Spatio-Temporal Evolution of Ecological Sensitivity in the Desert of China from 1981 to 2022

Abstract

The northern desert of China plays an important strategic role in land resource security and national economic development. Research on the spatio-temporal changes of ecological sensitivity can provide a scientific reference for desert management and ecological restoration in arid and semi-arid areas in northern China. This paper takes the northern desert of China as the research area, uses the spatial distance model to build a comprehensive ecological sensitivity evaluation index system, and discusses the spatio-temporal evolution characteristics of ecological sensitivity in the area from 1981 to 2022. The results show the following: (1) The land use types in the northern desert of China are mainly sandy land, grassland and other lands. The changing areas of grassland and other lands are 74,353.14 km² and 50,807.97 km², which is an important factor affecting the ecological sensitivity in the northern desert of China. (2) Five aspects, including terrain, climate, hydrology, soil and vegetation, influence and restrict each other, and jointly create the background conditions for the distribution and change of ecological sensitivity in the northern desert of China. Climate and terrain are the most important influencing factors affecting the ecological sensitivity of northern desert of China. Vegetation is the most active and basic factor affecting the ecological sensitivity of northern desert of China. Hydrology and soil have a certain limiting effect on the ecological sensitivity of northern desert of China. (3) The spatial heterogeneity of ecological sensitivity in the northern desert of China is significant, showing the characteristics of high volatility in the west, low volatility in the central region and low volatility in the east. (4) For nearly 42 years, ecological sensitivity of the northern desert of China shows first increasing and then decreasing characteristics. The area of the fluctuation reduction zone accounts for 26.34% of the total research area, of which the area of extreme sensitivity and mild sensitivity varies by 11.84% and 65.28%, respectively. (5) The spatial aggregation characteristics of ecological sensitivity have changed significantly, and the area of high–high and low–low agglomeration areas has also been decreasing, indicating that the environment is obviously improving. In the future, we should pay attention to the efficient use of natural resources in the northern desert of China and strengthen the protection of all kinds of land to achieve the sustainable development of the regional environment.

1. Introduction

Ecological sensitivity refers to the sensitivity of the ecosystem to various adverse factors when it is restricted by the natural environment and human activities. It can not only reflect the susceptibility of environmental imbalance but also identify the intrinsic relationship of various influencing factors within the ecosystem [1,2]. With the rapid development of the human population and social economy, the scope and intensity of the influences of human activities on the natural environment and ecosystem are increasing. The intensification of desertification has led to a sharp reduction in available land and a serious compression of human production and living space, restricting the coordinated and sustainable development of the local environment and the social economy [3]. Ecological sensitivity evaluation can effectively identify the quality and stability of the potential environment, which is of great guiding significance for the rational use and protection of desert resources, and also provides scientific reference for the construction and management of China’s desert environment [4].

At present, scholars have mainly studied the spatio-temporal characteristics of ecological sensitivity in regions, basins, cities and countries [5,6,7,8]. In terms of the selection of the evaluation index system, scholars select different indicators according to the needs of the research to build the evaluation system, and then reveal the intrinsic correlation of ecosystem changes in the region [9]. They mainly build an ecological sensitivity evaluation index system from the perspective of climate, soil erosion and land desertification [10,11,12]. Eggermont et al. [13] selected lake sediment and average temperature indicators to build a variety of inference models to explore the spatio-temporal distribution characteristics of the ecological sensitivity of lakes and swamps in the high-altitude Rwenzori area. Tsou et al. [14] used remote sensing data and geographic information system technology to select geomorphological factors, land use types and other indicators to analyze the ecological sensitivity of Hangzhou, China. However, the research on ecological sensitivity is still in the stage of exploration and development. There is no standard specification for the selection of indicators, there is uncertainty and arbitrariness, the degree of spatial dependence is high, and the scale effect has a great impact on the accuracy of the research results.

At present, there are many evaluation methods for ecological sensitivity research, such as the hierarchical analysis method, coefficient of variation method, expert scoring method, etc. [15,16,17]. Wu et al. [16] used the objective confirmation method–variation coefficient method to determine the weight, and analyzed the ecological sensitivity of Lingbao City, Henan Province, with a focus on soil type and terrain. Wei et al. [7] selected four comprehensive indicators of soil erosion, water content, salinization and desertification, and used the principal component analysis method to explore the characteristics of spatio-temporal changes of the ecological sensitivity of Wuwei City. There are shortcomings in each evaluation method, and the calculation steps are complex and susceptible to subjective factors, which cannot fully reflect the actual ecological sensitivity in the northern desert of China and the details of different sensitive areas. The spatial distance model does not require experts to subjectively judge the size of the weight coefficient, which reduces arbitrariness in the calculation process of the index, and considers the amplification and tightening effect of the critical value on the comprehensive index of the system, which can effectively reflect the coordination and integration of the ecological system in the northern desert of China.

China’s desert has important ecological functions, which are mainly distributed in the northern region [18]. It is not only a natural underground reservoir and a treasure of biodiversity, but is also the second ecological security barrier to protect the north of China [19]. However, the environment of the desert is relatively sensitive. Under the influence of human activities such as overgrazing and reclamation, the balance of regional ecosystems has been destroyed, there has been serious land desertification, and a series of ecological security problems have emerged, such as grassland desertification, reduced water conservation capacity, land desertification and soil erosion [20]. These seriously restrict the coordinated development of the social economy and environment [21]. In recent years, the Chinese government has attached great importance to environmental protection and restoration projects, and has successively implemented policies, such as returning farmland to grazing, returning forests to grasslands, and ecological compensation, which have significantly improved ecological construction. In summary, taking the northern desert region of China as the research area, this paper selects six sensitivity indicators of terrain, climate, hydrology, soil, vegetation and land use; builds a comprehensive ecological sensitivity evaluation index system based on the spatial distance model; evaluates the ecological sensitivity of the northern desert region of China from 1981 to 2022; quantitatively reveals the spatial pattern and temporal evolution ins; and provides a scientific reference for desert-area governance and industrial development.

2. Study Area

The northern desert of China (30°–50° N, 75°–130° E) is mainly distributed in China’s inland arid and semi-arid climate zone (Figure 1), including Heilongjiang, Jilin, Liaoning, Inner Mongolia, Shaanxi, Ningxia, Gansu, Qinghai and Xinjiang [22]. Being affected by the climatic conditions of the Qinghai–Tibet and Mongolian Plateaus, as well as atmospheric circulation, annual precipitation gradually decreases from the southeast to the northwest. The average annual precipitation in the northwest arid areas is generally less than 200 mm, and the average annual precipitation in the eastern semi-arid areas generally does not exceed 500 mm. The wind force of the wind season is generally above 5 to 6, wind and sand days are 20–100 d, the average annual temperature is −5–20 °C, sunshine hours are 2600–3400 h and the frost-free period is 150–260 d. The river alluvials, alluvial lake deposits, flood alluvial deposits and bedrock weathering residues from underlying strata in desert areas provide rich sand sources for desert formation. The types of sand dunes, such as the Taklamakan Desert, the Badain Jaran Desert and the Tengger Desert, are mainly mobile sand dunes. The Gurbantunggut Desert, Mu Us Sandy Land, Otindag Sandy Land and Horqin Sandy Land are mainly fixed and semi-fixed sand dunes [23].

3. Materials and Methods

3.1. Selection of Indicator Factors

Ecological sensitivity means that under the dual interaction of natural factors and human activities, the various elements are spatially interconnected, influenced and mutually restricted to determine the heterogeneity of the spatio-temporal distribution in the northern desert of China [24]. With reference to the research results of ecological sensitivity analysis, combined with the natural, socio-economic conditions and environment characteristics of the research area itself, 15 index factors are selected from six aspects: terrain, climate, hydrology, soil, vegetation and land use [7,13]. There are three main indicators of terrain ecological sensitivity (TES), namely elevation, aspect and slope. The northern desert of China has a large longitude span, and its geographical location and environmental characteristics are significantly different. Different elevation have a certain impact on climatic conditions, vegetation types and soil types, while different slope and aspect will directly change the wind speed, precipitation, sunshine time, soil stability, water conservation capacity and vegetation coverage [24]. Therefore, the three indicators have a certain impact on the ecological sensitivity of the northern desert of China. Climate ecological sensitivity (CES) mainly selects the average annual temperature, average annual precipitation, average annual evaporation and average annual wind speed [7]. The climatic characteristics of different regions are significantly different. Under the joint action of the four climate indicators, it directly affects the hydrothermal balance conditions, soil erosion degree and vegetation coverage in the northern desert of China, thus affecting the ecological sensitivity. Hydrological ecological sensitivity (HES) mainly selects rivers, lakes and reservoirs, which not only reflects the degree of dryness in the northern desert of China, but also affects the diversity of vegetation and organisms, thus affecting the ecological sensitivity characteristics of the regional environment [25]. Soil ecological sensitivity (SES) mainly selects soil texture, soil type and soil erosion degree. Most of the soil environment in the northern desert of China is poor, which can affect the capacity of water conservation and vegetation coverage of the research area to a certain extent, thus changing the ecological sensitivity [26,27]. Vegetation ecological sensitivity (VES) mainly selects two indicators, namely, the normalized vegetation index and the net primary productivity. The vegetation in different regions of the northern desert of China improves and promotes the regional climate, hydrology, soil and other conditions to varying degrees, and has a certain inhibitory effect on the desertification process, thus reducing ecological sensitivity [28,29]. Land use ecological sensitivity (LUES) is mainly affected by human activities. Human activities can change the environment and make it develop in a direction that is not conducive to human survival, but it can also promote the coordinated and sustainable development of the regional environment. [30,31].

3.2. Data Sources and Processing

From the perspective of natural conditions and human activities, six first-level indicators for terrain, climate, hydrology, soil, vegetation and land use factors are selected to construct the ecological sensitivity evaluation index system of the area (

Table 1. Indicator source and processing.

Primary Indicators	Secondary Indicators	Data Sources	Model and Methods
Terrain factors	Elevation	Geospatial Data Cloud (http://www.gscloud.cn/ accessed on 19 January 2023)	Digital elevation model [7]
	Aspect
	Slope
Climate factors	Average annual temperature	China Meteorological Data Network (http://data.cma.cn/ accessed on 15 January 2023)	Kriging interpolation method [8]
	Average annual wind speed
	Average annual precipitation
	Average annual evaporation
Hydrological factors	River	National Geomatics Center of China (http://www.ngcc.cn/ accessed on 19 January 2023)	Spatial distance model [24]
	Lake	Geospatial Data Cloud (http://www.gscloud.cn/ accessed on 25 January 2023)
Soil factors	Soil erosion intensity	World Soil Database (http://www.fao.org/ accessed on 31 January 2023)	Vector data to raster data [24]
	Soil type	Resource and Environmental Science Data Center of the Chinese Academy of Sciences (http://www.resdc.cn/ accessed on 10 February 2023)
	Soil texture
Vegetation factors	Net primary productivity of vegetation	NASA website (https://www.nasa.gov/ accessed on 24 February 2023)	CASA model [26]
	Vegeration coverage		Dimidiate pixel model [13]
Land use factors	7 types of land use	Resource and Environmental Science Data Center of the Chinese Academy of Sciences (http://www.resdc.cn/ accessed on 26 March 2023)	Reclassification [30]

) based on environmental quality of the northern desert of China and the principles of science and integrity. The data of five periods in 1981, 1990, 2000, 2010 and 2022 are selected to carry out our research. The spatial reference is the Krasovsky coordinate Albers projection, and the spatial resolution of all indicators obtained and calculated through the data source is resampled to 1 km × 1 km raster data.

3.3. Research Methods

3.3.1. Single-Factor Ecological Sensitivity

European distance is to measure the absolute distance between points in multidimensional space. With objectivity, science and versatility, it has been widely used in various fields. In geography, European distance has been applied to environmental quality assessment and arid-zone detection. The spatial distance model refers to the basic point of the lowest level based on the research objective in multidimensional space, calculates the European distance from each point of the multidimensional space to the reference point of the lowest level, and judges the degree of ecological sensitivity of each single factor by the size of the distance value. This paper uses ArcGIS 10.2 software to study six primary indicators, terrain, climate, hydrology, soil, vegetation and land use, and uses the extreme difference change method to standardize each secondary indicators so as to remove the impact of different dimensions on the evaluation results, and build a one-factor ecological sensitivity index model in the northern desert of China with the help of the space distance model [32]. The formula is

$$S=\sqrt{{\displaystyle \sum _{i=1}^n{(I_i-I_{low})}^2}}$$

(1)

where S is the single-factor ecological sensitivity, I is the i-th ecological sensitivity factor, and I_low is the lowest value of each ecological sensitivity.

3.3.2. Comprehensive Ecological Sensitivity

In the current research, to calculate comprehensive ecological sensitivity in the evaluation of ecological sensitivity, a certain index is selected, different weights to the index are assigned, and multi-factors are weighted and superimposed. In order to overcome the subjectivity of the traditional empowerment method, this paper uses ArcGIS 10.2 software to deal with terrain ecological sensitivity, climate ecological sensitivity, hydrological ecological sensitivity, soil ecological sensitivity, vegetation ecological sensitivity and land use ecological sensitivity index. Select the lowest value of the six processed primary indicators as the reference point, calculate the distance from other points in space to the lowest point of sensitivity, and build a comprehensive ecological sensitivity index model in the northern desert of China based on the spatial distance model. The farther away from the reference point, the lower the ecological sensitivity [33]. The formula is

$$P={\displaystyle \sqrt{\begin{array}{l}{(TES-TE{S}_{low})}^2+{(CES-CE{S}_{low})}^2+{(HES-HE{S}_{low})}^2+\\ {(SES-SE{S}_{low})}^2+{(VES-VE{S}_{low})}^2+{(LUES-LUE{S}_{low})}^2\end{array}}}$$

(2)

where P is the comprehensive ecological sensitivity; TES, CES, HES, SES, VES and LUES represent terrain ecological sensitivity, climate ecological sensitivity, hydrological ecological sensitivity, soil ecological sensitivity, vegetation ecological sensitivity and land use ecological sensitivity, respectively. TES_low, CES_low, HES_low, SES_low, VES_low and LUES_low are the lowest reference points, respectively. Referring to the environment of the northern desert and the natural breakpoint method, the ecological sensitivity of the northern desert of China is divided into five categories, namely, extreme sensitivity, severe sensitivity, moderate sensitivity, mild sensitivity and insensitivity.

3.3.3. Local Spatial Autocorrelation

Local spatial autocorrelation represents the correlation of spatial adjacent areas. If there is spatial correlation, its spatial aggregation characteristics can be determined. If it does not exist, it can reflect the spatial correlation phenomenon that may be masked. This paper uses the Moran index to analyze the spatial aggregation degree of comprehensive ecological sensitivity in the northern desert of China from 1981 to 2022, and uses ArcGIS 10.2 software to set the spatial relationship to Inverse Distance and the distance method to Euclidean Distance [34]. The formula is

$$I=\frac{(X_i-\overline{X})}{S^2}{\displaystyle \sum _j{W}_{ij}}(X_j-\overline{X})$$

(3)

where I is the Moran index and is the sensitivity mean of the i unit, the ecological sensitivity mean of the j-th evaluation unit and the ecological sensitivity mean of all evaluation units; W_ij is the spatial weight matrix; and S is the sum of the elements of the spatial weight matrix.

4. Results and Analysis

4.1. Changes in Land Use Types in Northern Desert Areas

Sandy land, grassland and other lands are the main types of land use in the northern desert of China (Figure 2). From 1981 to 2022, the change of land use type was more obvious, with the largest change in grassland (46,016.33 km²), followed by other lands (30,402.17 km²), and relatively small changes in farmland (9996.36 km²), forest (1982.75 km²), water (826.15 km²), construction land (1487.06 km²) and sandy land (2974.13 km²). From 2000 to 2010, the area used for different types of land changed greatly, and the areas of grassland and water decreased by 19.08% and 15.22%, respectively; the areas of sandy land and other lands increased by 0.47% and 32.78%, respectively. Reductions to water and grassland areas in the eastern part of the study area were mainly due to the increase in farmland, forest and construction land area, and the reduced part of the grassland area in the west and central region of the study area were mainly due to the increase in other lands and sandy land area.

There was a significant change between different land use types in the northern desert of China from 1981 to 2022 (Figure 3). The grassland transfer area was the highest (74,353.14 km²), followed by the sand transfer area (50,807.97 km²). The areas of grassland transfer into farmland, forest, water, construction land and sandy land were relatively high, at 9 673.07 km², 5644.62 km², 651.64 km², 811.10 km² and 43,515.81 km², respectively. The areas of sandy land transfer into grassland and other lands were relatively high, at 13,253.18 km² and 34,643.18 km², respectively. The area of other lands transfer into water area was 663.87 km². From 2000 to 2010, the areas of grassland and sandy land transferred out to other land use types were relatively high, at 63,406.32 km² and 47,354.93 km², respectively, of which the area of grassland transferred to sandy land was 4.75%, and the area of sandy land transferred to other lands was 4.16%.

4.2. Single-Factor Ecological Sensitivity Spatial Distribution Characteristics

The spatial distribution characteristics of single-factor ecological sensitivity are closely related to the geographical location of each desert (Figure 4). Areas with high terrain ecological sensitivity are mainly the Qaidam Basin Desert, Kumukuli Basin Desert and Gonghe Basin Desert located in the Qinghai–Tibet Plateau in the western part of the study area. The area is above 2700 m above sea level and the sunny slope is relatively high, while the terrain sensitivity of the Gurbantunggut Desert in the west and eastern desert of the study area is low, and the maximum difference in terrain sensitivity in 2022 was 1.176. Climate ecological sensitivity gradually declines from west to east. The ecological sensitivity of the western part of the study area is high, and the maximum value of climate sensitivity in 2022 was 1.008. This is because the region is mainly located in the Eurasian continent and is greatly affected by the Mongolian and Qinghai–Tibet Plateaus. Hydrological ecological sensitivity is mainly related to the distribution of rivers and lakes. The sensitivity of Mu Us Sandy Land and Tengger Desert in the center region of the study area is low. The minimum hydrological sensitivity in 2022 was 0.015. This is due to the superior water conditions in this area, as there are many lakes and rivers. The difference in soil ecological sensitivity is relatively low in the northern desert of China. The difference between the maximum and minimum values of soil ecological sensitivity in 2022 was 0.608. This is due to the poor soil background conditions in this area. The soil is mainly made up of loose sand grains, which is not conducive to plant growth, the sand gap is large, precipitation seeps into the ground so it is not easy to form surface runoff, and the surface layer of dry sand prevents water evaporation. Vegetation ecological sensitivity decreases from west to east. The ecological sensitivity of the eastern part of the study area is low, and the minimum vegetation sensitivity in 2022 was only 0.002. This is due to the good climatic conditions in the eastern part of the study area, the superior water quality conditions and the luxuriant plant growth.

4.3. Comprehensive Ecological Sensitivity of Space-Time Evolution Characteristics

4.3.1. Spatial Differentiation Characteristics

From 1981 to 2022, the ecological sensitivity of the northern desert of China was characterized by high in the west, low fluctuation in the center region and low in the east, and the ecological sensitivity was variable from west to east (Figure 5). On the whole, the ecological sensitivity of the northern desert of China is mainly extreme and severe sensitivity, accounting for more than 60% of the study area. The area of extreme and severe sensitivity in 2000 was 619,456.21 km², which was the maximum of nearly 42 years. It was mainly distributed in the western and central study area, and there was also a small distribution in the eastern part of Horqin Sandy Land in the eastern part of the study area. The insensitive areas are relatively small. In 1990, the insensitive area only accounted for 3.98% of the study area, which has been the minimum value of nearly 42 years. In 2022, the maximum area of insensitivity accounted for 20.60% of the area of the research area. It was mainly distributed in the eastern and central parts of the study area, and there were also a small number in Kumukuli Basin Desert, Qaidam Basin Desert and Gonghe Basin Desert in the western part of the study area. The moderate and mild sensitive areas fluctuated during the research period. In 1990, the area of the two was the highest in nearly 40 years, accounting for 35.45% of the study area. In 2022, the area of the moderate and mild sensitive was the lowest in nearly 40 years, accounting for 15.52% of the study area, and was mainly distributed in the eastern and center region of the study area.

4.3.2. Time Series Evolution Characteristics

From 1981 to 2022, the ecological sensitivity of the northern desert of China as a whole first increased and then decreased (Figure 6 and Figure 7). The areas of extreme sensitivity and mild sensitivity decreased by 11.84% and 65.28%, respectively. The insensitive area increased by 133.46%. This shows that the ecological sensitivity of the 42-year study area is decreasing and that the environment is obviously improving. In 2000, the extreme sensitivity area reached its maximum, accounting for about 66.40% of the northern desert of China. From 1981 to 2000, the extreme sensitive area of the northern desert of China increased by 81,140.39 km², or 17.46%, while the areas of severe, moderate, mild and insensitive areas decreased to high sensitivity to varying degrees. From 2000 to 2022, ecological sensitivity decreased overall and areas of extreme and mild sensitivity decreased by 16.40% and 55.31%, respectively. The areas of severe, moderate and insensitivity increased by 54.65%, 10.05% and 262.03%, respectively, indicating that the ecological sensitivity of the research area is being transformed to low sensitivity to varying degrees.

According to the analysis of ecological sensitivity in the northern desert of China from 1981 to 2022, and it is divided into four types: perennial invariant area, fluctuation invariant area, fluctuation increase area and fluctuation reduction area. The ecological sensitivity perennial invariant area in the northern desert of China accounts for 50.73% of the total study area, of which the area of extreme sensitivity within the perennial invariant area accounts for 83.22% of that area. The unchanged area of ecological sensitivity fluctuation accounts for 18.93% of the study area, of which heavy and insensitive areas account for 7.26% and 5.35% of study area, respectively. The fluctuation increases area accounts for 26.34% of the study area. Among them, the transformation from mild sensitivity to insensitivity accounts for 9.66% of the study area, and the transformation from extreme sensitivity to severe sensitivity accounts for 6.84%. However, the fluctuation increases area accounts for 4.00%, indicating that the ecological sensitivity in the northern desert of China has gradually decreased.

4.4. Correlation Analysis of Ecological Sensitivity

Using the local Moran′s I index, the spatial clustering analysis of ecological sensitivity of the northern desert of China from 1981 to 2022 was carried out, and ecological sensitivity was further divided into five different types: high–high aggregation area, high–low abnormal aggregation area, low–high abnormal aggregation area, low–low aggregation area and non-significant area distribution (Figure 8). In 1981, high–high and low–low agglomeration areas accounted for 16.32% and 1.89% of the study area, respectively. In 2022, the high–high and low–low agglomeration areas were reduced, accounting for 12.95% and 0.70%, respectively. From 1981 to 2022, the overall trend of high–high agglomeration area decreased by 20.65%, and it was mainly extreme, severe and moderate sensitivity. It was mainly distributed in the Gurbantunggut Desert, the Shanshan Kumtag Desert, the Kumtag Desert, the Qaidam Basin Desert, the Tengger Desert and the Ulan Buh Desert. In 1981–2000, the high–high agglomeration area decreased by 31.07%, and the Gurbantunggut Desert contracted significantly. In 2000–2022, the high–high agglomeration area increased slightly by 15.11%, and the Kumtag Desert expands significantly. For nearly 42 years, the low–low agglomeration area decreased by 62.96%. It was mainly distributed in the Taklimakan Desert, the Mu Us Sandy Land and the Horqin Sandy Land, and the Taklamakan Desert shrank significantly. The high–low abnormal aggregation areas were mainly in the Taklamakan Desert, the Gurbantonggut Desert, the Badain Jaran Desert and the Tengger Desert. This was mainly due to the poor environment, relatively high ecological sensitivity and relatively weak human activities in the region. The low–high abnormal aggregation areas were mainly in East Sandy Land of the Yellow River and Mu Us Sandy Land. This was due to the relatively good environment in the region, relatively low ecological sensitivity, and was greatly affected by human activities. On the whole, the ecological sensitivity of high–high and low–low aggregation areas has developed in the direction of dispersion and shows a trend of continuous weakening, indicating that the environment tends toward improvement.

5. Discussion

5.1. Space-Time Evolution of Desert Ecological Sensitivity

Since the 1980s, natural resources, as an important means of production, have been continuously acquired by human beings, and vegetation has been destroyed, thus affecting the balance of other ecosystems and making the originally fragile northern desert of China more sensitive to the environment [35,36]. The desert ecosystem itself is complex and involves many fields, such as ecological, environmental and societal. This study finds that places with high ecological sensitivity are mainly located in the western part of the northern desert of China, which is consistent with the research results of Li et al. [26] and Guo et al. [24]. From 1981 to 2000, extreme sensitivity increased, but from 2000 to 2022, it declined. Jiang et al. [37] also found that the ecological sensitivity of the Badain Jaran Desert and the desert west of Langshan was relatively high, and the transition occurred to varying degrees with the change of time. Xu et al. [38] believed that the ecological sensitivity of the northern desert of China was generally reduced from west to east, such as Horqin Sandy Land and Songnen Sandy Land, where the ecological sensitivity was significantly lower than that of the Taklimakan Desert. Wei et al. [8] believed that the distribution of ecologically sensitive areas in northern China was continuous, but there are differences between extreme and severe sensitive areas and this study. This may be due to main types of soil erosion, desertified land and stony desertification considered in the study, as other factors were not considered.

5.2. Factors Influencing Desert Ecological Sensitivity

The ecological sensitivity of the northern desert of China is affected by many influencing factors, and the interaction between different factors is mutually restricted [39,40]. In this study, the ecological sensitivity of factors, such as terrain, climate, hydrology, soil, vegetation and land use, is selected to explore its impact on the ecological sensitivity of the northern desert of China [41]. From the perspective of the comprehensive environment system, terrain fluctuations affect the significant differences in climatic conditions, vegetation cover and soil types of geographical units in desert areas. The desert area is in the inland area with a dry climate, less precipitation and where it is not easy to form rivers. Under the erosion of wind and water, deserted grasslands are gradually formed. The soil in the northern desert of China is dominated by wind-sand soil and desert soil that is loose and small in size, and is easily eroded by water and wind power, which aggravates the desertification of desert areas. Due to the large area of the northern desert of China, the large difference in the spatial distribution of vegetation is related to factors such as terrain and climate. This creates an ecologically sensitive spatial distribution pattern of the northern desert. In general, climate and terrain play a direct role in the impact on the ecological sensitivity of the northern desert of China and are the most important influencing factors. Vegetation is the most active and basic factor affecting the ecological sensitivity of the northern desert of China, and high vegetation coverage can inhibit ecological sensitivity. Hydrology and soil have a certain limiting effect on the ecological sensitivity of northern desert areas. Shi et al. [42] argued that vegetation, humidity and desertification factors have a great impact on ecological sensitivity, and are mainly affected by hydrology, which in turn affects humidity, soil desertification and salinization. Qiao et al. [43] analyzed the ecological sensitivity of the northeast forest and grassland transition zone using the four evaluation indicators of climate, terrain, soil and vegetation based on GIS technology, and found that vegetation cover and terrain played a leading role in ecological sensitivity.

In order to meet the needs of socio-economic development, human beings have changed land use types to a certain extent. The combined effects of human activities and the environment have changed ecological sensitivity in the study area, which is consistent with the research results of Wang et al. [44] and Liu et al. [45]. Land use change has a great impact on comprehensive ecological sensitivity. The land use types in the eastern part of the research area are mainly grassland. Its environment is good, and its ecological sensitivity is relatively low, but the interference caused by human production and life to the surrounding environment is the reason for the change of ecological sensitivity. The land use types in the central and western parts of the research area are mainly sandy land and other land. Due to the poor background conditions of the environment, its ecological sensitivity is also high. Li et al. [46] analyzed the ecological sensitivity of the Shiyang River Basin from a variety of factors such as ecological risk, hydrothermal environment and biodiversity, and found that the environment in areas with poor natural background conditions is relatively fragile and easily affected by the external environment, while the environment in places with strong human activities is also relatively fragile.

5.3. Desert Environmental Management and Restoration

After 2000, the state and local governments have improved the desertification of the land, increased vegetation coverage and further improving the desert environment through the establishment of nature reserves and the implementation of ecological compensation programs, such as the conversion of farmland to forests and grasslands [47]. The extreme and severe sensitivity areas in the study area are mainly distributed in the western part. Due to the relatively weak human activities, the restoration of the environment should be mainly based on natural recovery and supplemented by artificial recovery. The vegetation on the edge of the desert is covered mainly by grassland and can prevent the further expansion of sand. According to the regional characteristics of the desert, a unique natural protection mechanism is established to protect the centralized zoning of desert plants and animals, thus affecting the species diversity of desert areas in China [48,49]. Human activity is relatively strong in areas with large changes in ecological sensitivity, and the environmental changes are also relatively large. This is due to the unreasonable development and utilization of land types by human beings, the increase in arable land area, and the reduction in desert rivers and lakes. Therefore, the development of farmland should be reduced in the desert. When returning farmland to forests, grasslands and lakes, the local environment and climatic conditions should be taken into account, and trees and grass should be planted reasonably to prevent the intensification of groundwater consumption and the expansion of desertification [50,51]. On the whole, the ecological sensitivity in the northern desert of China is still at a high level, and environmental management still has a long way to go. For areas with increased ecological sensitivity, environment management projects should be carried out to increase the protection of the environment, reduce the damage of human activities, and prevent the further deterioration of the environment. For areas with less ecological sensitivity, the local governance model should be maintained.

In this study, the Moran index can be used to distinguish the positive and negative correlations of space, but it cannot further distinguish between the hot spot area and the cold spot area of the positive correlation of the space. The research results of Zou et al. also confirmed this [52]. This study explores the ecological governance of the northern desert of China and reveals its governance effectiveness. However, the ecological impact of the natural environment and human activities on the northern desert of China is a long-term and very complex process, involving ecology, geography, climatology and other disciplines. To practice ecological governance more scientifically and effectively, its specific internal mutual mechanism still needs to be further investigated and studied.

6. Conclusions

The spatial distribution characteristics of terrain, climate, hydrology and vegetation ecological sensitivity in the northern desert of China are related to the geographical location of each desert, and the ecological sensitivity index gradually increases from west to east. In the past 42 years, the land use types in the northern desert of China were mainly sandy land, grassland and other lands, and the areas of grassland and other lands vary greatly. The areas of grassland and other lands transferred to other land use types were 74,353.14 km² and 50,807.97 km², respectively. Grassland was mainly transferred into farmland, forest, water, construction land and sandy land. Sandy land was mainly transferred into grassland and other lands. Other lands is mainly transferred into water.

The ecological sensitivity space of the northern desert of China is high in the west, low and variable in the middle and low in the east, and this first increases and then decreases over time. On the whole, ecological sensitivity is dominated by extreme and severe sensitivity, accounting for more than 60% of the study area. In 2022, the insensitive area reached a maximum and accounted for 20.60% of the study area. For nearly 42 years, the areas of extreme and mild sensitivity have decreased by 11.54% and 65.28%, respectively, and the insensitive area have increased by 133.46%. The fluctuation reduction area and the fluctuation increase areas account 26.34% and 4.00% of the study area, respectively, and the environment tends toward improvement.

In 2022, the high–high and low–low agglomeration areas accounted for 12.95% and 0.70% of the study area, respectively. From 1981 to 2022, the high–high and low–low aggregation areas decreased by 20.65% and 62.96%, respectively. Ecological sensitivity develops in the direction of discretization with a trend of continuous weakening. Terrain, climate, soil, hydrology, vegetation and land use affects the ecological sensitivity of desert areas in northern China to varying degrees. According to the characteristics of different deserts, corresponding protection mechanisms can be established, and desert resources can be reasonably utilized to promote the further development of the desert environment.