Spatial Distribution Pattern and Natural Causes Analysis of Sandy Desertification Land in Ali Area

Abstract

In order to explore the spatial distribution pattern and natural causes of the sandy desertification land in the Ali area, on the basis of clarifying the dynamic change laws of sandy desertification land areas, sandy desertification degree and spatial distribution pattern, the main controlling factors of sandy desertification land distribution are analyzed from three aspects of landform, climate and vegetation. During the 22 years from 1992 to 2014, the sandy desertification land area in the Ali area shows a law of first increase and then decrease, reaching a peak of 61,054.14 km² in 2004, accounting for 20.57%, and decreasing to 60,892.65 km² in 2014, accounting for 20.51%, which do not return to the level of 1992. Sandy land desertification is mainly slight and moderate, accounting for 53.29% and 45.73%, respectively, in 2014. Sandy desertification land in the Ali area is mainly distributed among intermountain basins, river valleys, lake basins, piedmont plains and other landform units. The landform and wind speed are the main natural factors that determine the spatial distribution pattern of sandy desertification land in the Ali area, that is, the spatial distribution pattern of sandy desertification land in the Ali area is the coupled result of sand source and wind speed.

1. Introduction

The Ali area is located in the west of Tibet and is known as the “third pole of the world”. It is one of the regions with the harshest natural conditions in Tibet. High altitude, low temperatures and high drought make the local ecological environment particularly fragile. In addition, strong winds and rich surface sand materials provide natural conditions and a foundation for the occurrence and development of sandy desertification [1,2]. Overcutting, overgrazing and rat activities are factors that can stimulate and accelerate sandy desertification [3]. At present, sandy desertification has become one of the most prominent ecological and environmental problems in the Ali area.

The current research on desertification focuses on the cause, process, damage assessment, and control measures. Many studies show that the causes of desertification can be divided into two categories, namely, natural causes and human causes. It is generally accepted that natural causes include climatic warming, aridification, and strong and frequent wind erosion. The human causes are mainly high-intensity land use, including over-cultivation, overgrazing, overcutting fuelwood and over-extracting groundwater. According to Li`s study [4] on the dynamic changes and development trend of land sandy desertification in the Tibet Plateau in the 1990s, the degree of sandy desertification was significantly aggravated during this period, and the driving forces were climate and human activities of excessive reclamation, grazing and woodcutting. His other research [3] shows that the sandy desertification land area in the Ali area has decreased by 219.58 km² in the last 10 years. The development and evolution of sandy desertification in the Ali area in the last 40 years is the result of the comprehensive action of climate drying and human and rat activities. The natural factors affecting land sandy desertification are mainly climate, soil and vegetation. In terms of the spatial distribution pattern of sandy desertification land, Jin et al. [5] consider that the distribution pattern of wind-sandy land in Tibet is determined by both natural and human factors. However, some studies consider that the formation and development of land sandy desertification in Tibet is mainly driven by natural factors, not caused by excessive human activities [6]. Li et al. [7] considers that landform and climate are important factors controlling the spatial distribution pattern of sandy desertification land on the Qinghai-Tibet Plateau. Some studies suggest that land use, relating to land cover, is one of the influential factors associated with desertification risk [8,9,10,11,12]. In Ge’s research [8] on Horqin Sandy Land (HSL) in exploring the association between sandy desertification and its driving factors in the context of the synergistic roles of both climatic and anthropogenic variables, he developed an indicator for land use intensity, and considered that the continuous increase in socioeconomic demands will increase land use intensity in the long run, so that the risk of sandy desertification remains high. A similar study [9] uses spatial analysis and the MEDALUS model to investigate the extent of land degradation, land use changes and desertification risk, and also considers that land use changes can influence desertification risk [11]. This is based on GIS and RS techniques, using the Environmental Sensitivity Area Index (ESAI) to assess the land degradation sensitivity of an area contributing to desertification risk in the Lam Ta Kong Watershed. His research pointed out that the key factors causing land degradation and desertification are related to the soil factors, the land utilization and vegetation factors. In summary, the causes of desertification mainly include natural causes and human causes, and each contains complex and diverse subfactors. The main driving factors of desertification may be different in different research areas, which should be analyzed according to the natural and cultural conditions of the study area.

At present, the main research method for desertification monitoring and assessment is remote sensing interpretation, which is based on the constructed desertification land grading index system, and interprets remote sensing images through the GIS platform. The desertification land grading index system should be comprehensively considered from three aspects: natural, biological, agricultural (including land use status), social and economic indicators [10]. However, most of the research ignores social and economic indicators. Some scholars propose that the surface morphological change is a main indicator to assess desertification state and trend. Many other factors directly related to surface morphological changes can be used as additional indicators, such as vegetation cover, plant community structure, organic matter content and soil moisture content [4], using vegetation coverage, percentage of bare sand occupied and surface gravel content as the main indicators, and surface landscape information, such as surface erosion, sand dune morphology and soil type, as auxiliary indicators to identify the type and degree of desertification land on the Qinghai Tibet Plateau. We have used vegetation coverage, bare sand area percentage and surface gravel content [13] to construct a similar index system to study the sandy desertification degree and regional assessment in Tibet, and believe that the sandy desertification degree in Tibet is relatively light.

Chinese scholars have studied sandy desertification in Tibet for more than 30 years and consider sandy desertification as one of the most prominent ecological and environmental problems in Tibet [14]. Wang et al. [1] believe that the Ali area is one of the main distribution areas of sandy desertification land in Tibet, and the degree of land degradation has increased significantly. The grassland area in the Ali area accounts for about 70%, and the grassland degradation rate has increased from 7.98% in the 1980s to 20.75% in 2012 [15]. The degraded grassland has gradually developed into potential sandy desertification land, which seriously hinders regional ecological security and sustainable economic development. Data from China’s five sandy desertification and sandy desertification monitoring and Tibet’s sandy desertification census show that the sandy desertification land area in the Ali area accounts for more than 20% of the total area, which is one of the two regions with the highest proportion of sandy desertification land in Tibet. Land sandy desertification has seriously affected the sustainable development of society, the economy and the environment in the Ali area and has had a significant and far-reaching impact on the production environment of the Qinghai-Tibet Plateau, East Asia, Southeast Asia and South Asia by affecting climate change and material circulation [14,16]. At present, most of the relevant studies are on sandy desertification in the whole of Tibet, and there are few relevant reports on the Ali area where sandy desertification land is concentrated. In addition, there are different standards for determining the degree of sandy desertification, and problems such as outdated data are common. Furthermore, few studies have analyzed the relationship between vegetation and sandy desertification land distribution patterns. This study is based on the China desertification survey and Tibet desertification survey from 1992 to 2014, the 1:100,000 sandy desertification land distribution dataset in China in 2000, the 1:1,000,000 Vegetation dataset in China in 2001, and the Shiquanhe, Gêrzê and Purang meteorological data in the Ali area from 1981 to 2010. RS (Remote Sensing) and GIS (Geographic Information System) technologies are used to interpret and interpolate the data. On the basis of clarifying the dynamic change law of sandy desertification land area, sandy desertification degree and spatial distribution pattern, this paper systematically analyzes the relationship between landform, climate, vegetation and the spatial distribution pattern of sandy desertification land, and clarifies the main controlling factors of the spatial distribution pattern of sandy desertification land in the Ali area, so as to provide a theoretical basis for sandy desertification prevention and control in the Ali area.

2. Materials and Methods

2.1. Study Site

The Ali area (See Figure 1) is the only administrative office of the Tibet Autonomous Region. It is located between 78°24′ E~86°20′ E and 29°41′ N~35°52′ N, with a span of 742 km from east to west and 688 km from north to south, covering an area of 345,000 square kilometers. It is the driest region in Tibet and also one of the areas with the most serious sandy desertification, accounting for more than 20%. The average altitude of the Ali area is 4500 m, the average annual temperature is 0 °C, there is drought and the average annual rainfall is less than 100 mm. The vegetation coverage is only 24.52%. The local rainfall is not only low, but also shows an uneven distribution of time and space. The average annual precipitation of Purang is 168 mm, while that of Ritu is only 54 mm. Rainfall is concentrated from May to September, accounting for 80% of the annual rainfall. Strong wind is also the climate characteristic of the Ali area, with an average annual wind speed of 3.23 m/s.

2.2. Data Sources and Methods

This study used five periods of China’s desertification and sandy desertification monitoring data and the general survey data of desertification in Tibet from 1992 to 2014, the 1:100,000 sandy desertification land distribution dataset in China in 2000 [17], the 1:1,000,000 Vegetation dataset in China in 2001 [12], and the Shiquanhe, Gêrzê and Purang meteorological data in the Ali area from 1981 to 2010 (https://data.cma.cn/, accessed on 10 January 2022).

The sandy desertification land distribution dataset [13] is based on the land use map of China and the TM (Thematic Mapper) digital image information in 2000. After interpretation, extraction and revision, it is obtained by using RS and GIS technology. The vegetation dataset [18] is digitized according to the 1:1,000,000 Vegetation Atlas of China [19]. The 60 maps in the atlas are digitized (polygon properties), then projected, matched and spliced. Finally, each polygon is assigned vegetation properties. The vegetation dataset is classified according to vegetation type groups, and the Ali area mainly includes alpine vegetation, desert vegetation, grassland and meadow. All data were processed with ArcGIS10.6 software.

The sandy desertification dynamic degree is an index reflecting the annual growth rate of sandy desertification land in a period of time, and the calculation formula is as shown in Equation (1) [4].

$$\mathit{SD}=\frac{U_b-U_a}{U_a} \times \frac{\mathit{1}}{t} \times 100\%$$

(1)

where: SD is the sandy desertification dynamic degree of a certain place or type; U_a is the sandy desertification land area in the initial year; U_b is the sandy desertification land area in the final year; t is the calculation interval.

Aeolion sandy desertification index(ADI) is an index that can comprehensively evaluate the degree of sandy desertification [20,21]. The calculation formula is as Equation (2):

$$\mathit{ADI}=\frac{S_{\mathit{sl}}{+2S}_{\mathit{mo}}{+3S}_{\mathit{se}}{+4S}_{\mathit{ex}}}{S_A}$$

(2)

where, ADI is aeolion sandy desertification index; S_sl is the land area of slight sandy desertification; S_mo is the land area of moderate sandy desertification; S_se is the land area of severe sandy desertification; S_ex is the land area of extremely severe sandy desertification; S_A is the total area.

3. Results and Discussion

3.1. Dynamic Changes of Sandy Desertification Land Area

From 1992 to 2004, in the Ali area, the area of sandy desertification land increased from 60,506.69 km² to 61,054.14 km², an increase of 0.90%, and the proportion of sandy desertification land increased from 20.38% to 20.57%. From 2004 to 2014, the area and proportion of sandy desertification land continued to decline. In 2014, the area of sandy desertification land decreased to 60,892.65 km², a decrease of 0.26%, and the proportion decreased from 20.57% to 20.51%. In general, the area and proportion of sandy desertification land increased first and then decreased from 1992 to 2004, and 2004 was the peak year. Although the period from 2004 to 2014 did not increase but decreased, there was still a slight increase in 2014 compared with 1992. The area of sandy desertification land increased by 385.96 km², an increase of 0.64%, and the proportion increased from 20.38% to 20.51%.

There is a similar law in Tibet and China as in the Ali area. The area of sandy desertification land increases first and then decreases, and 2004 is the peak year. However, compared with Tibet and China, the proportion of sandy desertification land in the Ali area is significantly higher (see Figure 2). The average area of sandy desertification land in the Ali area in the past 22 years accounted for 20.46%, 3.15% higher than that in Tibet and 2.49% higher than that in China. This indicates that the sandy desertification land in the Ali area is more widely distributed than that in Tibet and China.

From 1992 to 2014, the dynamic degree of sandy desertification in the Ali area at different time intervals is shown in Figure 3. During the period from 1999 to 2014, the annual growth rate of sandy desertification land in the Ali area was relatively high, reaching 0.25%, indicating that the area of sandy desertification land increased rapidly during this period. During the 10 years from 2004 to 2014, the SD value continued to decrease, indicating that the sandy land desertification in the Ali area had been controlled since 2004 and had a decreasing trend year by year. However, the SD value calculated from 1999 to 2014 is still positive at 0.029%, indicating that the sandy desertification land area has not recovered to the level of 1999.

3.2. Sandy Desertification Land Classification and Degree Evaluation

In the current technical regulations on desertification monitoring in China, the main indicator of desertification degree is vegetation coverage. Based on the ecological base theory and referring to the existing research results [3,4,14,22], the indicator of bare sand area proportion is increased. At the same time, in order to integrate with the general survey data on desertification in Tibet, the sandy desertification land in the Ali area is divided into three levels and 10 categories (See

Table 1. Sandy desertification land degree classification and landscape characteristics.

Degree of Sandy Desertification	Vegetation Cover	Bare Sand Area Proportion	Sandy Desertification Type	Landscape Characteristics of Sandy Desertification Land
Severe-extremely severe	≤25%	≥30%	Mobile sand dune or drifting sand land	Wind-sand activity are widespread, sand dunes are widely distributed, dense sand dunes in the form of barchan, transversal longitudinal dunes, reticulate and climbing dunes, the soil is mobile aeolian sandy soil, generally presents desert vegetation landscape.
			Eroded inferior lands	Wind erosion is intense, the landform is broken, and there is basically no vegetation.
			Semi-mobile sand dune or drifting sand land	Wind-sand activity is widespread, sand dunes are widely distributed, sparse sand dunes in the form of barchan, transversal longitudinal dunes, reticulate and climbing dunes. The soil is mobile aeolian sandy soil and semi-fixed aeolian sandy soil, which generally presents a desert vegetation landscape.
			Wind erosion residue	Wind erosion is serious, and there are landform combinations such as wind erosion hills, wind eroded relic mounds and blowout troughs.
Moderate	25–40%	10–30%	semi-fixed sand dune	Wind-sand flow activity is common, dune patches or sporadic distribution, mostly semi-fixed sand dunes, sand dunes and semi-fixed climbing dunes. The soil is semi-fixed aeolian sandy soil, generally showing the desert or desert grassland landscape.
			Bare gravel land	The surface is roughened by wind erosion, covered with coarse sand and gravel, with Nebkhas, presenting the Gobi landscape as a whole.
Slight	>40%	<10%	Fixed dune or sand land	There are wind-sand flow activities, dune patches or sporadic distribution and mostly fixed sand dunes, sand dunes and fixed climbing dunes. The soil is aeolian sandy soil, generally showing the grassland or desert grassland landscape.
			Semi-bared gravel land	The surface is roughened by wind erosion, with Nebkhas, presenting the desert grassland landscape as a whole.
			Wind erosion cultivated land	Cultivated land with wind erosion traces on the surface and sporadic quicksand or small sand dunes.
			Engineering sand control land	Sand dunes or sand lands fixed by abiotic measures.

). The areas of slight, moderate, severe and extremely severe sandy desertification land in the Ali area are 32,128.43 km², 104,266.75 km² and 3775.11 km², respectively. The proportion of slight and moderate sandy desertification land is 53.29% and 45.73%, respectively, higher than that of severe and extremely severe, and the proportion of severe and extremely severe was only 0.98%. The areas of slight, moderate, severe and extremely severe sandy desertification land in Tibet are 93,716.15 km², 27,568.04 km² and 590.75 km², respectively. The proportion of slight and moderate sandy desertification land is 51.68% and 47.45%, respectively, higher than that of severe and extremely severe, and the proportion of severe and extremely severe was only 1.87%. The land areas of slight, moderate, severe and extremely severe sandy desertification land in China are 261,143 km², 25,3619 km², and 1,206,412 km², respectively, of which severe and extremely severe account for the highest proportion, reaching 70.09% (See Figure 4). The Ali area is similar to Tibet, and the proportion of slight and moderate sandy desertification degrees is significantly higher than that of severe and extremely severe desertification degrees, reaching 53.29% and 45.73%, respectively. The proportion of severe and extremely severe is only 0.98%, which is significantly lower than that of 70.09% in China. It can be seen that although the desertification land in the Ali area is widely distributed, the degree of sandy desertification is light, mainly slight and moderate and the proportion of severe and extremely severe is very small.

In order to more accurately evaluate the development status of regional sandy desertification, the comprehensive quantitative index ADI, including sandy desertification land area and sandy desertification degree information, is used to evaluate the desertification degree. ADIs calculated for the Ali area, Tibet and China are shown in Figure 5. The ADIs of the Ali area, Tibet and China are 0.258, 0.255 and 0.457, respectively. There is almost no difference in ADI between the Ali area and Tibet, which is only 1.18% higher than Tibet, but 43.54% lower than China, indicating that the degree of sandy desertification in Ali is equal to Tibet and significantly lower than China.

3.3. Spatial Distribution Pattern and Natural Causes of Sandy Desertification Land

The sandy desertification land in the Ali area is widely distributed, accounting for 20.51% of the total area, mainly distributed in intermountain basins, river valleys, lake basins, piedmont plain and other landform units (See Figure 6). The formation of aeolian sandy land requires three conditions: abundant sand sources, an arid and windy climate and sparse vegetation [5]. These conditions are all available in the Ali area. This study is based on sandy desertification land distribution, landform distribution, vegetation distribution and meteorological data, to analyze the natural causes of the spatial distribution pattern of sandy desertification land in the Ali area from the topography, climate and vegetation.

3.3.1. Landform

Sandy desertification requires abundant sources of sand materials. The sources of sand in the Ali area are mainly local loose sediments. These loose sediments are detrital materials produced by the glacial and freezing weathering of mountain bedrock around basins and valleys [5,7]. The basins, mountains and lakes are alternately distributed in the Ali area. The lake basins and intermountain basins are low-lying and rich in loose sediments. Affected by landforms and surface sediments, the sandy desertification land in this area presents a large area of patches. In the river valley area, during the wet season, rivers can carry sediment and accumulate in the wide valley area, and during the dry season, they will be exposed in the floodplain, river center bar and low terraces, providing abundant sand material for sandy desertification land. These sand materials form sandy desertification land under the transport of wind and water flow [7]. In conclusion, landform is the key factor affecting the distribution pattern of sandy desertification land in the Ali area, and this conclusion is similar to the research of the Qinghai-Tibet Plateau and Tibet. Under the influence of landform, the sandy desertification land in the Ali area is mainly distributed in intermountain basins, river valleys, lake basins, piedmont plains and other landform units [5,7,14], such as the Shiquanhe Basin, the wide valley area of Garzangbu and Senge Zangbu, Lamucuo, the Nawucuo Lake Basin, etc.

3.3.2. Climate

Drought, low temperatures and strong winds are the main characteristics of the climate in the Ali area. The 30-year rainfall data of Shiquanhe, Gêrzê and Purang stations from 1981 to 2021 shows that the average annual rainfall in the Ali area is only 129.5 mm, which is generally lower in the northwest and higher in the southeast, with obvious regional differences (See Figure 7). The average annual rainfall of Shiquanhe in the northwest is only 66.4 mm. The average annual rainfall of Purang in the south and Gêrzê in the east reach 150.7 mm and 171.7 mm, 2.27 times and 2.59 times that of Shiquan river. The spatial distribution of sandy desertification land in the Ali area shown in Figure 6 does not show a similar pattern to that of the Qinghai-Tibet Plateau: from southeast to northwest, with the decrease in rainfall, the proportion of sandy desertification land decreases. Previous studies have shown that the wind erosion climatic erosivity has a good consistency with the spatial distribution of sandy desertification land [7,23]. Although the rainfall in the Ali area is higher in the southeast and lower in the northwest, it is not consistent with the spatial distribution of sandy desertification land. There is still a lot of sandy desertification land in Purang and Coqen in the south. The reason is that the places with relatively high rainfall in the Ali area still have an average annual rainfall of less than 200 mm for many years, and almost the whole area belongs to arid areas. Relevant studies show that the wind erosion climatic erosivity is affected by both wind speed and rainfall in non-arid regions [23], and is mainly affected by wind speed in arid regions [24]. Therefore, under the drought conditions in the Ali area, even if there are significant differences in rainfall between regions, the wind erosion climatic erosivity is less affected by the rainfall, so the rainfall is not the main meteorological factor affecting the spatial distribution pattern of sandy desertification land in the Ali area.

The average annual wind speed in the Ali area is greater than 3 m/s, and is generally higher in the eastern and central regions, and lower in the western region (see Figure 8). In Figure 6, the severe–extremely severe sandy desertification land is mainly distributed in the southeastern and central regions, which is related to the distribution of wind speed. In fact, similar to [7,23] on the spatial distribution of sandy desertification land on the Qinghai-Tibet Plateau, this study believes that the wind erosion climatic erosivity is strongly correlated with the spatial distribution of sandy desertification land. The difference is that almost all of the Ali area is an arid area, and only wind speed is the main controlling factor affecting the wind erosion climatic erosivity. In addition, some studies believe that the correlation between the wind erosion climatic erosivity and the spatial distribution pattern of sandy desertification land is not high [24,25].

3.3.3. Vegetation

Vegetation Spatial Distribution and Analysis of Main Controlling Factors

The vegetation types in the Ali area mainly include desert vegetation, alpine vegetation, grassland and meadow, and the demand for annual rainfall of the four vegetation types is increasing. The rainfall and vegetation distribution in the Ali area are in good consistency. Perpendicular to the rainfall isoline, the rainfall increases from northwest to southeast, and the vegetation types appear successively as desert vegetation, alpine vegetation, grassland and meadow in turn.

It is obvious from the vegetation distribution map (see Figure 9) that the vegetation in Gêrzê is mainly grassland and desert vegetation, while the vegetation in Purang is mainly grassland and meadow. The rainfall of Gêrzê is 13.93% higher than that of Purang (see Figure 10). This vegetation type has less rainfall but is more suitable for a humid environment, which shows that there are other factors affecting the distribution pattern of vegetation types in the Ali area. The average altitude of Purang is 3900 m, and that of Gêrzê is 4700 m, with an altitude difference of 800 m. Although the average annual rainfall in Purang is slightly lower than that in Gêrzê, the average annual temperature and humidity are 3.26 °C and 13.3% higher than those in Gêrzê, which are more suitable for meadow growth. Altitude has a strong correlation with air temperature and air humidity, and these two factors have a greater impact on vegetation growth. Therefore, differences in air temperature and humidity caused by altitude differences are also factors that affect the spatial distribution of vegetation types.

The three meteorological factors of rainfall, air temperature and air humidity are analyzed on a monthly scale (see Figure 11). The annual rainfall of Gêrzê is 21 mm higher than that of Purang, which is concentrated in summer with high air temperature and humidity (See Figure 11a) and the rainfall from June to September accounts for 89.44% of the whole year. Compared with Gêrzê, the rainfall in Purang is relatively scattered throughout the year, showing a bimodal distribution (see Figure 11b), with maximum values in both spring and summer, and the rainfall from October to the next May accounted for 55.81%. Concentrated rainfall is prone to surface runoff, reducing the amount of water available for vegetation, which is one reason why Purang has more meadow than Gêrzê. In addition, the monthly average temperature and humidity of Gêrzê are obviously lower than those of Purang. In terms of air temperature, the annual monthly average temperature of Gêrzê is 3.27 °C lower than that of Purang. In terms of air humidity, the average monthly air humidity of Gêrzê is 13.33% lower than that of Purang, especially from January to May before the rainy season, the humidity is significantly lower than that of Purang (see Figure 11c), and the monthly average air humidity is 17.80% lower than that of Purang.

The spatial distribution pattern of vegetation in the Ali area is the result of the comprehensive action of three main control factors: rainfall, air temperature and air humidity. The overall pattern is that the vertical to the annual rainfall contour line is distributed from northwest to southeast with desert, alpine vegetation, grassland and meadow. The higher the air temperature and humidity, the greater the proportion of meadow and grassland.

Effect of Vegetation Spatial Distribution on Spatial Distribution Pattern of Sandy Desertification Land

The spatial distribution pattern of vegetation types in the Ali area is clarified (see Figure 9), and it is found that there is no good consistency with the distribution of sandy desertification land. Even in Coqen and Purang in the south of Ali, which are relatively warm and humid and widely distributed with meadows and grasslands, 1175.51 km² of sandy desertification land is still distributed, and the severe–extreme severe degree is 332.94 km², accounting for 28.32% (see Figure 6).

Existing studies show that the main plant formations in the Ali area include Stipa basiplumosa, Stipa purpurea, Orinus thoroldii and Stipa gobica [26]. The total coverage of most vegetation communities is less than 30% and the height is less than 20 cm (see

Table 2. Main Plant formations, height of field layer and total coverage in the Ali Area.

Plant Formation	Plant Association	Field Layer Height Range (cm)	Total Coverage
Stipa basiplumosa	Stipa basiplumosa	8~20	10~65%
	Stipa basiplumosa + Solms-laubachia pulcherrima	3~17	3~14%
	Stipa basiplumosa + Orinus thoroldii	3~45	3~45%
	Stipa basiplumosa + Chamaerhodos sabulosa	1~4	4~8%
	Stipa basiplumosa + Stipa purpurea	5~14	2~25%
Stipa purpurea	Stipa purpurea	4~9	4~15%
	Stipa purpurea + Chamaerhodos sabulosa	5~9	4~15%
	Stipa purpurea + Solms-laubachia pulcherrima	2~10	17~27%
Orinus thoroldii	Orinus thoroldii	7~20	6~60%
	Orinus thoroldii + Stipa purpurea	11~45	10~40%
	Orinus thoroldii + Artemisia wellbyi	7~22	4~30%
	Orinus thoroldii + Solms-laubachia pulcherrima	4~11	8~35%
Stipa gobica	Stipa gobica	2~22	5~20%
	Stipa gobica + Ajania fruticulosa	3~13	3~7%
	Stipa gobica + Oxytropis microphylla	3~6	10~20%

). The vegetation is sparse and low, coupled with strong wind in the Ali area (see Figure 7), the effect of vegetation on wind prevention and sand fixation is not good. In addition, the vegetation has obvious seasonality. The climate in the winter half year (November to May) is dry and cold with strong winds, which is the period of vegetation withering. During this period, the wind-sand activity is active and accelerates the sandy desertification [5]. In general, due to the sparse, low and seasonal variability of vegetation, coupled with strong winds, the effect of vegetation on wind prevention and sand fixation is not good, and has little impact on the spatial distribution pattern of sandy desertification land.

4. Conclusions

Sandy desertification is one of the most prominent ecological and environmental problems in Tibet, and Ali is one of the two regions with the highest proportion of sandy desertification land in Tibet. The area of sandy desertification land in Ali has continued to decrease from 2004 to 2014, but it has not yet recovered to the level of 1992. Although the policy of forest rehabilitation from slope agriculture helps to reduce sandy desertification land, the situation is still grim.

This study systematically analyzes the natural causes of the spatial distribution pattern of sandy desertification land in Ali from the aspects of landform, climate and vegetation. We consider that the landform determines the distribution of sand materials in the Ali area. In terms of climate factors, we find that there is no good consistency between rainfall and the spatial distribution of sandy desertification land in Ali, which is obviously different from that in semi-arid areas. The reason is that the wind erosion climate erosivity has been proven to be in good consistency with the spatial distribution of sand desertification land. In arid areas, the wind erosion climate erosivity is mainly affected by wind speed rather than rainfall. Therefore, the main climatic factor affecting the spatial distribution of sand desertification land in Ali is wind speed. This conclusion is different from similar studies on the spatial distribution of sand desertification land in the Yellow River Basin. In terms of vegetation, the vegetation in the Ali area is sparse, low and has seasonal variability. In addition, the wind is strong and the effect of vegetation on wind prevention and sand fixation is poor. The vegetation has little impact on the spatial distribution pattern of sandy desertification land in the Ali area.

At present, it is still controversial whether the main driving factor of land desertification in the Ali area is due to human factors or natural factors. It needs to be systematically studied and accurately quantified whether the intensity of human activities in this area exceeds the ecological carrying capacity, and to explore the contribution of human factors and natural factors to sandy desertification.