ArticlePDF Available

MODELING PROCESS OF THE SPATIAL-TEMPORAL CHANGES OF VEGETATION COVER AND ITS RELATIONSHIP WITH DRIVERS IN DRYLANDS AND WETLANDS IN XIANJIANG (CHINA)

June 2024

June 2024

Authors:

Omid Shobairi

Show all 8 authorsHide

Content uploaded by Samira Hemmati

Content may be subject to copyright.

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

MODELING PROCESS OF THE SPATIAL-TEMPORAL CHANGES OF

VEGETATION COVER AND ITS RELATIONSHIP WITH DRIVERS IN

DRYLANDS AND WETLANDS IN XIANJIANG (CHINA)

Seyed Omid Reza SHOBAIRI *, **, Lingxiao SUN *, ** (C.A.), Haiyan ZHANG *, **,

Chunlan LI *, **, Jing HE *,**, Behnam Asghari BEIRAMI ***,

Samira Hemmati ROUDBARI **** and Qorghizbek AYOMBEKOV *

* Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, 818 South Beijing Road,

Urumuqi, China, CN-830011, omidshobeyri214@gmail.com, ORCID: 0000-0002-6528-8653 (S. O. R.

S.); sunlx@ms.xjb.ac.cn, ORCID: 0000-0002-1702-5445 (L. S.); hyzhang@ms.xjb.ac.cn, ORCID:

0000-0002-5250-1865 (H. Z.); lichunlan@ms.xjb.ac.cn, ORCID: 0009-0001-9837-7556 (C. L.);

hejing@ms.xjb.ac.cn, ORCID: 0001-0001-7013-1266 (J. H.); ayombekqarghizbek@gmail.com,

ORCID: 0009-0002-2400-1481 (Q. A.).

** University of Chinese Academy of Sciences, 19 Yuquan Road, Beijing, CN-100049, China, CN-

100049, omidshobeyri214@gmail.com, ORCID: 0000-0002-6528-8653 (S. O. R. S.);

sunlx@ms.xjb.ac.cn, ORCID: 0000-0002-1702-5445 (L. S.); hyzhang@ms.xjb.ac.cn, ORCID: 0000-

0002-5250-1865 (H. Z.); lichunlan@ms.xjb.ac.cn, ORCID: 0009-0001-9837-7556 (C. L.);

hejing@ms.xjb.ac.cn, ORCID: 0001-0001-7013-1266 (J. H.).

*** K. N. Toosi University of Technology, 470 Mirdamad Ave. West, Tehran, Iran, IR-19697,

behnam.asghari1370@gmail.com, ORCID: 0000-0002-0314-1912 (B. A. B.).

**** University of Zanjan, Faculty of Agriculture, University Blvd., 45371-38791, Iran, IR-45371-

38791, ss.hemmati82.sh@gmail.com, ORCID: 0009-0000-0002-1702-0907, (S. H. R.).

DOI: 10.2478/trser-2024-0003

KEYWORDS: vegetation, NDVI, drivers, ecoregions, spatial distribution, OLS.

ABSTRACT

Findings reveal that the majority of studied areas are classified as bare lands, while the

lowest amount is covered by lichens and mosses. Grassland and cropland occupy major areas

of the region, with highest normalized difference vegetation index (NDVI) value saw in 2020,

showing dense vegetation in the western, northwestern and northern regions. Afforestation

efforts shown positive results, with a 4% increase in forested area between 2000 and 2022.

RÉSUMÉ: Processus de modélisation des changements spatio-temporels de la

couverture végétale et de sa relation avec les facteurs de changement dans les zones arides et

humides du Xianjiang (Chine).

Les résultats révèlent que la majorité des zones étudiées sont classées comme des

terres nues, tandis que la plus petite partie est couverte de lichens et de mousses. Les prairies et

les terres cultivées occupent une grande partie de la région, la valeur la plus élevée de l’indice

de végétation par différence normalisée (NDVI) étant observée en 2020, ce qui indique une

végétation dense dans les régions de l’ouest, du nord-ouest et du nord. Les efforts de

reboisement ont donné des résultats positifs, avec une augmentation de 4% de la superficie

forestière entre 2000 et 2022.

REZUMAT: Procesul de modelare a schimbărilor spațio-temporale ale acoperirii

vegetale și relația acesteia cu factori ai zonelor uscate și zonelor umede din Xianjiang (China).

Rezultatele arată că majoritatea ariilor studiate sunt clasificate ca terenuri goale, cea

mai mică proporție este acoperită de licheni și mușchi. Pajiștile și terenurile cultivate ocupă

arii mari ale regiunii, cu o valoare mare a indicelui de vegetație (NDVI) în 2020, indicând

vegetație densă în regiunile de vest, nord-vest și nord. Eforturile de împădurire au arătat

rezultate positive, cu o creștere cu 4% a suprafeței împădurite între 2000 și 2022.

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

INTRODUCTION

Vegetation plays a key role in Earth’s biosphere, influencing biogeochemical cycles,

carbon balance regulation, and climate stability (He et al., 2023; Zhuang et al., 2019; Jiang et

al., 2017). It also provides environmental protection, defence of soil from erosion, wind and

sand damage, and preventing desertification (Wang et al., 2023; Zhuang et al., 2019).

Additionally, vegetation contributes to the energy cycle by reducing greenhouse gas,

supporting global carbon balance, and ensuring climate stability (Zhuang et al., 2019). Thus, it

is vital to monitor changes in vegetation over time and understand the underlying mechanisms

to effectively manage ecological processes and protect the environment (Li et al., 2021).

Many studies conducted worldwide saw significant changes in vegetation across

various scales (Gadiga, 2015). These studies have highlighted the influence of both natural and

human factors, such as climate change, land use modifications, and ecological interventions on

the dynamics of vegetation patterns (Wang et al., 2023; Wang et al., 2022; Gadiga, 2015).

Natural factors exert long-term control, for instance, weather conditions determine the type of

vegetation, and alterations in rainfall deeply impact vegetation growth and health. On the other

hand, social factors play a significant role in short-term scales. Growth of urban economies and

rise in urban population contribute to the expansion of urban areas (Hu and Hu, 2019).

When a large-scale change in vegetation occurs in an area, it affects the energy

balance of the surface, which may cause heating or cooling effects (Han et al., 2022), land use

change is one of the main challenges that affects natural landscapes and has raised concerns

about sustainable development and food security (Hu and Nacun, 2018; Li et al., 2015). That is

why, countries and organizations around the world have accepted the risks of reducing

vegetation cover and considered measures for protection (Wang et al, 2023). IGBP

(International Geosphere-Biosphere Program) and IHDP (Global Change Human Factors

Program) proposed the research program “land use/change of cover” (LULC) in 1995 (Liu et

al., 2022). As one of the most active economies in the world, China has undergone many

changes in land use over the past decades, increasingly contributing to environmental crises

(Zhu et al, 2022; Yin et al, 2018). In the last fifty years, China has implemented some

measures to protect the Loess plateau, aiming to mitigate soil erosion and enhance ecological

conditions. Among these initiatives, the “project of restoring cultivated land to forests and

grasses” has emerged as a prominent ecological revitalization scheme. This project, initiated in

1999, has yielded ample improvements in vegetation quantity and quality on the Loess plateau

(Zhang et al., 2020). However previous studies have shown that between 1700 and 1950 there

was an increasing trend in arable land in China, although they show diverse magnitudes and

rates (Miao et al., 2016). Consequently, vegetation is a comprehensive indicator of

environmental change, and therefore, spatial and temporal changes in vegetation and its

response to climate change have become key issues in global change research (Xianfeng et al.,

2014). Xiao and Weng (2007) examined land use changes in southern China from 1991 to

2001, showing that agricultural land decreased and forest land increased slightly. In the study

of spatial-temporal dynamic changes in vegetation, the plant Normalized Difference index

(NDVI), is an important factor for assessing plant growth status and vegetation distribution

pattern (Duan et al., 2021). Gadiga (2015) studied the state and dynamics of vegetation in the

Mubi Region in Nigeria using the NDVI index, and the vegetation in the region has suffered

record degradation due to the growth of urbanization. Zhuang et al. (2020) conducted a study

investigating vegetation dynamics in China’s Xinjiang province from 1981 to 2018 using the

normalized vegetation difference index (NDVI). Their findings revealed that multiple factors,

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

including precipitation, influenced vegetation changes in the region. Areas with notable

vegetation dynamics were primarily concentrated on the northern and southern slopes of the

Tianshan Mountains, the Ely River Valley, and the Altai region. Meng et al. (2020) assessed

the temporal and spatial variations in the NDVI index in Mongolia and examined natural and

human factors influencing these changes. The results indicated major degradation in desert

steppe and the Gobi Desert within arid regions, while grassland steppe and Alpine steppe

exhibited significant upward trends. The researchers also observed that climatic parameters,

such as precipitation, had a positive impact on vegetation distribution. Overall, these studies

highlight the influence of factors such as precipitation on vegetation dynamics and underscore

the specific regions affected by these changes in China’s Xinjiang province and Mongolia.

Xinjiang Province, in northwestern China, has been susceptible to the effects of

climate change, experiencing warming over the last four decades. The region has witnessed

notable transformations in vegetation patterns during last years (Han et al., 2022; Zhuang et al.,

2019). Understanding the changes in vegetation cover is crucial in realising the temporal and

spatial dynamics of this region and the feedback between vegetation and the atmosphere.

This study aimed to analyze and model the detection of vegetation cover changes in

various areas of Xinjiang Province over a 23-year period. By harnessing the connectivity of

sensor datasets from remote sensing platforms, time series data was utilized to track these

changes. Examining the relationship between vegetation dynamics and climate and

environmental stimuli is vital in the context of global warming and climate change research, as

it provides insights into the dynamic responses of terrestrial ecosystems to these phenomena.

The study employed the ArcGIS platform to predict the impact of climate and environmental

factors on plant indicators, leveraging open-source data and coding techniques.

MATERIAL AND METHODS

Study area

Xinjiang Province is situated in northwestern China, spanning from 73°20’E to

96°25’E and 34°15’N to 49°10’N (Jiapaer et al., 2015). With a total land area of 1.66 million

km2, it accounts for approximately one-sixth of China’s land area (Yu et al., 2020). The

province encompasses a delicate ecological zone characterized by a complex arid environment,

with mountainous areas comprising 51.4% and plain areas comprising 48.6% of the total land

area (Luo et al., 2019; Jiapaer et al., 2015). Xinjiang exhibits diverse landforms, including the

Altai Mountains in the north, the Kunlun Mountains and A-erh-chin Mountains in the south,

and the Tian Shan Mountains spanning the central part of the province. These major mountain

ranges encircle the Junggar Basin and the Tarim Basin (Zhao et al., 2022), which serve as the

sources of numerous rivers (Luo et al., 2019). The Tian Shan Mountains, Altai Mountains, and

Kunlun Mountains harbor extensive forest and grassland vegetation, while oases and cities are

located in the valley plains. As of 2018, forest land and grassland accounted for 31.29% of the

land area, with grassland being the predominant vegetation type (Zhuang et al., 2020). The

Junggar and Tarim Basins are characterized by typical temperate desert vegetation. Despite

belonging to a temperate continental climate, Xinjiang exhibits distinct differences between its

northern and southern regions due to its extensive north-south span. The annual precipitation in

northern Xinjiang ranges from 100 to 500 mm, with average temperatures ranging from 4 to

8°C. In southern Xinjiang, annual precipitation ranges from 20 to 100 mm, while average

temperatures range from 10 to 13°C (Cao and Gao, 2022). Figure 1 provides a visual

representation of Xinjiang Province’s geographical location and its ecoregions.

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

Preparation of vegetation data from 2000 to 2023

Modeling the study is based on calculating satellite data. Calculations were completed

with the help of the Google Earth Engine (GEE) platform and catalogs of raster files on the

platform.Other open sources of data were used such as MODIS/061/myd13a1-NDVI (16 daily-

2000 to 2023 ‒ with a spatial resolution of 250 meters ‒ global scale, and data sets of drivers

such as daily LST (Kelvin), precipitation (mm/d), global land cover extracted from Sentinel-1

and 2 data and Global 30 Arc-Second Elevation to extract DEM and slope. Data through

ArcGIS platform was extracted to track vegetation cover changes, areas, and calculation of the

drivers affecting the vegetation. The main sources and characteristics of the data collected are

displayed in table 1.

Table 1: Main sources of data with the satellite products specifications.

Sattelite/

Model

Product

name

Date

name

Unit

Spatial

resolution

Temporal

resolution

Time

extent

Modis/

Terra

MOD11A1.061

Terra Land

Surface

Temperature

and Emissivity

Daily Global

1 km

LST_Day 1

LST_Night

1 km

Kelvin

1,000 m

Daily

2000-02-24

Modis/

Terra

Vegetation

Indices

NDVI

‒

500 m

Daily

2000-02-18

Modis/

Terra

Vegetation

Indices

NDVI

‒

500 m

Daily

2000-02-18

CHIRPS

CSB-

CHG/CHIRPS/

DAILY

Precipitation

mm/d

‒

Daily

1981-01-

01T00:00:00

Z–2023-06-

30T00:00:00

Sentinel-1

and 2 data

ESA World

Cover 2020

Global

Land

cover

product

‒

10 m

Annually

2020-2021

GTOPO30

Global 30 Arc-

Second

Elevation

DEM

1 km

NaN

1996

Ecoregion data

Ecoregions, defined as regions with similar ecosystems, have become increasingly

significant in the evaluation and administration of environmental matters. They offer a

comprehensive framework that allows for flexible and comparative analysis of intricate

environmental issues (Lovland and Merchant, 2004). We downloaded Ecoregion data from

https://ecoregions.appspot.com and cropped it for Xinjiang Province. There are 14 terrestrial

biomes and 846 ecoregions worldwide. Plant communities within a given biome may exhibit

comparable structural characteristics while harboring distinct species compositions. Table 2,

list the 17 ecoregions and five biomes in Xinjiang Province. Figure 1 shows the positions of

each of the ecoregions of Xianjian Province.

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Figure 1: Positions of each of the ecoregions in the Xianjiang, China.

Taklimakan Desert ecoregion with 742,657 km2 accounts for 18.1% of the province

and Tian Shan montane conifer forests ecoregion with 27,568 km2 covers 0.67% of the

province, representing the largest and smallest areas among the Xinjiang-Tian Shan ecoregions

(Tab. 2).

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

Table 2: Xinjiang ecoregions, classified by name, area, percentage.

Ecoregions

Area km2

Alashan Plateau semi-desert

674,352

16.498

Altai alpine meadow and tundra

90,434

2.212

Altai montane forest and forest steppe

142,875

3.495

Altai steppe and semi-desert

83,192

2.035

Central Tibetan Plateau alpine steppe

629,190

15.393

Emin Valley steppe

65,135

1.594

Junggar Basin semi-desert

304,938

7.460

Karakoram-West Tibetan Plateau alpine steppe

143,265

3.50

North Tibetan Plateau-Kunlun Mountains alpine desert

374,494

9.162

Pamir alpine desert and tundra

118,072

2.889

Qaidam Basin semi-desert

192,147

4.701

Rock and Ice

34,830

0.852

Taklimakan Desert

742,657

18.169

Tarim Basin deciduous forests and steppe

54,533

1.334

Tian Shan foothill arid steppe

129,231

3.162

Tian Shan montane conifer forests

275,68

0.674

Tian Shan montane steppe and meadows

280,611

6.865

Implementation phases

Tracking and analysis of the models took place in several phases: Identifying the

current state of the study border; Calculation and zoning of vegetation at the study boundary

(time series 2000 to 2023) and determination of maximum and minimum density; Analysis of

spatial-temporal distribution of vegetation since 2000 by classification method; Areas where

the value of NDVI is greater than 0.5 and vice versa areas where the value of NDVI is less than

0.5.; Estimation of vegetation density using kernel density and point density: 1) Monitoring and

evaluation of vegetation cover from the past 23 years using land surface measurements

(Divisions by municipalities), 2) Monitoring and evaluation of vegetation cover from 2000-

2023 using land surface measurements (Divisions by ecoregions); Estimation of the area and

percentage of afforestation and deforestation at the beginning and end of the desired time

period (2000-2022); Identifying the pattern of spatial distribution of vegetation in the study

border with spatial autocorrelation with a confidence level above 90%, Average Nearest

Neighbor; Estimation of hot spots and cold spots of vegetation in the border of the study with a

confidence level above 90% using hot spot analysis, clusters and outliers; Identification of

factors affecting the distribution of vegetation in the study border (land use, land surface

temperature, rainfall, slope, elevation); Spatial modeling of factors affecting vegetation (OLS).

RESULTS

Identifying the current state of the study border

Figure 2 shows patterns for the digital elevation model (A), slope (B), precipitation

parameter (C) and LULC variations calculated. According to the digital elevation model, the

area studied has expanded between 152 m and 7,553 m, which will create a wide range of plant

species, each plant specific to an elevation surface.

As much as altitude increase, precipitation increase, temperature decrease, dry stress

decrease, and the vegetation develops, this trend continues till higher altitudes where low

temperatures reduces the vegetation. The gradient variation pattern (Fig. 2B) also represents

gradient fluctuations of 0 to 356 degrees in the study area, which is very diverse. There is a

question of which slopes in the Xinjiang ecoregions have more vegetation?

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

For the northern hemisphere of the Earth, including Xinjiang, which is wide and rich

in diversity, the slopes facing south receive more sunlight and become drier and warmer,

support drought-resistant vegetation and are less favorable for trees, while the slopes facing

north retain moisture and are cold and humid and are suitable for humid plants.

The rain pattern is displayed in figure 2C. The highest rainfall was 704 mm and the

lowest was 13 mm. Precipitation is one of the most important factors affecting the growth and

development of vegetation. According to studies conducted in many Xinjiang ecosystems,

especially grasslands and arable land, the growth of seasonal plants occurs entirely within the

rainy season. In times of drought, vegetation grows poorly in these ecosystems. If it rains, the

sudden growth of vegetable plants actually turns. Unsure what is being communicated.

By manipulating the Sentinel-1 and 2 satellite data sets, tracking land use and land

cover changes, for 2020, LULC was considered as one of the drivers that reflected vegetation

in the ecoregions of Xinjiang. Figure 2D shows this classification visually, and table 3

examines the classification of LULC and their area and percentage.

A B

C D

Figure 2: Digital elevation model (A), slope (B), precipitation (C) and LULC (D).

Table 3 shows, the largest area is the Bare area class, (139,433.7 km2, or ying

61.49%), and the smallest area is the Lichens and Mosses class (41.8 km2, or 0.018%). Bare

land refers to areas where dominant vegetation covers under 90%. This category contains areas

without artificial cover and includes regions with less than 4% vegetative cover. Examples

include bare rock regions, sandy areas, and deserts. Table 3 shows that mosses and lichens has

the smallest area, however it is argued that it plays a very important role in study area. Both

mosses and lichens have a significant role in maintaining a healthy environment as they

contribute to the absorption of carbon dioxide and other air pollutants. Lichens, in particular,

possess valuable detoxifying properties that make them highly beneficial to various organisms.

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

Table 3: Classification of LULC alignments, area and percentage.

LULC

Area

km2

Crop land

16,263.74

7.17

Herbaceous cover

13,615.37

Tree cover

2,038.06

0.89

Tree and shrub

1,989.67

0.87

Grassland

48,112

21.2

Lichens and mosses

41.8

0.018

Urban areas

261.83

0.11

Bare areas

139,433.7

61.49

Water bodies

1,306.74

0.57

Snow and ice

3,692.84

1.62

Table 3 is divided into valuable LULC classes and their respective area and percent

coverage of the study area. These classes together create an incredibly diverse and complex

BioCycle and any changes in urban area, cropland and bare area classes will greatly affect the

quality and quantity of valuable classes.

Calculation and zoning of vegetation at the study boundary in time series 2000 to

2023 and determination of maximum and minimum density

Satellites carry numerous sensors that measure the reflection of red and near-infrared

light waves from land surfaces. By utilizing mathematical algorithms, scientists convert the

raw satellite data related to these light waves into vegetation indices. These indices serve as

indicators of vegetation greenness, representing the relative density and health of vegetation

for each pixel in a satellite image. Among various vegetation indices, the Normalized

Difference Vegetation Index (NDVI) is widely employed. NDVI values range from +1.0 to

‒1.0. Barren areas such as rocks, sand, or snow typically exhibit very low NDVI values, often

0.1 or less. Regions with sparse vegetation like shrubs, grasslands, or aging crops show

moderate NDVI values ranging from approximately 0.2 to 0.5. High NDVI values, around 0.6

to 0.9, correspond to dense vegetation found in temperate and tropical forests or crops during

their peak growth stage (Tahir et al., 2019; Elsu et al., 2017). Researchers can utilize raw

satellite data to convert it into NDVI values, which enables the creation of images and other

products that provide a general assessment of vegetation characteristics, quantity, and

condition on a global scale. NDVI is particularly valuable for monitoring vegetation at

continental to global levels, as it can account for variations in lighting conditions, surface

slope, and viewing angles. However, it should be noted that NDVI may become saturated in

areas with dense vegetation and is influenced by the color of the underlying soil. By averaging

NDVI values over time, researchers can establish a baseline for normal growing conditions in

a specific region during a particular time of year. Subsequent analysis can then evaluate the

relative health of vegetation compared to this baseline. Over time, NDVI analysis can identify

areas where vegetation is flourishing or experiencing stress, as well as detect changes in

vegetation patterns resulting from human activities like deforestation, natural disturbances

such as wildfires, or shifts in plants’ phenological stages (*).

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Figure 3 shows the visual patterns of vegetation classification calculated with

the average NDVI value for the time series 2000 to 2023. The highest value of NDVI is for

2020 (Fig. 3E), with a value of 0.933 predicted, and according to the classification of NDVI

value balances; it represents areas with dense vegetation. The lowest value of -0.199

corresponds to 2010 (Fig. 3C), representing areas with the presence of abundant sand, rocks,

and snow.

Based on the NDVI value patterns depicted in figure 3, the western, northwestern, and

northern areas of Xinjiang Province exhibit the highest NDVI values. These regions

encompass various ecoregions, including the arid steppe of the Tian Shan foothills, conifer

forests of the Tian Shan Mountains, steppe and meadows of the Tian Shan montane region,

steppe in the Emin Valley, steppe and semi-desert in Altai, forest and forest steppe in Altai

Mountains, and alpine meadow and tundra in Altai.

The average value of NDVI is estimated to be approximately between 0.2 and 0.5

which are spread in the central, western and northwestern and southwestern regions, are also

broadcasted in the form of spots in the eastern regions of Xinjiang Province, which this type of

classification represents scattered vegetation, savanna, shrubs, meadows, and agricultural

lands. The classification of certain areas within the Tian Shan montane steppe and meadow

ecoregion serves as a notable illustration of this categorization.

Ecoregions like Taklimakan Desert, north Tibetan Plateau-Kunlun mountains

alpine desert and junggar basin semi desert positioned in the central, central south,

southern and northern regions of Xinjiang Province respectively experienced the lowest

value of NDVI, and are covered with very scattered vegetation, sand, barren rock and rarely

snow (Fig. 3).

Analysis of spatial-temporal distribution of vegetation 2000-2023 by classification

method (Areas with NDVI > 0.5 and NDVI < 0.5).

The temporal and spatial distribution of the average annual NDVI value

was examined. This classification specifically identifies regions where the NDVI value

exceeds 0.5, and regions where the NDVI value is below 0.5, based on data from 2000 to 2023

(Fig. 5). 2015 had the highest area of the regions, with a NDVI value greater than 0.5 in

Xinjiang Province, which has an area of 149,450.6 km2 (Tab. 4, Fig. 5D). 2000 also

experienced the largest area with a NDVI less than 0.5, this value includes 1,614,503.6 km2

(Tab. 4, Fig. 5A). In summary, the range of NDVI values spans from ‒1.0 to 1.0. Negative

values represent clouds and water, values close to zero indicate bare soil, and higher positive

values of NDVI correspond to varying levels of vegetation density, ranging from sparse

vegetation (0.1-0.5) to dense green vegetation (0.6 and above). 2015 has experienced the

highest level of dense vegetation, and 2000 is also covered with a sporadic and vast vegetation

community.

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

A B

C D

E F

Figure 3: Visual patterns of temporal and spatial changes of vegetation

calculated by the mean value of NDVI for the time series 2000 to 2023.

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Table 4: Statistics on area in km2 from 2000 to 2023.

Year

2000

2005

2010

2015

2020

2023

Area (km2) of arenas where

the value of the NDVI < 0.5

1614503.61

1600279.06

1587131.76

1574324.40

1583306.92

1578566.28

Area (km2) of arenas where

the value of the NDVI > 0.5

109396.39

123345.94

136518.24

149450.60

142168.08

146908.72

Figure 4 describes changes in boundary areas from 2000-2023. Figure 4A clearly

shows a decrease in area for arenas with a NDVI of less than 0.5. On the other hand figure 4B

shows a unique phenomenon, where area of areas with NDVI more than 0.5 is increasing,

indicating the existence of suitable conditions for vegetation development as well as correct

management of the arenas. The increasing quantity of positive NDVI values signifies a rise in

the presence of green vegetation. What are the factors that affect NDVI? NDVI is an important

indicator that reflects the state of the regional climate and the environment, which is affected

by precipitation, temperature, the water content of the soil, humans, etc. For instance, Yao et

al. (2023) conducted a study utilizing GeoDetector to investigate the impact of various factors

on NDVI, such as climate, soil, topography, and human activities. Their findings revealed that

the primary contributing factors were night light brightness (51.9%), annual average air

temperature (47%), and annual average atmospheric pressure (45.8%). Additional research has

demonstrated that NDVI exhibits a stronger correlation with changes in rainfall (R = 0.75)

compared to temperature and elevation. Generally, NDVI values tend to decrease under

extremely high and low temperature conditions. Furthermore, the growth pattern of NDVI

varies with altitude, with a more pronounced response observed between 70 m and 1,500 m

above sea level. These aspects will be further examined in the subsequent sections of this

article, exploring the relationship between NDVI and rainfall, land use and land cover changes,

elevation, slope, and LST, as well as modeling the interconnectedness of these factors.

Figure 4: Changes in the area of studied boundaries (km2) where the NDVI value is less

than 0.5.(A) and NDVI value is more than 0.5 (B).

Figure 5 reports visual patterns of NDVI value, for the northern regions of Xinjiang

such as Altai Mountains forest and forest steppe, Altai alpine meadow and tundra, parts of the

Emin Valley steppe and large parts of the Tian Shan ecoregions. Likewise it reports the

montane steppe meadows, the montane conifer forest and the foothill arid steppe and the

central parts of this province are seen as spots that have experienced NDVI values >5 from

2000 to 2023. This value reached its highest value in 2015 (Tab. 4, Fig. 4B), After a hiatus in

2020, it increased again until 2023. Areas that had NDVI values <5 and have scattered

vegetation, bare land, sand and agricultural land (displayed in red), with the exception of 2020,

have experienced a downward trend (Fig. 4A). Overall the very positive ecological event is

taking shape and every year we see an increase in conservation, restoration, and development

of vegetation coverage in all parts of the Xinjiang Province. This constructive process is more

prominent in central and western parts such as the Tian Shan Mountains and northern areas

such as the Altai Mountains and the Emin Valley (Fig. 5).

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

A B

C D

E F

Figure 5: Visual patterns of NDVI temporal and spatial distribution along with area of arenas

with a value greater than 0.5 and less than 0.5 for 2000 to 2023.

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Estimation of vegetation density using kernel density and point density

monitoring and evaluation of vegetation cover from 2000-2023 using land surface

measurements (Divisions by municipalities)

Figure 6 shows temporal and spatial patterns of monitoring and evaluating vegetation

changes using land surface measurements, which are borders of counties, and have modeled

changes since 2000. The NDVI value is an average of the total months of the year. The

modeling process relies on assessing vegetation density and leverages the kernel density

method. This approach involves calculating the density of point features surrounding each

output raster cell. Essentially, a smooth curved surface is fitted over each point in the model.

Kernel density estimation, a statistical technique, is utilized for estimating the probability

density function of a random variable using kernel smoothing. In this case, the variable of

interest is the annual average NDVI value.

Kernel estimator results argue that the maximum values close to NDVI >0.55 with

green, and the minimal values are close to NDVI <0.07, highlighted in red. The range between

maximum and minimum is aligned with yellow and orange colors (Fig. 6). In general, the

northern and western counties of Xinjiang Province have experienced the development of

vegetation, in other words the southern and eastern counties have lost the harmony of warm

red and orange colors and turned into pale yellow and green colors (Fig. 6).

According to table 5, the highest average NDVI value is related to Ili Kazakh County,

which experienced a value of 0.54 in 2000. The lowest average NDVI value of 0.07 also

belongs to Khotan County.

According to NASA Earth Observatory website (***), extremely low NDVI values

(0.1 and below) indicate barren areas such as rock, sand, or snow. Moderate values (0.2 to 0.3)

correspond to shrub and grassland, while high values (0.6 to 0.8) signify temperate and tropical

rainforests. Applying this principle, it can be observed that Khotan County predominantly

consists of very sparse vegetation, desert, sand, and rock, whereas Ili Kazakh County and

Shihezi County exhibited a relatively moderate vegetation situation in 2000, characterized by

grasslands, shrubs, as well as broadleaf and coniferous species (Fig. 7).

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

A B

C D

E F

Figure 6: Examination of spatial and temporal patterns of vegetation density

using Kernel density estimation.

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Table 5: Classification of vegetation cover (NDVI) based on county divisions in two.

No.

Name

Minimum

Maximum

Mean

Aksu

‒0.052921

0.691787

0.136181

Altay

‒0.123229

0.859978

0.270487

Bayin’gholin

‒0.191100

0.878285

0.104023

Bortala

‒0.171962

0.874815

0.290748

Changji Hui

‒0.033705

0.803515

0.233974

Hami

‒0.027253

0.678575

0.088444

Ili Kazakh

‒0.043728

0.885946

0.54512

Karamay

0.03571

0.541906

0.124138

Kashgar

‒0.075600

0.536531

0.118364

Khotan

‒0.117503

0.677344

0.070696

Kizilsu

‒0.054392

0.547

0.137629

Shihezi

0.201837

0.640216

0.484435

Tacheng

‒0.156277

0.819484

0.253239

Turfan

‒0.038627

0.58874

0.082853

Urümqi

‒0.038954

0.724495

0.307904

Figure 7: Statistics of average annual NDVI value

based on the division of county boundaries in 2000.

Tables 5-10 as well as figures 7-12 show the NDVI values of the cities of Xinjiang

Province 2000-2003. The results showed that Ili Kazakh city and then Shihezi city had

the highest NDVI value during the period studied. The lowest NDVI value was for Khotan

city. Examining the NDVI index values in Ili Kazakh city 2000 to 2023 showed that the

amount of vegetation has grown significantly and improved. In this city, the NDVI value in

2000 was 0.54, which gradually increased in 2005, 2010, and 2015 (0.55, 0.57, and 0.57,

respectively).

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

Table 6: Classification of vegetation cover (NDVI) based on couty divisions in 2005.

No.

Name

Minimum

Maximum

Mean

Aksu

‒0.122748

0.742965

0.148695

Altay

‒0.192430

0.927757

0.279236

Bayin’gholin

‒0.184302

0.869611

0.10498

Bortala

‒0.155703

0.875777

0.307521

Changji Hui

‒0.027991

0.773594

0.229612

Hami

‒0.073582

0.669499

0.088761

Ili Kazakh

‒0.046839

0.88735

0.555844

Karamay

‒0.042529

0.670178

0.169616

Kashgar

‒0.175655

0.563468

0.135556

Khotan

‒0.074800

0.60319

0.079338

Kizilsu Kirghiz

‒0.055451

0.5445

0.1639

Shihezi

0.170395

0.704664

0.521317

Tacheng

‒0.048387

0.808383

0.279653

Turfan

0.011559

0.578973

0.084733

Urümqi

‒0.030313

0.695617

0.269192

Figure 8: Statistics of average annual NDVI value

based on the division of county boundaries in 2005.

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Table 7: Classification of vegetation cover (NDVI) based on couty divisions in 2010.

No.

Name

Minimum

Maximum

Mean

Aksu

‒0.065500

0.687213

0.160299

Altay

‒0.116237

0.864199

0.286228

Bayin’gholin Mongol

‒0.187300

0.869814

0.111299

Bortala Mongol

‒0.190533

0.820894

0.315226

Changji Hui

‒0.023896

0.81202

0.274984

Hami

‒0.072722

0.68649

0.099629

Ili Kazakh

‒0.033036

0.878573

0.570325

Karamay

‒0.025764

0.712739

0.204186

Kashgar

‒0.080600

0.684593

0.143876

Khotan

‒0.074623

0.694637

0.081276

Kizilsu Kirghiz

‒0.064005

0.662137

0.156734

Shihezi

0.261811

0.7112

0.56347

Tacheng

‒0.128363

0.807561

0.305837

Turfan

0.011221

0.593036

0.086799

Urümqi

‒0.033983

0.75133

0.304136

Figure 9: Statistics of average annual NDVI value based on the division of county boundaries in 2010.

Table 8: Classification of vegetation cover (NDVI) based on couty divisions in 2015.

No.

Name

Minimum

Maximum

Mean

Aksu

‒0.060800

0.666599

0.167056

Altay

‒0.199042

0.866951

0.279071

Bayin’gholin Mongol

‒0.199100

0.857465

0.111562

Bortala Mongol

‒0.150868

0.83892

0.337688

Changji Hui

‒0.038100

0.841333

0.271983

Hami

‒0.085294

0.737637

0.11500

Ili Kazakh

‒0.041731

0.877921

0.579145

Karamay

0.007897

0.780512

0.210299

Kashgar

‒0.086180

0.682347

0.151257

Khotan

‒0.080100

0.592517

0.079768

Kizilsu Kirghiz

‒0.056029

0.521963

0.152814

Shihezi

0.19053

0.760908

0.510827

Tacheng

‒0.189183

0.830494

0.30230

Turfan

0.005873

0.629756

0.096969

Urümqi

‒0.049939

0.726139

0.30828

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

Figure 10: Statistics of average annual NDVI value

based on the division of county boundaries in 2015.

Table 9: Classification of vegetation cover (NDVI) based on couty divisions in 2020.

No.

Name

Minimum

Maximum

Mean

Aksu

‒0.069200

0.779886

0.18251

Altay

‒0.141897

0.848825

0.27150

Bayin’gholin Mongol

‒0.127306

0.822712

0.115179

Bortala Mongol

‒0.135692

0.933632

0.330082

Changji Hui

‒0.019654

0.797883

0.246454

Hami

‒0.030933

0.652469

0.091052

Ili Kazakh

‒0.039898

0.865697

0.541766

Karamay

0.036021

0.781246

0.211123

Kashgar

‒0.057445

0.646838

0.160216

Khotan

‒0.076941

0.614853

0.08908

Kizilsu Kirghiz

‒0.051024

0.52605

0.166212

Shihezi

0.134233

0.817762

0.5278

Tacheng

‒0.043740

0.825483

0.30307

Turfan

0.013841

0.592871

0.0888

Urümqi

‒0.033376

0.67819

0.267829

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Figure 11: Statistics of average annual NDVI value

based on the division of county boundaries in 2020.

Table 10: Classification of vegetation cover (NDVI) based on couty divisions in 2023.

No.

Name

Minimum

Maximum

Mean

Aksu

‒0.060800

0.736774

0.178739

Altay

‒0.182885

0.83864

0.271577

Bayin’gholin Mongol

‒0.162868

0.833813

0.113792

Bortala Mongol

‒0.177081

0.847629

0.3529

Changji Hui

‒0.009055

0.800372

0.24254

Hami

‒0.058832

0.640684

0.08512

Ili Kazakh

‒0.037388

0.87388

0.579248

Karamay

0.038863

0.790834

0.210868

Kashgar

‒0.076800

0.651084

0.16133

Khotan

‒0.076644

0.558047

0.08468

Kizilsu Kirghiz

‒0.052250

0.661522

0.166611

Shihezi

0.136324

0.814858

0.5312

Tacheng

‒0.039781

0.841734

0.29065

Turfan

0.019744

0.576473

0.088044

Urümqi

‒0.011569

0.687454

0.264764

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

Figure 12: Statistics of average annual NDVI value

based on the division of county boundaries in 2023.

However, in 2020, the NDVI value in Ili Kazakh city area decreased slightly

(0.54), then increased again in 2023 (0.58). In Shihezi city, this significant trend of

vegetation growth was observed, with average NDVI value reaching from 0.48 in 2000 to 0.53

in 2023.

This growth trend was seen NDVI index values for Khotan County, which had the

lowest NDVI rate among the different counties of Xinjiang Province. According to tables 5-10,

as well as figures 7-12, the NDVI value in Khotan County in 2000 and 2005 averaged 0.07,

which gradually reached 0.08 in 2010, 2015, 2020 and 2023.

According to the information provided on the NASA Earth Observatory website (***),

it is stated that low NDVI values (0.1 and below) indicate barren areas consisting of rock,

sand, or snow. Moderate values (0.2 to 0.3) suggest the presence of shrubs and meadows,

while high values (0.6 to 0.8) are associated with temperate and tropical rainforests. Applying

this principle, it can be argued that the region of Khotan has minimal vegetation,

predominantly comprising desert, sand, and rock. On the other hand, the cities of Ili Kazakh

and Shihezi exhibit a relatively moderate level of vegetation, including meadows, shrubs, as

well as broadleaf and coniferous species.

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Monitoring and evaluation of vegetation cover from 2000-2023 using land surface

measurements (Divisions by ecoregions)

Figure 13 shows the pattern of spatial and temporal changes in vegetation cover in the

ecoregions of Xinjiang Province during 2000-2023 using the kernel density function.

A B

C D

E F

Figure 13: Patterns of temporal and spatial changes in vegetation cover

using kernel density estimators in the borders of the ecorigens Xinjiang Province.

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

Most of the Taklimakan Desert is completely devoid of vegetation, but hardy plants

such as Halogeton, Halostachys, and camelthorn grow in its foothills. Although the value of

NDVI in this ecoregion is low, it has had a growing trend during the studied years, the value of

NDVI increased from 0.08 in 2000 to 0.09 in 2005 and 2010 and gradually increased to 0.1 in

2015, 2020 and 2023 (Tabs. 11-16, Fig. 13).

This growing trend in the amount of vegetation was also observed in other ecoregions

that have thin and scattered vegetation. Among them are the ecoregions of North Tibetan

Plateau-Kunlun Mountains alpine, Junggar Basin semi-desert, Tarim Basin deciduous forests

and steppe, and Pamir alpine desert and tundra, respectively in the southern, northern, central

and southwest of Xinjiang Province. Therefore, in these ecoregions, the vegetation is sparse

and the NDVI is low. However, in these ecoregions, the amount of vegetation has improved

since 2000, with mean NDVI levels going from 0.08, 0.15, 0.08 and 0.11 in 2000 to 0.10, 0.18,

0.12 and 0.13 respectively in 2023.

Table 11: Classification of vegetation cover (NDVI) based on ecoregions divisions in

2000.

Name

Minimum

Maximum

Mean

Area of

arenas

with NDVI

classify-

cation

(km2)

Alashan Plateau semi-desert

0.020716

0.167403

0.066595

1.549515

Altai alpine meadow and tundra

‒0.041297

0.8356

0.447685

1.847604

Altai montane forest and forest steppe

0.110255

0.859978

0.586297

2.038968

Altai steppe and semi-desert

0.24145

0.799791

0.521414

0.254039

Central Tibetan Plateau alpine steppe

0.0532

0.0789

0.069438

0.015808

Emin Valley steppe

0.055365

0.819484

0.290983

5.343731

Junggar Basin semi-desert

‒0.171962

0.874815

0.152806

27.366403

Karakoram-West Tibetan Plateau alpine steppe

‒0.062345

0.218932

0.031641

0.779973

North Tibetan Plateau-Kunlun Mountains alpine desert

‒0.164422

0.513398

0.086163

19.851451

Pamir alpine desert and tundra

‒0.054019

0.424546

0.119441

3.417256

Qaidam Basin semi-desert

‒0.191100

0.278864

0.080643

2.209025

Rock and Ice

‒0.075600

0.269455

‒0.012968

1.099327

Taklimakan Desert

‒0.181898

0.793929

0.086815

76.751566

Tarim Basin deciduous forests and steppe

0.016457

0.61935

0.089074

5.763608

Tian Shan foothill arid steppe

0.10338

0.789768

0.439169

0.972174

Tian Shan montane conifer forests

0.086819

0.885946

0.530611

1.425375

Tian Shan montane steppe and meadows

‒0.093751

0.875281

0.353765

21.228148

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Table 12: Classification of vegetation cover (NDVI) based on ecoregions divisions in

2005.

Name

Minimum

Maximum

Mean

Area of

arenas

with NDVI

classify-

cation

(km2)

Alashan Plateau semi-desert

0.024269

0.137483

0.063685

1.549515

Altai alpine meadow and tundra

‒0.035260

0.83096

0.451796

1.847604

Altai montane forest and forest steppe

0.10491

0.863939

0.57805

2.038968

Altai steppe and semi-desert

0.267149

0.808383

0.544267

0.254039

Central Tibetan Plateau alpine steppe

0.0578

0.0776

0.069378

0.015808

Emin Valley steppe

0.064729

0.80623

0.317591

5.343731

Junggar Basin semi-desert

‒0.192430

0.927757

0.168551

27.365448

Karakoram-West Tibetan Plateau alpine steppe

‒0.064121

0.26982

0.026307

0.779973

North Tibetan Plateau-Kunlun Mountains alpine desert

‒0.184302

0.550939

0.094414

19.846451

Pamir alpine desert and tundra

‒0.054584

0.422466

0.13886

3.417256

Qaidam Basin semi-desert

0.171300

0.26967

0.082013

2.201525

Rock and Ice

‒0.074800

0.344393

‒0.021671

1.099327

Taklimakan Desert

‒0.175655

0.78328

0.095594

76.756566

Tarim Basin deciduous forests and steppe

‒0.122748

0.742965

0.101872

5.763608

Tian Shan foothill arid steppe

0.128404

0.761067

0.499164

0.972174

Tian Shan montane conifer forests

0.042952

0.872852

0.492354

1.425375

Tian Shan montane steppe and meadows

‒0.108059

0.88735

0.342893

21.209103

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

Table 13: Classification of vegetation cover (NDVI) based on ecoregions divisions in

2010.

Name

Minimum

Maximum

Mean

Area of

arenas with

NDVI

classify-

cation

(km2)

Alashan Plateau semi-desert

0.0251

0.156013

0.069917

1.549515

Altai alpine meadow and tundra

‒0.030100

0.8483

0.447748

1.847604

Altai montane forest and forest steppe

0.153305

0.864199

0.567695

2.038968

Altai steppe and semi-desert

0.238124

0.807561

0.587318

0.254039

Central Tibetan Plateau alpine steppe

0.066

0.0851

0.077617

0.015808

Emin Valley steppe

0.067906

0.798935

0.335793

5.343731

Junggar Basin semi-desert

‒0.128363

0.820894

0.188803

27.378903

Karakoram-West Tibetan Plateau alpine steppe

‒0.065062

0.195633

0.025916

0.779973

North Tibetan Plateau-Kunlun Mountains alpine

desert

‒0.163778

0.537248

0.102822

19.848951

Pamir alpine desert and tundra

‒0.064005

0.415137

0.12624

3.417256

Qaidam Basin semi-desert

‒0.187300

0.348592

0.098393

2.186525

Rock and Ice

‒0.080600

0.261311

‒0.019348

1.099327

Taklimakan Desert

‒0.125400

0.7671

0.099849

76.726566

Tarim Basin deciduous forests and steppe

‒0.011591

0.65088

0.101618

5.763608

Tian Shan foothill arid steppe

0.132332

0.795914

0.516423

0.972174

Tian Shan montane conifer forests

0.075842

0.876967

0.548598

1.425375

Tian Shan montane steppe and meadows

‒0.190533

‒0.878573

0.365406

21.240648

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Table 14: Classification of vegetation cover (NDVI) based on ecoregions divisions in

2015.

Name

Minimum

Maximum

Mean

Area of

arenas

with

NDVI

classify-

cation

(km2)

Alashan Plateau semi-desert

0.02805

0.152062

0.073175

1.549515

Altai alpine meadow and tundra

‒0.014300

0.849298

0.46039

1.847604

Altai montane forest and forest steppe

0.1477

0.866951

0.571273

2.038968

Altai steppe and semi-desert

0.2203

0.808246

0.576432

0.254039

Central Tibetan Plateau alpine steppe

0.0643

0.0867

0.074508

0.015808

Emin Valley steppe

0.065934

0.830494

0.333953

5.343731

Junggar Basin semi-desert

‒0.199042

0.83892

0.187073

27.371403

Karakoram-West Tibetan Plateau alpine steppe

‒0.064099

0.218998

0.031506

0.779973

North Tibetan Plateau-Kunlun Mountains alpine desert

‒0.199100

0.544989

0.098162

19.856451

Pamir alpine desert and tundra

‒0.045398

0.469946

0.137724

3.417256

Qaidam Basin semi-desert

‒0.185452

0.296632

0.091435

2.201525

Rock and Ice

‒0.086180

0.290036

0.017878

1.099327

Taklimakan Desert

‒0.167808

0.7579

0.10416

76.734066

Tarim Basin deciduous forests and steppe

0.030569

0.655555

0.111113

5.763608

Tian Shan foothill arid steppe

0.109939

0.825393

0.53483

0.972174

Tian Shan montane conifer forests

0.04382

0.858642

0.5565

1.425375

Tian Shan montane steppe and meadows

‒0.085294

0.877921

0.368744

21.230648

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

The distinct topographical features found in the Tianshan and Altai mountains have

resulted in the presence of abundant grassland and forest vegetation. Consequently, the

ecoregions situated in this region, namely Tian Shan foothill arid steppe, Altai montane forest

and forest steppe, Altai steppe and semi-desert, Tian Shan montane conifer forests, Altai alpine

meadow and tundra, and Emin Valley steppe, exhibit substantial vegetation with high NDVI

values. As can be seen, the revitalization and development of vegetation in these ecoregions is

more prominent during the period under study. Only in 2020, NDVI decreased in these

ecoregions, but it again has an upward trend and normal vegetation has increased. This

revitalization and development of vegetation is more prominent in the Tian Shan foothill arid

steppe ecoregion, so that its NDVI has increased from 0.48 in 2020 to 0.56 in 2023 (Fig. 13,

Tabs. 15 and 16).

Table 15: Vegetation cover classification (NDVI) based on ecoregions divisions (2020).

Name

Minimum

Maximum

Mean

Area of

arenas

with

NDVI

classify-

cation

(km2)

Alashan Plateau semi-desert

0.031

0.129193

0.066389

1.549515

Altai alpine meadow and tundra

‒0.020200

0.8201

0.425986

1.847604

Altai montane forest and forest steppe

0.11874

0.848825

0.54862

2.038968

Altai steppe and semi-desert

0.20505

0.76781

0.560979

0.254039

Central Tibetan Plateau alpine steppe

0.0695

0.0887

0.080397

0.015808

Emin Valley steppe

0.068533

0.819078

0.324058

5.343731

Junggar Basin semi-desert

‒0.141897

0.825483

0.184406

27.421403

Karakoram-West Tibetan Plateau alpine steppe

‒0.042049

0.190949

0.036133

0.779973

North Tibetan Plateau-Kunlun Mountains alpine desert

‒0.074712

0.614853

0.11043

19.873951

Pamir alpine desert and tundra

‒0.051024

0.405781

0.13389

3.417256

Qaidam Basin semi-desert

‒0.092845

0.535508

0.102987

2.241525

Rock and Ice

‒0.076941

0.259547

‒0.009148

1.099327

Taklimakan Desert

‒0.127306

0.779886

0.107867

76.786566

Tarim Basin deciduous forests and steppe

0.034547

0.728055

0.124142

5.763608

Tian Shan foothill arid steppe

0.142023

0.859359

0.483168

0.972174

Tian Shan montane conifer forests

0.044904

0.845019

0.488253

1.425375

Tian Shan montane steppe and meadows

‒0.102500

0.933632

0.34722

21.240648

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Table 16: Vegetation cover classification (NDVI) based on ecoregions divisions

(2023).

Name

Minimum

Maximum

Mean

Area of

arenas

with

NDVI

classify-

cation

(km2)

Alashan Plateau semi-desert

0.031885

0.112891

0.06282

1.549515

Altai alpine meadow and tundra

‒0.017536

0.813509

0.423099

1.847604

Altai montane forest and forest steppe

0.122865

0.83864

0.54210

2.038968

Altai steppe and semi-desert

0.21795

0.778232

0.53419

0.254039

Central Tibetan Plateau alpine steppe

0.0655

0.0856

0.078207

0.015808

Emin Valley steppe

0.060161

0.828862

0.301536

5.343731

Junggar Basin semi-desert

‒0.182885

0.841734

0.18347

27.421403

Karakoram-West Tibetan Plateau alpine steppe

‒0.047465

0.218858

0.046283

0.779973

North Tibetan Plateau-Kunlun Mountains alpine desert

‒0.135032

0.557652

0.106581

19.873951

Pamir alpine desert and tundra

‒0.057781

0.411547

0.13789

3.417256

Qaidam Basin semi-desert

‒0.162868

0.353611

0.094669

2.241525

Rock and Ice

‒0.076800

0.356202

‒0.007643

1.099327

Taklimakan Desert

‒0.145283

0.760485

0.106296

76.786566

Tarim Basin deciduous forests and steppe

0.031389

0.733905

0.12035

5.763608

Tian Shan foothill arid steppe

0.171138

0.870284

0.563711

0.972174

Tian Shan montane conifer forests

0.049584

0.864204

0.48319

1.425375

Tian Shan montane steppe and meadows

‒0.060800

0.87388

0.353977

21.240648

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

Figure 14: Column chart of average NDVI values (annual)

By ecoregion boundaries from 2000, 2005, 2010, 2015, 2020 and 2023.

Tables 11-16 as well as figure 14 show the NDVI values of different eco-origins

during the studied years. As can be seen, the ecoregions located in the range of Tianshan and

Altai mountains have the highest NDVI. According to the NDVI values in the mentioned

tables, it can be said that there have been many changes in the vegetation of the region during

different years. In 2000 and 2005, the Altai montane forest and forest steppe ecoregions had

the highest NDVI values of 0.58 and 0.57 among the Xinjiang ecoregions, respectively. The

highest NDVI value in 2010, 2015 and 2020 with values of 0.58, 0.57 and 0.56 belonged to

Altai steppe and semi-desert ecoregion and in 2023 with value of 0.56 belonged to Tian Shan

foothill arid steppe ecoregion.

Estimation of the area and percentage of afforestation and deforestation (2000 to

2022)

Changes in land use and land cover have significant impacts on ecology, the

environment, and politics at both global and regional levels. Therefore, land cover monitoring

and modeling is important in environmental planning and management. Figure 15 shows the

spatial distribution of deforestation and afforestation in Xinjiang Province (Figure 15A),

ecoregions (Fig. 15B) and counties of Xinjiang Province (Fig. 15C) from 2000 to 2022.

Additionally, the table show the changes in vegetation cover in Xinjiang Province

between 2000 and 2022. Specifically, when examining the vegetation in the northern regions

of Xinjiang Province, such as Altai montane forest and forest steppe, Altai steppe and semi-

desert, Emin Valley steppe, and Tian Shan montane steppe and meadows ecoregions, no

significant alterations have been observed. These regions collectively encompass an average

area of approximately 1.665 million km² within Xinjiang Province. According to figure 15 and

tables 17-18, most parts of Xinjiang Province lack vegetation (1,403,680.527 km2). In figure

15, the areas without vegetation are highlighted with a yellow color, mainly representing the

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

southern and central regions of Xinjiang Province. These areas include the Taklimakan Desert,

North Tibetan Plateau-Kunlun Mountains alpine desert, Central Tibetan Plateau alpine steppe,

Karakoram-West Tibetan Plateau, as well as alpine steppe and Pamir alpine desert and tundra.

It is worth noting that there has been no significant change in the vegetation status of these

regions. From 2000 to 2022, approximately one percent of the land area in Xinjiang Province

has experienced vegetation degradation.

Figure 15: Visual patterns of forestry and deforestation, their area and percentage from 2000 to

2022 (Classification based on border of Xinjiang Province (A), the borders of eco-regions ( B),

the borders of the counties (C), the pie chart of the percentage changes in the area of

afforestation and deforestation activities and other changes (D).

Shobairi S. O. R. et al.– Modeling process of the changes vegetation cover in in Xianjiang (China) (27 ~ 74)

Deforestation and vegetation destruction have primarily occurred within the

ecoregions of Tian Shan montane steppe and meadows, Emin Valley steppe, Tian Shan

montane conifer forests, and Altai alpine meadow and tundra, covering an area of

21,749.34084 km2. However, due to protective measures and efforts towards vegetation

revitalization and development, significant positive ecological changes have taken place in

various parts of Xinjiang Province. Notably, in an area of 59,743.718885 km2, approximately

4% of the total area of Xinjiang Province, predominantly in the Junggar Basin semi-desert,

Tian Shan foothill arid steppe, Tian Shan montane steppe and meadows, Taklimakan Desert,

and North Tibetan Plateau-Kunlun Mountains ecoregions, there has been substantial

restoration of alpine desert vegetation (Fig. 15).

Table 17: Area of revival and development activities and destruction of vegetation

coverage in Xinjiang province from 2000 to 2022.

Vegetation changes

Area (km2)

No Vegetation to No Vegetation

1,403,680.527142

Vegetation to No Vegetation (Deforestation)

21,749.340839

No Vegetation to Vegetation (Afforestation)

59,743.718885

Vegetation to Vegetation

223,681.424438

Figures 16 and 17 as well as table 18 show the changes in vegetation in 2022

compared to 2000. Figure 16 shows the vegetation cover has had a significant growing trend,

with, 37988.61 km2 of the target areas without vegetation cover being reduced. Also, vegetated

areas in 2022 was 283,462.35 km2, which has increased by an average of 37,988.467 km2

compared to 2000.

Figure 16: Visual patterns of vegetation changes based on the presence or absence of

vegetation from 2000 (A) to 2022 (B).

Table 18: The area of vegetation changes in Xinjiang province from 2000 to 2022.

Vegetation changes in 2000

Area (km2)

No Vegetation

1,463,463.10104

Vegetation

245,473.884643

Vegetation Changes in 2022

Area (km2)

No Vegetation

1,425,474.48557

Vegetation

28,3462.351692

Transylv. Rev. Syst. Ecol. Res. 26.1 (2024), "The Wetlands Diversity"

Figure 17: Models of changes in vegetation area based on the presence or absence of

vegetation in square kilometers from 2000 (A) to 2022 (B).

Figure 18: Visual patterns of afforestation and deforestation based on county boundaries

and ecoregions in square kilometers from 2000 (A) to 2022 (B).

In contrast, deforestation has been observed in the northern and southeastern parts of

Xinjiang Province, which account for only 27% of the total area. These regions correspond to

the ecoregions of Altai montane forest and forest steppe, Altai alpine meadow and tundra,

Emin Valley steppe, Tian Shan montane steppe and meadows, and Tian Shan montane conifer

forests. In general, it can be said that vegetation has grown significantly and environmental

protection measures and policies have had positive ecological effects. The area of afforestation

and deforestation was 59,743.72 and 21,749.34 km2, respectively, which means that the area of

restored areas is on average 37,994.38 km2 more than the destroyed areas (Tab. 19).