Sen’s slope estimation was used to analyze the trend and significance of NDVI changes on Changbai Mountain from 2001 to 2020.

$$\mathrm{Z}=\left\{\begin{array}{cc}\frac{\mathrm{S}-1}{\sqrt{\mathrm{var}\left(\mathrm{S}\right)}},\hfill & \mathrm{S}>0\hfill \\ 0,\hfill & \mathrm{S}=0\hfill \\ \frac{\mathrm{S}+1}{\sqrt{\mathrm{var}\left(\mathrm{S}\right)}},\hfill & \mathrm{S}<0\hfill \end{array}\right.$$

(1)

$$\mathrm{S}={\sum }_{\mathrm{i}=1}^{\mathrm{n}-1}{\sum }_{\mathrm{j}=\mathrm{i}+1}^{\mathrm{n}} \mathrm{sgn}\left({\mathrm{NDVI}}_{\mathrm{j}}-{\mathrm{NDVI}}_{\mathrm{i}}\right)$$

(2)

$$\mathrm{sgn}\left({\mathrm{NDVI}}_{\mathrm{j}}-{\mathrm{NDVI}}_{\mathrm{i}}\right)\left\{\begin{array}{cc}1,\hfill & {\mathrm{NDVI}}_{\mathrm{j}}-{\mathrm{NDVI}}_{\mathrm{i}}>0\hfill \\ 0,\hfill & {\mathrm{NDVI}}_{\mathrm{j}}-{\mathrm{NDVI}}_{\mathrm{i}}=0\hfill \\ -1,\hfill & {\mathrm{NDVI}}_{\mathrm{j}}-{\mathrm{NDVI}}_{\mathrm{i}}<0\hfill \end{array}\right.$$

(3)

$$\mathrm{var}\left(\mathrm{S}\right)=\frac{\mathrm{n}\left(\mathrm{n}-1\right)\left(2\mathrm{n}+5\right)}{18}$$

(4)

where Z is the trend change test value, S is the MK statistic, n is the length of the time series, here 20 years. NDVI_i and NDVI_j denote the NDVI for the growing seasons of year i and year j.

Sen’s slope estimator, the β value determines whether vegetation is undergoing an upward or downward trend, where β < 0 indicates a downward trend, and vice versa for an upward trend. The NDVI significance classification criteria are shown in

Table 2. Significance classification of the NDVI based on the MK test.

β	|Z|	Significance
β < 0	|Z| > 2.58	highly significant decrease
	1.96 < |Z| ≤ 2.58	significant decrease
	1.65 < |Z| ≤ 1.96	low significant decrease
	|Z| ≤ 1.65	no significant decrease
β > 0	|Z| ≤ 1.65	no significant increase
	1.65 < |Z| ≤ 1.96	low significant increase
	1.96 < |Z| ≤ 2.58	significant increase
	|Z| > 2.58	highly significant increase

Set the significance level α = 0.05 and accept the original hypothesis when |Z| ≤ Z1 − α/2. Otherwise, the original hypothesis is rejected, indicating a significant change in the trend.

.

$$\mathsf{\beta }=\mathrm{Median}\left\{\frac{{\mathrm{NDVI}}_{\mathrm{j}}-{\mathrm{NDVI}}_{\mathrm{i}}}{\mathrm{j}-\mathrm{i}}\right\},\forall \mathrm{j}>i$$

(5)

2.3.2. Extraction of Alpine Timberlines Based on the Support Vector Machine Method

Support vector machine (SVM) is a machine learning model based on statistical learning theory; the main idea is to map the input data to a high-dimensional space through the kernel function to realize linear regression. SVM is robust to outliers and has a high prediction accuracy, and it is widely used in the classification of features [31,32,33]. Firstly, preprocessing work such as atmospheric correction and radiometric calibration was carried out on Landsat data, and then forests, alpine tundra, Tianchi water, and bare rocks were selected as samples for classification via visual interpretation method in ENVI5.6 software combined with the available information on the study area, respectively. The results showed that the selected samples were more different, and the “separability” of the feature samples was high. Then, the classification results were verified with high-resolution remote-sensing images, and the misclassified or omitted pixels were corrected. Finally, in ArcGIS 10.8 software, based on the DEM, the summit points and contour lines were extracted [34], and the demarcation line between the forest and the alpine tundra was extracted as the alpine timberline using vector topology analysis and superposition analysis [35].

2.3.3. Surface Frost Number Model

The surface frost number model, as proposed by Nelson [36], utilizes the surface frost number F_n formula, represented in Equation (8), to analyze, simulate, and predict permafrost distribution [37,38,39,40]. Zhang and colleagues [41,42] provided a more comprehensive delineation for Northeast China using this model. The surface freezing index (DDF) is defined as the cumulative absolute values of LST below 0 °C from July of the first year to June of the second year, while the surface melting index (DDT) is calculated as the cumulative absolute values of LST above 0 °C from January to December.

$$\mathrm{DDF}={\sum }_{182}^{\mathrm{n}}\mid {\overline{\mathrm{LST}}}_{\mathrm{t}}\mid +{\sum }_1^{181}\mid {\overline{\mathrm{LST}}}_{\mathrm{t}+1}\mid \left(\overline{\mathrm{LST}}<0 °\mathrm{C}\right)$$

(6)

$$\mathrm{DDT}={\sum }_1^{\mathrm{n}}\overline{\mathrm{LST}}\left(\overline{\mathrm{LST}}>0 °\mathrm{C}, \mathrm{n}\le 365\right)$$

(7)

$${\mathrm{F}}_{\mathrm{n}}=\frac{\sqrt{\mathrm{DDF}}}{\sqrt{\mathrm{DDF}}+\sqrt{\mathrm{DDT}}}$$

(8)

where $\overline{\mathrm{LST}}$ is the mean value of MOD11A1 ground temperature data, t denotes the year, and n is the number of days. F_n is the surface frost number.

Vegetation in the Changbai Mountain region shows obvious vertical distribution characteristics, and the alpine timberline has obvious boundaries, and vegetation regulates the atmosphere, soil, and hydrothermal state of the ecosystem. Therefore, vegetation is an important factor influencing surface temperature [43].

$${\mathrm{F}}_{\mathrm{nc}}=\left({\mathrm{E}}_{\mathrm{t}}\times \mathrm{NDVI}+1\right){\mathrm{F}}_{\mathrm{n}}$$

(9)

where F_nc is the surface frost number under the influence of the vegetation factor, and E_t is the vegetation factor taken as 0.306, which is derived using the later calculation.

3. Results

3.1. Vegetation Change

Climate is a crucial environmental factor that significantly influences the growth of vegetation during the growing season. The trend of climate interannual variability in the study area was calculated via counting the annual mean values for the NDVI in the growing season in the Changbai Mountain alpine timberline elevation 1660–2240 m area, and combining them with the mean temperature and mean precipitation in the growing season at the four nearest meteorological stations in the Changbai Mountain area, namely, Changbai, Donggang, Erdao, and Linjiang. Subsequently, a linear model was applied to fit the interannual climate variability in the study area.

The monthly mean air temperature and precipitation data from the weather stations from 2000 to 2020 showed that Changbai Mountain had a significant rain–heat synoptic phenomenon (Figure 2a), and both temperature and precipitation reached their highest values in July, which were 20.2 °C and 148.8 mm, respectively. We plotted and linearly fitted the interannual relationships of the NDVI with temperature and precipitation during the growing season (Figure 2b,c). The results show that air temperature and precipitation showed an increasing trend between 2000 and 2021, with an average rate of change of 0.00707 °C/year and 0.277 mm/year, and that air temperature and precipitation reached a minimum in 2009 and a maximum in 2010. Over the course of 22 years, the NDVI has shown good consistency with air temperature and precipitation trends during the growing season. Vegetation is sensitive to changes in air temperature, and the overall trend in both changes is similar. The Pearson correlation coefficient between the NDVI and air temperature is 0.39 (p = 0.072 < 0.1), indicating that there is a strong correlation between vegetation and air temperature; the peaks of precipitation correspond to the troughs of the NDVI, which indicates that there is a certain lag in the change in vegetation and precipitation, and there is about a one-year lag period, and the Pearson correlation coefficient of the NDVI and precipitation is 0.076 (p = 0.076 < 0.1). The correlation coefficient of the NDVI and precipitation was 0.076 (p = 0.74 > 0.1), indicating that the correlation between vegetation and precipitation was weak.

3.2. Coupled Permafrost and Vegetation Changes

According to Equation (8), the distribution of surface frost number F_n from 2001 to 2020 is shown in Figure 3; F_n reflects the status of permafrost; and the larger the value of F_n, the more stable the permafrost. It can be seen that the overall stability of permafrost decreases from high-altitude to low-altitude areas with Tianchi as the center, and the permafrost is more stable in the northern part of Changbai Mountain and the northeastern part, while, in the southeastern part, the permafrost is degraded to different degrees in some years.

The stability of permafrost is significantly correlated with vegetation cover. The interannual relationship between F_n and the mean growing season NDVI in the study area from 2001 to 2020 was plotted, as shown in Figure 4a. Combining the interannual variation in the NDVI and precipitation in the growing season (Figure 2c), it can be seen that, with insignificant changes in mean annual precipitation, the NDVI shows an upward trend with global warming, while the surface frost number F_n in the Changbai Mountain region shows a downward trend. Temperature and precipitation are crucial influencing factors in the development of permafrost over multiple years. To eliminate the potential errors caused by strongly correlated factors under extreme conditions, several sets of data with maximum NDVI, temperature, and precipitation are excluded. Subsequently, a linear regression using the least-squares method is applied to fit the remaining years’ data for the NDVI and F_n (Figure 4b). The results show that F_n and the NDVI were significantly negatively correlated, and the correlation coefficient between F_n and the NDVI was −0.61 (p < 0.05).

F_nc is the surface frost number after considering the vegetation factor, which can accurately simulate the distribution and stable state of permafrost, and a larger value of F_nc indicates a more stable permafrost. From Figure 4b, it can be seen that when the NDVI increases by 0.1, F_n decreases by 0.0306, from which it can be determined that the vegetation factor E_t is 0.306. All other conditions being the same, the surface frost number model can reflect the distribution of permafrost very well when the ground surface is bare soil, but when the ground surface is covered by vegetation, the canopy of the vegetation will block the heat from entering into the ground. At this time, the surface frost number in the understory will be higher than the vegetation canopy. Therefore, the fraction of heat absorbed by the vegetation canopy needs to be taken into account when calculating F_nc. Substituting the 20-year annual average F_n and the growing season NDVI into Equation (9), the F_nc of the study area was calculated. Trimming away rocks above 2040 m in elevation and the Tianchi area, the remaining portion represents the distribution of permafrost in the Changbai Mountain region, as depicted in Figure 4c. The percentage and area of each type of surface thermal state in the study area are divided according to F_nc, and the results are displayed in

Table 3. Percentage and area of each type of surface thermal state on Changbai Mountain.

Type of Permafrost	Surface Thermal State	Fnc Value	Percentage	Area (/km2)
Continuous permafrost	Extremely stable surface thermal state	≥0.66	33.6%	425.47
Discontinuous permafrost	Stabilized surface thermal state	0.55–0.66	65.25%	826.30
	Sub-stabilized surface thermal state	0.50–0.55	1.15%	14.50
Aggregate	-	-	100%	1266.27

.

3.3. Spatial Distribution of Vegetation on Changbai Mountain and Trends in Change

The vegetation in the study area shows significant vertical distribution characteristics. The average values of the NDVI growing season from 2001 to 2020 in the study area are shown in Figure 5a. The NDVI can be divided into four levels of irregular concentric circles centered on Tianchi on Changbai Mountain, and the NDVI gradually increases with decreasing altitude. Firstly, the first level of NDVI is below 0.3, which is the water and alpine rock layer centered on Tianchi; the second level of NDVI is around 0.3–0.6, where the alpine tundra is mainly distributed at higher elevations; the third level of NDVI is around 0.6–0.9, at the periphery of the alpine tundra, where the alpine tundra and forest intertwine; and the fourth level of NDVI is above 0.9, which is mainly the coniferous forest zone with denser vegetation cover.

The overall NDVI in the study area is mainly increasing. The significance test for the trend in NDVI change in the part of Changbai Mountain above 2040 m altitude is shown in Figure 5b. Based on the results of the MK test, it is evident that significant and highly significant increases are concentrated in the southeast of the alpine timberline in the year 2021. Conversely, areas with a non-significant decrease in the NDVI are concentrated in the northwest region. In addition, the percentage of significantly decreasing regions is very small, and there is no region of highly significant decrease in the region above 2040 m. Combined with the regions where the 2021 alpine timberline is distributed, the southern part of Changbai Mountain, which is the sunny slope of Changbai Mountain, the overall trend in NDVI in the alpine timberline region is increasing. Meanwhile, on the northern part of Changbai Mountain, which is the dorsal slope of Changbai Mountain, the overall trend in NDVI in the alpine timberline region is decreasing.

3.4. Changbai Mountain Alpine Timberline Changes

Changbai Mountain’s alpine timberline change is related to the distribution of permafrost, and the discontinuous-type permafrost distribution area, and the alpine timberline change is significant. The Changbai Mountain alpine timberline intertwining zone transition is obvious from the forest directly transitioning to tundra vegetation. The alpine timberline in the Changbai Mountain region was extracted via selecting the demarcation line between forest and alpine tundra, and the alpine timberline on Changbai Mountain between 2000 and 2021 was extracted using the support vector machine method based on ENVI5.6, and the results of the alpine timberline extraction were superimposed onto the Landsat RGB true-color imagery (Figure 6a). The results show a significant upward trend in the southwest and, combined with Figure 4c, it can be seen that the southwest region of Changbai Mountain is a discontinuous type of permafrost, and the alpine timberline has a significant upward trend in this region. The northwestern region is continuous permafrost, and the change in alpine timberline in this part is not obvious.

Geographic location is an important factor contributing to the variability of different aspects of the alpine timberline on Changbai Mountain. The southeastern part of Changbai Mountain is close to the sea, and it is easy to form topographic rain under the influence of oceanic air currents, so the rain and heat conditions on the southeastern part of the mountain are significantly better than those in other regions. The slope aspect of the terrain factor was extracted from the DEM, and the mean and maximum values of the alpine timberlines (Figure 6d,e) were counted for each slope aspect. The results showed that the western and southeastern slopes suffered the greatest degree of natural disasters, and the mean values for alpine timberlines were significantly lower than those of other slope aspects. In terms of time, the mean value for the alpine timberline on the south and southwest slopes showed an increasing trend year by year during the 22-year period, and the mean value for the alpine timberline on the southeast aspect was degraded to a certain extent and then recovered slowly to the original level after 2000. The maximum value of the Changbai Mountain alpine timberline on the eastern slope appeared in 2000; the maximum value of the alpine timberline on the northern and northeastern slopes appeared in 2005; the maximum value on the southwestern slope appeared in 2021; and the maximum value of all other aspects appeared in 2015.

There is a strong correlation between the spatial distribution characteristics of alpine timberline and elevation. By analyzing the top view of the alpine timberline from south to north (Figure 6b) and from north to south (Figure 6c) in the year 2021 in this direction, it is concluded that the distribution of the alpine timberline ranges from 1650 to 2040 m. Alpine timberlines at different elevations are uniformly distributed around a certain elevation, among which, the alpine timberlines in the north and the south of Changbai Mountain are located around the 2040 m contour line, and those in the northwestern part of Changbai Mountain are located around the 1650 m contour line.

The distribution frequency of alpine timberlines at elevation approximates a Gaussian distribution. The elevation frequency statistics of alpine timberlines on Changbai Mountain in each year are shown in Figure 6f. The results show that the alpine timberlines in the study area were mainly concentrated at the elevation of 1880–2080 m, and the proportion in this interval was as high as more than 75%; in the part below 2050 m, the proportion of alpine timberlines increased steadily with the elevation; however, in the part above 2050 m, the distribution density of alpine timberlines decreased rapidly with further elevation, and the high-altitude climate severely limited the growth of vegetation.

The elevation information for the alpine timberline from 2000 to 2021 is shown in

Table 4. Changbai Mountain alpine timberline elevation information.

Year	Elevation Range	Average Elevation	Standard Deviation
2000	1661–2185	1939.02	101.12
2005	1663–2180	1941.99	102.31
2011	1662–2198	1944.67	100.92
2015	1665–2241	1948.70	99.92
2021	1661–2179	1955.27	99.74

, from which it can be seen that the elevation of the distribution range of the alpine timberline in the Changbai Mountain region is 1661–2241 m, and the average elevation of the alpine timberline is in the range of 1939–1955 m, of which the elevation span is larger in 2005, with a standard deviation of 102.31 m. In the last 22 years, the years of the maximal value of the elevation of the alpine timberline were in the order 2015 > 2011 > 2000 > 2005 > 2021, with the maximal value of 2241 m reached in 2015. The maximum value of the alpine timberline can reflect the ability of trees to adapt to extreme altitude environments, which is consistent with an increasing trend in time from year to year. However, with the gradual increase in altitude, the survival environment of vegetation will be more severe, and the vegetation at high altitude will face more complex challenges. Therefore, the maximum value of alpine timberline is unstable and, with the passage of time, the vegetation at the highest altitude may be destroyed once it exceeds the limit of tolerance of tree species. The alpine timberline maximum will then be replaced by vegetation at lower elevations in subsequent years. A warming climate promotes migration of Changbai Mountain alpine timberline to higher elevations. From the statistical results in Table 4, it can be seen that the average elevation of the alpine timberline has risen by about 16 m during the last 22 years, with the rate of upward movement of the timberline being 0.73 m/year.

4. Discussion

4.1. Stability Analysis of Permafrost in the Changbai Mountain

The distribution and characteristics of permafrost are influenced by both geomorphology and vegetation [44,45]. Changbai Mountain is a circular volcanic terrain, so its different aspect steepness varies, and there are large differences in elevation, temperature, precipitation, and soil type at different locations. Environmental geology affects the vegetation distribution in the Changbai Mountain region mainly through soil components and volcanic activities. Several volcanic eruptions have been recorded in the history of Changbai Mountain [46], and some scholars have suggested that volcanic eruptions and sediment erosion in Changbai Mountain have caused differences in vegetation in different directions [47]. The northern part of Changbai Mountain is steep, with a lesser degree of vegetation destruction, and has a complete vertical distribution zone of vegetation; the eastern part is gentle, with the most serious destruction, almost destroying the original distribution of vegetation, and the dominant tree species in the eastern part, larch, is considered to be the second-generation forest vegetation after the volcanic activity [48]; the western part of Changbai Mountain has a lesser degree of destruction compared to the eastern part.

The spatial distribution of surface frost numbers calculated after considering the vegetation factor (Figure 4c) can reflect the spatial distribution status of permafrost and the stability of permafrost [49]. The area of continuous permafrost is about 425.47 km², which is concentrated in the northwestern part of Changbai Mountain, and this area is mainly characterized by leeward slopes, which are less affected by windstorms [50]. In addition, the northwestern part has higher elevation, well-preserved vegetation, and shady slopes. The vegetation canopy can effectively block solar radiation and reduce the heat directly into the ground, and the root system of the vegetation has the effect of water retention to prevent the water from penetrating downward and evaporating. The dense vegetation and high altitude create favorable conditions for the development of permafrost.

4.2. The Distribution Characteristics of Changbai Mountain Permafrost and Alpine Timberline at Elevation Are Correlated

The surface frost number and vegetation cover in the Changbai Mountain region show a negative correlation at elevation. Permafrost stability increases with elevation, but an environment with too low a temperature will no longer be suitable for vegetation growth.

Topographic factors control the pattern of vegetation distribution at small scales through elevation, slope, and aspect [51]. Vegetation in Changbai Mountain shows obvious vertical distribution characteristics [52], with the red pine mixed coniferous forest belt mainly from 700 to 1100 m above sea level, the coniferous forest belt from 1100 to 1800 m above sea level, the Yuehua birch forest belt from 1800 to 2100 m above sea level, and the alpine tundra belt above 2100 m above sea level, and the alpine timberline has a clear boundary. From Figure 4c, it can be seen that the degree of stability of permafrost occurrence decreases from the center of Tianchi to the surroundings with the decrease in elevation, while the vegetation in the study area has obvious stratification phenomena generally, and the NDVI gradually increases with the decrease in elevation. With the warming of the climate, the permafrost is degraded to higher elevations and, combined with the frequency distribution of alpine timberlines at elevation (Figure 6f), it can be seen that the percentage of alpine timberline in the higher-elevation region has a tendency to increase from year to year.

4.3. Response of Alpine Timberlines on Changbai Mountain to Permafrost Degradation

The occurrence state of permafrost shows a degradation trend under the influence of external factors [53], and vegetation has a hindering effect on the degradation of permafrost. Combined with Figure 3, it can be seen that the stability of permafrost changes when external factors change. The degradation of permafrost in the study area has its own regularity, which is specifically manifested in the direction of degradation from low elevation to high elevation and from southeast to northwest. In the northwest direction, stable permafrost is characterized by high elevation and well-preserved vegetation while, in the southeast direction, the terrain transition is gentle, and the vegetation cover is sparse. This indicates that unstable permafrost is more sensitive to climate change and will be affected first when the climate warms, while the dense vegetation in the northwest region can hinder the degradation of permafrost.

Due to the geographic location, the southwestern region of Changbai Mountain is a sunny slope with better sunshine conditions than the northern part, and the surface freezing number is lower in the southwestern region. The alpine timberline on the sunny slopes of the Changbai Mountain is steadily migrating to higher elevations. The results of the MK trend significance test also indicate that the vegetation near the alpine timberline in the southwest-to-southeast region in 2021 shows an increasing trend of different degrees. There was no significant change in the alpine timberline in the area of continuous permafrost at the higher elevations of the northern Changbai Mountain on the back shady slopes.

The degradation of permafrost promotes the upward shift of the alpine timberline. Climate is an important driver of permafrost degradation [7,54,55,56]. Under the background of global warming, the low-temperature limitation in high latitudes is gradually lifted, while the vegetation shows a certain trend of growth [57]. When permafrost thaws, it releases water and organic matter. These changes in hydrothermal conditions are rapidly reflected in vegetation growth and development [58,59], ultimately affecting shifts in alpine timberlines. The alpine timberline has risen by 80 m from 1850 to 2010, with a rate of change of 0.5 m/year [60]. From the statistics in Table 4, it can be seen that the average height of the alpine timberline has risen by about 16 m during the past 22 years, and the rate of upward movement of the timberline has increased from 0.5 m/year to 0.73 m/year.

5. Conclusions

The continuous degradation of permafrost will change the vegetation growth environment and affect the global carbon cycle process. In this study, the distribution of permafrost and alpine timberline on Changbai Mountain was simulated through calculations using multi-source remote-sensing data and a surface frost number model, based on the characteristics of permafrost highly sensitive to climate change, and taking into account the factors of topography, geographic location, climate, and natural disasters. The results of the study show that the distribution characteristics of permafrost and the alpine timberline on Changbai Mountain in terms of elevation are correlated. The relationship between vegetation and permafrost is very subtle. On one hand, vegetation has a hindering effect on the degradation of permafrost; on the other hand, the degradation of Changbai Mountain permafrost promotes the growth of vegetation, leading to an increase in the rate of upward movement of alpine timberlines.

This study aims to focus on permafrost and the alpine timberline, both of which are crucial aspects of environmental science and engineering research. We combined the alpine timberline and permafrost to analyze the response pattern of the alpine timberline to the degradation of permafrost at high altitude in northeastern China. It is our hope that this research will contribute new perspectives to understanding ecological changes in permafrost regions and support sustainable development, while also providing insights for addressing global climate change.