Next Article in Journal
The Vertical Distribution of Ice-Nucleating Particles over the North China Plain: A Case of Cold Front Passage
Previous Article in Journal
Deep Learning Methods for Semantic Segmentation in Remote Sensing with Small Data: A Survey
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Remote Sensing
Volume 15
Issue 20
10.3390/rs15204988
remotesensing-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Effects of Climate Variability and Human Activities on Vegetation Dynamics across the Qinghai–Tibet Plateau from 1982 to 2020

by
Yiyang Liu
,
Yaowen Xie
*,
Zecheng Guo
and
Guilin Xi
College of Earth and Environmental Sciences, Lanzhou University, Lanzhou 730000, China
*
Author to whom correspondence should be addressed.
Remote Sens. 2023, 15(20), 4988; https://doi.org/10.3390/rs15204988
Submission received: 15 September 2023 / Revised: 9 October 2023 / Accepted: 13 October 2023 / Published: 16 October 2023
(This article belongs to the Section Remote Sensing in Agriculture and Vegetation)
Browse Figures
Versions Notes

Abstract

:
In recent years, vegetation on the Qinghai–Tibet Plateau (QTP) has undergone significant greening. However, the causal factors underpinning this phenomenon, whether attributable to temperature fluctuations, precipitation patterns, or anthropogenic interventions, remain a subject of extensive scholarly debate. This study conducted a comprehensive analysis of the evolving vegetation across the QTP. The National Oceanic and Atmospheric Administration Climate Data Record Advanced Very High Resolution Radiometer Normalized Vegetation Difference Index (NOAA CDR AVHRR NDVI) dataset was employed to elucidate the intricate relationship between climatic variables and human activities driving vegetative transformations. The findings were as follows: The NDVI on the QTP has exhibited a significant greening trend at a rate of 0.0013/a (per year). A minor decline, accounting for only 17.6% of grasslands, was observed, which was primarily concentrated in the northwestern and northern regions. Through residual analysis, climate change was found to be the predominant driver, explaining 70.6% of the vegetation variability across the plateau. Concurrently, noticeable trends in temperature and precipitation increases were observed on the QTP, with the southern region demonstrating improved sensitivity to precipitation alterations. In summary, these results substantiate that a confluence of climatic warming, enhanced moisture availability, and a reduction in livestock population collectively creates an environment conducive to enhanced vegetation vigor on the QTP. This study highlights the significance of acknowledging the dual influence of climate and human agency in shaping vegetative dynamics, which is a critical consideration for informed land management strategies and sustainable development initiatives on this ecologically pivotal plateau.

1. Introduction

Due to increasing greenhouse gas concentrations, global warming, an incontrovertible reality, has resulted in alterations of precipitation patterns and increased temperatures, affecting ecosystems both directly and indirectly [1]. Vegetation plays a pivotal role in terrestrial ecosystems by linking the biosphere, pedosphere, atmosphere, lithosphere, and hydrosphere [2,3,4]. The QTP, recognized as the “Roof of the World” and the “Third Pole”, holds the highest elevation among global plateaus and is often referred to as the “Asian Water Tower.” It covers approximately 70% of the total plateau area and supports the most extensive alpine grassland ecosystem globally. Owing to its unique geography and complex environment, the QTP is one of the primary ecologically vulnerable regions in the world, serving as a responsive and sensitive zone to climate change, influencing not only Asia but also the broader Northern Hemisphere [5,6].
Over the past 50 years, warming in the QTP has doubled the global warming rate, particularly in winter. Hence, the QTP is becoming one of the foremost indicators of significant global warming [5,6,7]. Within the global warming trend, temperature escalations contribute to glacier shrinkage, snowpack melting, and a concomitant reduction in the total area of the glaciers and permafrost zones. This intricate interplay can trigger a series of unpredictable socioenvironmental and ecological crises [4]. These factors directly influence the hydrological patterns of the QTP, notably causing divergences in precipitation between the northern and southern regions. The enhanced variability in precipitation is an evident aspect of climate change on this plateau [8]. Overall, climate change in the QTP is characterized by the intertwined effects of warming and increased moisture despite regional variations [9]. This climate profile indicates a complex interplay of factors, necessitating a detailed examination and understanding of the interconnected factors that determine the ecological future of the QTP.
By 2015, the QTP, a significant national grazing zone, saw its actual livestock capacity rise to 158 million sheep units, surpassing the theoretical capacity by over 1.6 times. Overgrazing, especially in the northwest QTP, has increased since the 1990s and peaked in the 2000s [10]. Moreover, the QTP also sustains approximately 5.3 million pastoralists who had largely relied on grass-fed livestock farming by 2015 [11]. With population growth and infrastructure development, socioeconomic advancements have intensified pressures on ecosystems [6,12,13], mainly in the western and eastern areas of the QTP. The latest data indicate that alpine arid desert and sparse vegetation account for 34.9% of the QTP, exhibiting severe desertification, soil erosion, and freeze–thaw erosion. In addition, forest and shrub degradation reached 59%, primarily in the Hengduan Valley, while grassland degradation reached 80%, mainly in the northwest of the QTP [5,14]. There are also adverse effects, such as soil texture changes, nutrient depletion, and shifts in soil communities [15,16]. These changes can be attributed to climate change and human activities.
Grassland degradation on the Qinghai–Tibet Plateau has become an urgent ecological issue. The driving forces behind this can be subject to various perspectives in the academic community. Although several studies suggest that climatic factors play a dominant role [17], others believe that human activities exert a more pronounced influence than climate change [18,19]. Therefore, examining the roles of climate change and human activities in vegetation dynamics is essential for understanding the implications of climate warming and for offering insights into the ecological sustainability of the QTP.
Currently, remote sensing data with a high spatial–temporal resolution are extensively employed for monitoring surface vegetation dynamics and biomass assessment [20,21]. The NDVI and Enhanced Vegetation Index (EVI) are commonly applied indices for assessing the vegetation status in the QTP. Shen et al. [9] utilized three satellite-derived NDVI datasets (AVHRR NDVI, SPOT NDVI, and MODIS NDVI) to investigate temporal changes in QTP vegetation. The AVHRR NDVI data indicated a greening trend from 1982 to 2010, while all datasets highlighted a decline in the growing season NDVI in the southwest of the plateau during 2000–2010. In contrast, greening persisted in the northeastern region. Based on the GIMMS NDVI, Wei et al. [19] identified temperature rises and decreases in the high-altitude austere climate as factors expanding vegetation and improving high-altitude vegetation, particularly in the southwest, albeit with degradation in the northern and eastern QTP. Wang et al. [22] observed a general NDVI increase since the 1980s, albeit with notable spatial variability and localized decreases.
Many scholars have explored the relationship between vegetation and climate on the QTP and found that climate variability has a direct impact on the growth of vegetation [23]. Li et al. [13] utilized MODIS NDVI data to observe that vegetation in the northeastern and southwestern regions of the QTP responded prominently to precipitation, while the southern and central-eastern regions were more responsive to temperature. Utilizing the random forest algorithm, Wang et al. [22] assessed vegetation changes in the QTP and identified precipitation variability and the DEM (Digital Elevation Model) as primary drivers for both greening and browning across various vegetation types. Certain studies have indicated that, in addition to largely driving NDVI growth, rising temperatures can also intensify water shortages in low-precipitation areas of the QTP [24]. In addition, Lou et al. [25] employed a novel deep learning method and identified precipitation and temperature as the principal factors behind the greening of the QTP.
Numerous researchers have explored the primary drivers affecting vegetation growth in different subregions of the QTP. Bai et al. [26] discovered a modest upward trend in the NDVI in Sanjiangyuan from 1982 to 2015. Although the annual average temperature significantly increased, the precipitation remained largely unchanged. They concluded that the temperature rise facilitated vegetation growth and recovery in this area. Generally, on the QTP, heightened precipitation during warmer periods favors alpine vegetation development. Conversely, in arid periods temperature acts as the dominant constraint on vegetation growth [27]. These results indicate that, at the regional level, the influence of human activities on vegetation changes may not surpass that of climate change [13,17].
Numerous scholars have explored vegetation changes and their driving mechanisms across various time periods and study areas within the QTP, using diverse satellite datasets [18,19,28,29]. Based on the research above, we have used the NOAA CDR AVHRR NDVI long-term time series data to assess vegetation changes on the QTP from 1982 onwards, covering over four decades. In addition, scholars have mostly used methods such as machine learning and remote sensing model simulation to explore the factors affecting vegetation change. In this paper, utilizing a residual-based method, we evaluated the effects of climate change and human activity on vegetation changes for nearly 40 years. Our approach identified the impacts of climate change and human activities on NDVI trends over this extended period, capturing the relative contributions of climatic and anthropogenic factors to vegetation changes. Such differentiation is crucial for providing a scientific foundation for formulating effective vegetation restoration strategies for the QTP. The main objectives of this research were as follows: (1) to examine both temporal and spatial variations in vegetation across the plateau and its specific subregions; (2) to quantify the effects of climate change (CC) and human activities (HAs) on vegetation and to determine their individual contributions; and (3) to investigate the influencing mechanisms of meteorological elements, such as temperature increases and precipitation shifts, on vegetation, as well as the implications of contemporary human activities, including livestock counts and population fluctuations, for these dynamics.

2. Data and Methods

2.1. Study Area

Located between 26°–39°N and 73°–104°E (Figure 1b), the QTP has an average elevation of 4400 m. It extends from the Pamir Plateau in the west to the Hengduan Mountains in the east, bordered by the Kunlun, Alps, and Qilian Mountains to the north and the Himalayas to the south. Measuring 2800 km in length from east to west, with a width varying between 300 and 1500 km, the QTP covers an area of roughly 2.57 million km2. This region includes, either entirely or in part, six provinces: Tibet, Qinghai, Sichuan, Yunnan, Xinjiang, and Gansu [30].
In the northwestern regions of the QTP, an alpine desert steppe dominated by an arid climate is characterized by expanses of the Gobi Desert and visible rock formations (Figure 1a). Isolated pockets of alpine swamp meadows punctuate the landscape and cluster around water bodies, such as lakes and rivers. The vegetation of the QTP has shifted from evergreen broad-leaf forests in the southeast to desert grasslands in the northwest (Figure 1a), representing one of the most diverse vertical mountainous landscapes in China. This unique ecological structure indicates that the QTP plays a significant role as a critical carbon sink [31]. Agricultural activities in the QTP are tailored to the challenging environment. Due to the harsh climate and elevated altitude, pastoralism and animal husbandry dominate. Agricultural practices largely focus on the subsistence and cultivation of cold-resistant crops, such as barley and potatoes. Populations on the QTP cluster in valleys and river basins, which offer relatively mild conditions. Key administrative and economic centers include Lhasa and Xining, with smaller settlements dispersed throughout the plateau.

2.2. Data and Materials

We utilized the AVHRR NDVI 5th edition from the NOAA CDR as the remote sensing NDVI data source, featuring gridded daily NDVI measurements from 1981 to 10 days before the present using NOAA satellite data (7, 9, 11, 14, 16, 17, 18, and 19) and presented on a 0.05° × 0.05° global grid [32]. These data used AVHRR bands 1 and 2 for NDVI calculation (From Table 1). This high-quality, long-term dataset facilitates global vegetation growth studies and is applicable to areas such as arid region monitoring, agricultural forecasting, crop management, and biology. This dataset is one of the Land Surface CDR Version 5 products produced by the NASA Goddard Space Flight Center (GSFC) and the University of Maryland (UMD). Improvements for Version 5 include using the improved surface reflectance data; correcting the data for known errors in time, latitude, and longitude variables; and improvements in the global and variable attribute definitions. The dataset is in the netCDF-4 file format following ACDD and CF conventions. The dataset is accompanied by algorithm documentation, a data flow diagram, and the source code for the NOAA CDR Program.
In addition, for data quality validation (detailed in Section 3.1), we incorporated two other NDVI datasets. The GIMMS 3g NDVI dataset, the most recent release from NOAA’s Global Inventory Monitoring and Modeling System (GIMMS), covers July 1981 to December 2015. It provides bi-monthly data with a spatial resolution of 1/12° and is freely accessible at ecocast.arc.nasa.gov accessed on 1 January 2021. Moreover, the Moderate-resolution Imaging Spectroradiometer (MODIS) NDVI bridges the historical data of the NOAA’s AVHRR NDVI, ensuring continuous time-series analysis. The global MOD13A2 dataset offers 16-day intervals at a 1 km spatial resolution in a sinusoidal projection.
To complement the AVHRR CDR NDVI for QTP research, we selected the Chinese meteorological forcing dataset (CMFD) spanning from 1979 to 2018, sourced from the National Tibetan Plateau Data Center [33]. This dataset integrates various observational data, including remote sensing products, meteorological station records, and reanalysis data, and is primarily designed to support surface change research in China. The dataset captured seven near-surface meteorological elements from January 1979 to December 2018, including the 2 m air temperature, surface pressure, specific humidity, 10 m wind speed, downward short-wave and long-wave radiation, and precipitation rate. It has a 3 h temporal resolution and a spatial resolution of 0.1° × 0.1°.
To harmonize our survey with the focus of the study, DEM data featuring a resolution of 90 × 90 m were purchased from NASA (Figure 1b). Vegetation type data were derived from ESA’s Climate Change Initiative Land Cover dataset V2015 at a resolution of 300 × 300 m (Figure 1a). Socioeconomic data, including year-end livestock inventories, the Gross Domestic Product (GDP), and population figures, were primarily extracted from recent statistical yearbooks of Qinghai Province, the Tibet Autonomous Region, and other relevant provinces. These data were subsequently restructured at the county level for analysis. The eco-climatic zones followed the framework (as shown in Table 2) downloaded from Resource and Environmental Sciences Data Centre (https://www.resdc.cn).

2.3. Methods

(1)
Maximum Value Composition Method
Firstly, there is data preprocessing. Because the NDVI data are daily gridded NOAA CDR AVHRR NDVI data, we applied the maximum value composition method to transform daily data into a monthly NDVI to mitigate certain effects, such as the solar angle, cloud interference, and atmospheric residuals. This approach has been widely employed in NDVI product preprocessing [19]. The calculation procedure is as follows:
N D V I j = M a x ( N D V I i )
where N D V I i is the daily NDVI and N D V I j is the monthly NDVI. Through the above equation, we aggregated the daily NDVI data into monthly NDVI datasets. Subsequently, based on these monthly NDVI datasets, we synthesized four-season NDVI and annual NDVI datasets, from which the annual NDVImax is obtained for all the NDVIs involved in the following.
(2)
Trend Analysis
In this study, we utilized the Theil–Sen Median method to assess temporal patterns of regional factors, including the annual NDVI, mean annual temperature, and annual precipitation at the pixel level. The Theil–Sen Median method, commonly referred to as Sen’s slope estimation, is adept at identifying trends in time-series data. Notably, it can handle datasets with missing information, offers computational efficiency, and remains robust to measurement inaccuracies and outliers. Its widespread application in prolonged time-series trend analysis is well recognized. The calculation formula is as follows:
β = M e d i a n x j − x i j − i , ∀ j > i
where x i and x j denote the time-series data. If β > 0, the NDVI exhibits an upward trend; otherwise, it shows a downward trend. To validate this trend, the Mann–Kendall method was employed [34], which has the advantages of independence from assumptions about a normal distribution or linear trends and resilience against missing values and outliers. This nonparametric approach has been extensively utilized to analyze long-term time-series data trends. With this treatment, we obtained the spatial trends in the NDVI, temperature, precipitation, and other elements on the QTP.
(3)
Residual Analysis
Residual analysis has been widely employed to isolate the effects of both human activities and climate variables on vegetation change [35]. Typically, multiple linear regressions can establish relationships between the NDVI and climate variables, such as the mean temperature, precipitation, and solar radiation over a consistent period. This facilitates the estimation of the impacts of climate variables on vegetation, indirectly highlighting the crude impacts of human activities on the NDVI [36]. In this study, we delineated the linear relationship between the NDVI and monthly climate variables as follows:
N D V I h a = N D V I − N D V I c c
N D V I c c = a · P r e + b · T e m p + c · S o l a r
where climate elements include the Temp, representing the average annual temperature; Pre, representing the annual precipitation; and Solar, representing downward short-wave radiation. The NDVI is the real NDVI value, N D V I c c denotes the effect of climate factors on vegetation, and N D V I h a denotes the residual, which is a proxy for the impacts of human activities on vegetation (including pastoralism, population growth, and road infrastructure development). In this paper, we use the residual method to separate the effects of climatic elements and human activities on NDVI dynamics. In general, the annual NDVI of vegetation is mainly determined by climatic conditions and the intensity of human activities. Consequently, we disentangle the contribution of climate change to the NDVI by establishing a regression model based on the relationship between climate conditions and the NDVI, predicting the annual NDVI determined by climatic conditions. Under the assumption of there being no other nondeterministic factors, the residual variation between the observed NDVI values and the predicted values based on climate change represents the portion attributed to human activities. In the absence of a human activity influence, the interannual variation of residuals should exhibit random characteristics centered around zero. If there is a significant decreasing trend in the interannual variation of residuals, it indicates vegetation degradation induced by human activities. Conversely, it suggests that human activities have improved the ecological environment.
Subsequently, a trend analysis was conducted on the N D V I h a and N D V I c c over an extended time series. F N D V I denotes the linear slope of the NDVI, F C C denotes the linear slope of the NDVI under the impacts of climate variables, and F H A denotes the linear slope of the NDVI under the influence of human activities. Furthermore, the relative contributions of climate change and human activities were calculated using the equations shown in Table 3 [37,38].
(4)
Pearson Correlation Analysis
The Pearson correlation coefficient r is a standard metric used to measure the correlation between two variables. Therefore, to assess the response of vegetation to climatic and anthropogenic factors, correlation analysis was conducted. The correlation coefficients between the NDVI and meteorological data were determined as follows:
r x y = ∑ i = 1 n x i − x − y i − y − ∑ i = 1 n x i − x − 2 ∑ i = 1 n y i − y − 2
where x i and y i represent the NDVI and meteorological factor at the pixel scale during 1982–2020, respectively, and r x y denotes the correlation coefficient between indicators x i and y i . In this paper, we calculate the spatially distributed correlations of the NDVI with temperature and NDVI with precipitation, respectively. In addition, in the data validation section we also calculate the spatial correlation between NOAA CDR AVHRR NDVI and MODIS NDVI data and between NOAA CDR AVHRR NDVI and GIMMS-3gNDVI data to validate their usability.

3. Results

3.1. Data Quality Assessment

Owing to missing NOAA CDR AVHRR NDVI raw data for October, November, and December in 1994 and 2020 in China, a value of 0 was substituted during data processing. To ensure reliability, the acquired NOAA CDR AVHRR NDVI data were assessed for validation. During 1982–2015, the correlation coefficient between the NOAA CDR AVHRR NDVI and GIMMS3g NDVI was significant (r = 0.95, p < 0.05, n = 408). Given that both datasets rely on the same satellite sensor, a high correlation is anticipated. During 2001–2020, the correlation between the NOAA CDR AVHRR NDVI and MODIS NDVI was marginally lower (r = 0.88, p < 0.05, n = 240). During 2001–2015, the correlation coefficient between the MODIS NDVI and GIMMS3g NDVI was notably high (r = 0.93, p < 0.05, n = 180). Over the period 2001–2015, a high correlation coefficient was observed between the NOAA CDR AVHRR NDVI and MODIS NDVI (r = 0.93, p < 0.05, n = 180). For the period 2001–2015, the consistency between the NOAA CDR AVHRR NDVI and MODIS NDVI surpassed that of the GIMMS3g NDVI. Similarly, between 2001 and 2020 the consistency between the NOAA CDR AVHRR NDVI and MODIS NDVI was higher than that of the GIMMS3g NDVI. The reduced consistency of the MODIS NDVI can be attributed to the absence of NOAA CDR AVHRR NDVI data during 2015–2020, particularly the missing values in 2020.
To provide additional validation of the above process, we also calculated the spatial correlation between the individual NDVI data (Figure 2a,b). The NOAA CDR AVHRR NDVI exhibits a robust spatial correlation with the other two NDVI datasets across the entire QTP. Although slight variations are observed in different subregions, they do not significantly impact our ability to dynamically monitor vegetation changes across the entire plateau. Hence, the NOAA CDR AVHRR NDVI data, with their compatibility with other NDVI datasets in the QTP and extensive time series, can be well suited for subsequent analytical endeavors.

3.2. Time Series of Major Change

As shown in Figure 3a, the NDVI over the QTP has exhibited a significant increase since 1982 at a rate of 0.0013/a. By employing different monitoring data sources and diverse vegetation indices (e.g., the Leaf Area Index, Net Primary Production), numerous researchers have similarly reported increasing vegetation trends on the QTP, with NDVI rates ranging between 0.002 and 0.0020/a [39,40]. We divided the period into three distinct phases: an increase of 0.0012/a from to 1982 to 2000 and a decline of 0.0010/a between 2000 and 2015, followed by a pronounced increase at 0.0130/a thereafter.
In the QTP, alpine grassland and alpine meadows primarily dominate the vegetation distribution, typically exhibiting low NDVI values. Therefore, NDVI values below 0.10 were categorized as bare ground [35]. Figure 3b illustrates that winter and spring NDVI values have remained relatively stable in recent years at approximately 0.2. This stability could be attributed to the relatively dry winter climate induced by global warming, which counteracts the vegetation greening trends. However, NDVI trends during summer and autumn closely align with the interannual fluctuations of the QTP because most vegetation in this region consists of seasonal deciduous plants, exhibiting peak growth in July and August within a single growth cycle [41].

3.3. Spatial Change

3.3.1. Spatial Variability of Major Change

The QTP, with its high altitude; intricate topography; extensive latitudinal and longitudinal ranges; and diverse eco-climatic and vegetation zones, exhibits pronounced spatial heterogeneity in NDVI change (Figure 4). Notably, the NDVI progressively decreased from east to west owing to the particular vegetation distribution on the QTP. The alpine grasslands and meadows on the QTP were further influenced by global changes [42,43]. Figure 4 illustrates that only 17.6% of the vegetation NDVI demonstrated a decline along the west-north and northern edge of the QTP from 1982 to 2020. Conversely, the regions east of 95°E on the QTP demonstrate a more pronounced upward trend. The inset of the NDVI spatial distribution map indicates an NDVI trend increase of 0.00004–0.00005/a per 1° longitude, while the adjacent graph on the right represents an NDVI trend increase of 0.00003–0.00005/a per 1° latitude. These graphs suggest that an average NDVI trend exceeding 0.002/a is primarily situated south of 29°N and east of 98°E.

3.3.2. Spatial Changes in Different Zones

We conducted an in-depth analysis of vegetation trends across distinct zones to elucidate the spatial heterogeneity of vegetation changes in the QTP. Our study evaluated various eco-climatic zones, as illustrated in Figure 5a. The zones with minimal changes mainly included the eco-climatic regions H1D1, H2D3, H2D1, and H2D2. Notably, these zones were positioned within the arid expanse and western periphery of the QTP and were characterized by sparse vegetation cover. As shown in Figure 5d, the distribution across the eco-climatic zones was relatively uniform. However, the humid zones, particularly VA5 and VA6, had the lowest proportion. Although the H1C2 zone, centrally located in this domain, exhibited a mean NDVI trend just below 0.001/a, regions indicating significant enhancement (with NDVI trends surpassing 0.002/a) were predominantly identified on the eastern edge and central area of the QTP. Zones such as H2C1, VA5, VA6, H2AB1, and H1B1 were situated in lower alpine temperate areas or moist to subhumid climate regions.
We examined NDVI dynamics across diverse land cover types, including 10 primary categories. The analysis utilized land use data primarily sourced from the land use data shown in Figure 1a, where the original 23 classifications were combined into 10 primary categories. “Tree cover, flooded” and “other forest” collectively accounted for only 0.002%, and other vegetation types, such as wetland, settlement, and shrubland, each represent less than 1%. Consequently, NDVI trend analysis was not conducted for categories below 1%. Figure 5b distinctly highlights the increased growth in “agriculture,” “forest,” and “other forest” zones. Both cropland and tree cover demonstrated increases exceeding 0.002/a. In the second tier, grassland, wetland, and settlement exhibited increments of approximately 0.001/a, with shrubland and water reflecting modest increases of approximately 0.0008/a. The “water” category included surface features, such as glaciers and snow. The increases in snow and tree lines due to global warming have induced vegetated area expansion, thereby increasing the NDVI for aquatic regions. Conversely, bare land (including the desert and Gobi categories) lacking significant NDVI changes was consistent with the existing knowledge.
Figure 5c illustrates the NDVI trends per 500 m elevation increment. As shown in Figure 5f, the elevation profile of the QTP primarily laid between 4500 and 5500 m, representing a significant 70.3% plateau. With an increasing altitude, there was a noticeable decline in the NDVI. The QTP was divided into 15 subregions, with only four smaller zones demonstrating notable NDVI increases of approximately 0.002/a. These zones cover less than 1% of the entire QTP. Below 2500 m elevation, an NDVI increase of 0.002/a was evident. In contrast, above 6000 m the NDVI change was minimal at 0.0006/a. Notably, between 3500 and 4500 m there was a substantial NDVI increase of 0.0018/a. Compared with other elevational bands, the vegetation within the 3500–4500 m range demonstrates increased responsiveness to environmental changes, especially climatic variations.

3.4. Contribution of HA and CC to NDVI Change

As shown in Figure 6, we estimated the slope of human activities (HAs) and climate change (CC, encompassing Pre, Temp, and Srad) to the NDVI from 1982 to 2020 for each grid of the QTP, assessed using residual analysis. Figure 6a illustrates that HA positively influenced NDVI changes in 61.8% of the QTP. Specifically, the contribution of HA exceeded 0.0005/a in 30.5% of the regions and 0.001/a in 14.2%, predominantly in the southern and southeastern QTP. Conversely, the northwestern and northern parts of the QTP exhibited a negative contribution to HA. On average, HA contributed 0.0003/a to the NDVI changes across the QTP, representing 41.2% of the entire region. The cumulative impact of the three climatic factors under CC was positive for 75.7% of the QTP, with most areas exceeding 0.0005/a. The mean CC contribution to the NDVI change of the QTP was 0.0008/a, accounting for 44.1% of the total NDVI change.
As depicted in Figure 6c,d, we computed the relative contribution rates of CC and HA to NDVI changes. The total contributions of CC to NDVI were dominant (with relative contribution rates surpassing 50%) in 70.6% of the areas, mainly distributed in the central and northeastern regions of the QTP (Figure 6c). The dominant contribution rates of HA to NDVI change were observed in 29.4% of the QTP, primarily in the northwestern and southeastern regions (Figure 6d). Figure 7 illustrates the magnitude and relative contribution rate of each factor to the NDVI changes at the city scale. Most cities exhibited a CC contribution surpassing 0.001/a, with only three cities, Haixi, Rikaze, and Naqu, registering contributions below this value. Generally, HA contributions did not exceed 0.001/a, with the Ngari Prefecture notably presenting a negative HA contribution. Figure 7b shows that the contribution of CC to vegetation exceeded that of HA, with CC contributions averaging around 60% in cities, reaching a peak of 85.8%.

4. Discussion

4.1. Vegetation and Climate Change

Over the past four decades, the QTP has experienced an increase in both temperature and precipitation [24]. The overall temperature of the QTP has exhibited a relatively uniform and significant warming trend (from Figure 8a), particularly in zones H2D3, H1C2, H2D1, H2AB1, H1C1, and H1B1, where average temperatures have climbed by approximately 1.5–2.6 °C from the 1980s to the 2020s. However, areas such as H1D1 and H2D2, situated on the northwestern edge of the QTP, recorded a slight temperature decrease of approximately 0.9–1.1 °C. Precipitation patterns exhibited considerable temporal and spatial variation, with pronounced spatial heterogeneity [5,44]. Figure 8b highlights regions within a blue polygon where the coefficient of precipitation variation surpassed 35%. Overall, precipitation in the plateau exhibited an upward trend, especially in the central and western areas, including H1C1, H1C2, H1D1, H2C1, H2D2, and H2D3. In these regions, the average precipitation increased by 24–130 mm compared with 1982. Moreover, the southwestern and southeastern parts of the plateau experienced decreases in precipitation levels by an average of 20–35 mm compared with 1982.
Figure 8c illustrates that the NDVI predominantly exhibited a positive correlation with the overall temperature. Specifically, in the southwestern and eastern parts of Qinghai Province, notably northeast of Qinghai Lake, and along the Gansu–Sichuan border the correlation coefficient between the temperature and NDVI exceeded 0.65. These areas, characterized by lower elevations and higher temperatures, facilitated vegetation growth. Conversely, east of 85°E in Tibet, the temperature–NDVI correlation was mildly positive with a coefficient of approximately 0.45. However, certain regions within the Ngari Prefecture showed a negative correlation, with a coefficient of −0.35. In arid regions with sparse vegetation, during this stage, a decrease in temperature and a significant increase in precipitation promoted vegetation growth in the area, resulting in an increase in the NDVI [45].
In most parts of Tibet, the eastern region of Qinghai Province, and the western part adjoining Tibet, the correlation coefficient between the annual precipitation and NDVI was approximately 0.6 (Figure 8d). However, a slight negative correlation existed between the precipitation and NDVI in regions such as the Xinjiang Uygur Autonomous Region adjacent to Tibet, eastern Tibet, and certain parts of Yunnan. In precipitation-rich zones such as the southern Tibetan Plateau, vegetation exhibited a minimal response to precipitation changes. In the heavily degraded northern margin of the QTP, the NDVI indicated a negative correlation with both the temperature and precipitation, whereas in regions with robust vegetation growth the NDVI typically exhibited a positive correlation with both the temperature and precipitation.
Based on the above analysis, Qinghai and Central Tibet exhibited pronounced warming and humidification trends. Notably, in areas near the Qaidam Basin, there was a significant increase in the NDVI. This suggests that regional climate warming and humidification are conducive to surface greening and vegetation growth. The semi-arid and arid regions of the QTP house most alpine grasslands, which are susceptible to climate warming and are highly sensitive to precipitation changes. Considering the correlation between temperature and precipitation, an increase in temperature enhanced vegetation growth across the entire plateau. Additionally, north of 33°N, increasing precipitation further promoted vegetation growth. Conversely, south of 33°N, a decline in precipitation constrained vegetation growth, indicating an intricate relationship between the plateau ecology and precipitation patterns.
Subsequently, six counties in the western Naqu region were selected (as shown in Figure 1b). As a typical study area, the selected areas were situated in northern Tibet, constituting a key pastureland of the QTP. Research has indicated a prevalent greening trend in northern Tibet [3,46], possibly due to its historically sparse vegetation coverage and recent climate “warming and humidification.” This can facilitate the growth of alpine vegetation. Figure 9 highlights the noticeable vegetation enhancement in this region, albeit with varying degradations. MODIS data pinpoint degradation is mainly south of 32°N, particularly in the southeast. In contrast, the AVHRR data indicated degradation predominantly in the central and southern parts, with less accurate southeast monitoring. After 2000, the climate displayed a trend towards “warmer and drier” conditions, further intensifying vegetation degradation.
Temperature manipulation experiments conducted by Wang et al. [24] indicated that the influence of temperature on the NDVI depends on precipitation levels. Specifically, areas with annual precipitation of approximately 500 mm demonstrated pronounced NDVI growth. Increasing evidence has emphasized the crucial influence of precipitation on species richness, alpine grassland ecosystem diversity, primary production, and carbon and water cycle spatial distribution.

4.2. Human Drivers

Since the launch of the Qinghai–Tibet Railway and the Grand Western Development Program in 2000, the central and southern QTP has experienced accelerated socioeconomic development (see Figure 10). A surge in sectors such as mining, manufacturing, services, and tourism has led to population migration to the region, resulting in consistent population growth [19]. As depicted in Figure 11a, the QTP population has more than doubled, increasing from 4.3 million in 1970 to 9.4 million in 2021. With societal and economic advancements in China, the primary industry has witnessed substantial growth since 2000. In addition, the secondary and tertiary industries exhibit sharp increases in a linear trend.
The QTP primarily belongs to two different provinces, Qinghai and Tibet (Figure 11d). Due to the sociocultural differences between both regions, it should be evaluated whether their trajectories have diverged over recent decades. Therefore, this study sourced data on cattle, sheep, horses, and other livestock from the Tibet and Qinghai Province Statistical Yearbooks. These data were then uniformly converted into sheep units to analyze the livestock population trends in both provinces.
Based on the observed trends from 1982 to 2020, we categorized the period into three distinct phases. As depicted in Figure 11c, from 1982 to 2000 Tibet’s NDVI exhibited a positive trajectory, growing at 0.001/a (p > 0.05). Concurrently, livestock counts experienced a modest increase of 0.13 million sheep/a (Figure 11b). Both the temperature and precipitation significantly increased during this period (p < 0.05), making the conditions conducive to vegetation growth. From 2000 to 2005, the NDVI declined at a rate of 0.0076/a (p < 0.05). In contrast, livestock numbers increased at a rapid pace of 1.23 million sheep/a. Although temperature increased during this phase, precipitation remained relatively stable, potentially intensifying drought conditions and hampering vegetation greening. During 2005–2020, the livestock population experienced a dramatic decline at the rate of 0.63 million sheep/a (p < 0.05). Simultaneously, both the temperature and precipitation demonstrated no significant alterations, while the NDVI increased at 0.0034/a (p < 0.05). Overall, the livestock count in Tibet in the 2020s was close to that in the 1980s, suggesting a potential continued decline.
The NDVI and livestock population in Qinghai Province were divided into three distinct periods. During 1982–1991, as illustrated in Figure 11f, the NDVI in Qinghai exhibited minimal variation. Livestock figures displayed a negligible increase, peaking in 1991 (Figure 11e). As the temperature increased and precipitation remained relatively constant, both climatic and livestock changes were detrimental to vegetation growth. From 1992 to 2002, the NDVI in Qinghai demonstrated a positive trend, increasing at 0.005/a. Between 1992 and 1999, livestock counts declined substantially at a rate of 1.08 million sheep/a (p < 0.05). During this phase, both the temperature and precipitation increased. From 2003 to 2020, the NDVI in Qinghai showed a marginal growth of 0.001/a (p > 0.05). Between 2000 and 2020, the livestock population decreased (p > 0.05). During this period, both the temperature and precipitation increased, which enhanced vegetation growth. In summary, as inferred from Figure 11e, the livestock numbers in the 2020s were lower than those in the 1980s in Qinghai.
In general, “warming and humidification” contributed to vegetation greening [17]. On the one hand, a persistent rise in temperature might amplify extreme heat events, potentially mitigating the positive feedback between temperature and vegetation growth [40], as evidenced in Tibet from 2000 to 2005. On the other hand, the spatially dispersed grazing activities on the plateau suggest that changes in livestock numbers may not immediately influence vegetation feedback, as illustrated by vegetation trends in Qinghai from 1982 to 1990.
Restoring degraded grassland ecosystems can be an intricate and long-term endeavor. In response to highland grassland degradation, the Chinese government has implemented multiple measures since the 1950s, including initiatives such as the “Three-North Shelterbelt,” reforestation of neglected farmland, conversion of grazing areas to forests, and targeted grassland restoration projects. By 2012, the data confirmed that the fenced region in northern Qinghai Province reached 3.32 million hectares. In certain degraded alpine grasslands, fencing has been proven to be beneficial for vegetation recovery [47]. The “Grassland Ecosystem Compensation Policy” (GECP), introduced in 2010, has not only amplified livestock off-take rates, decreasing livestock counts, but also increased incomes for herders. This policy, in turn, has alleviated human-induced pressure on grasslands, enhancing their quality, although the extent of the improvement varies [46,48].

4.3. Limitations

By 2012, fencing enclosed approximately one-third of the northern Tibetan Plateau. Such initiatives have demonstrated the ability to promote vegetation consistency and may even improve plant diversity. Nonetheless, it is crucial to recognize that fencing can impede wildlife mobility, intensify grazing in unfenced areas, and pose significant financial challenges to regional and national authorities [41]. The effects of fencing differ across contexts, indicating the need for detailed evaluation in future research. Several studies have highlighted a reduction in vegetation near infrastructure, such as roads, rivers, and lakes [48]. Therefore, future research should consider the effects of transportation systems, herder settlement, and urban expansion on grassland dynamics.

5. Conclusions

Across the QTP, the NDVI exhibited an increase of 0.0013/a without significant variations during winter and spring. The eastern QTP experienced a more pronounced increase in NDVI, whereas only 17.6% of grasslands, primarily located in the northwest and northern periphery, demonstrated a declining trend. Residual analysis indicated that climate change was the primary driver of vegetation shifts for most of the plateau, accounting for approximately 70.6%. In contrast, human activities exhibited a more focal impact in the southeastern section of the plateau, accounting for 29.4% of the changes. Prospectively, the vegetation on the QTP can maintain the greening due to warming and increased humidity. After 2000, the positive impacts of human activities increased, reflecting the results of various ecological initiatives. Future studies should incorporate the consequences of such interventions on vegetation changes to accurately quantify the influence of human activity on vegetation at the pixel scale and to provide new solutions for the harmonious and friendly coexistence of human beings and nature in the context of climate variability.

Author Contributions

Validation, Z.G.; Resources, G.X.; Writing—review & editing, Y.L.; Project administration, Y.X. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Second Tibetan Plateau Scientific Expedition and Research Program (STEP) of the Ministry of Science and Technology of the People’s Republic of China (2019QZKK0603) and the Strategic Priority Research Program of Chinese Academy of Sciences, Pan-Third Pole Environment Study for a Green Silk Road (Pan-TPE) (XDA2009000001).

Data Availability Statement

All NDVI data are openly available via the Google Earth Engine.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. IPCC. Climate Change 2021: The Physical Science Basis; Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change; Cambridge University Press: Cambridge, UK; New York, NY, USA, 2021. [Google Scholar]
  2. Kong, D.; Zhang, Q.; Singh, V.P.; Shi, P. Seasonal vegetation response to climate change in the Northern Hemisphere (1982–2013). Glob. Planet. Chang. 2017, 148, 1–8. [Google Scholar] [CrossRef]
  3. Zhang, Z.; Chang, J.; Xu, C.-Y.; Zhou, Y.; Wu, Y.; Chen, X.; Jiang, S.; Duan, Z. The response of lake area and vegetation cover variations to climate change over the Qinghai-Tibetan Plateau during the past 30years. Sci. Total Environ. 2018, 635, 443–451. [Google Scholar] [CrossRef] [PubMed]
  4. Zhang, Q.; Shen, Z.; Pokhrel, Y.; Farinotti, D.; Singh, V.P.; Xu, C.-Y.; Wu, W.; Wang, G. Oceanic climate changes threaten the sustainability of Asia’s water tower. Nature 2023, 615, 87–93. [Google Scholar] [CrossRef]
  5. Chen, D.; Xu, B.; Yao, T.; Guo, Z.; Cui, P.; Chen, F.; Zhang, T. Assessment of past, present and future environmental changes on the Tibetan Plateau. Chin. Sci. Bull. 2015, 60, 1–2+3025–3035. [Google Scholar]
  6. Kuang, X.; Jiao, J.J. Atmospheres: JGR. Review on climate change on the Tibetan Plateau during the last half century. J. Geophys. Res. D. Atmos. JGR 2016, 121, 3979–4007. [Google Scholar] [CrossRef]
  7. Pauli, H.; Gottfried, M.; Reiter, K.; Klettner, C.; Grabherr, G. Signals of range expansions and contractions of vascular plants in the high Alps: Observations (1994–2004) at the GLORIA * master site Schrankogel, Tyrol, Austria. Glob. Chang. Biol. 2007, 13, 147–156. [Google Scholar] [CrossRef]
  8. Huss, M.; Hock, R. Global-scale hydrological response to future glacier mass loss. Nat. Clim. Chang. 2018, 8, 135–140. [Google Scholar] [CrossRef]
  9. Shen, M.; Piao, S.; Jeong, S.-J.; Zhou, L.; Zeng, Z.; Ciais, P.; Chen, D.; Huang, M.; Jin, C.-S.; Li, L.Z.X.; et al. Evaporative cooling over the Tibetan Plateau induced by vegetation growth. Proc. Natl. Acad. Sci. USA 2015, 112, 9299–9304. [Google Scholar] [CrossRef]
  10. National Bureau of Statistics of China. China Statistics Yearbook; China Statistics Press: Beijing, China, 2020. [Google Scholar]
  11. Bao, C.; Liu, R. Spatiotemporal Evolution of the Urban System in the Tibetan Plateau. J. Geo-Inf. Sci. 2019, 21, 1330–1340. [Google Scholar]
  12. Pang, G.; Wang, X.; Yang, M. Using the NDVI to identify variations in, and responses of, vegetation to climate change on the Tibetan Plateau from 1982 to 2012. Quat. Int. 2016, 444, 87–96. [Google Scholar] [CrossRef]
  13. Li, L.; Zhang, Y.; Liu, L.; Wu, J.; Wang, Z.; Li, S.; Zhang, H.; Zu, J.; Ding, M.; Paudel, B. Spatiotemporal Patterns of Vegetation Greenness Change and Associated Climatic and Anthropogenic Drivers on the Tibetan Plateau during 2000–2015. Remote Sens. 2018, 10, 1525. [Google Scholar] [CrossRef]
  14. Fu, B.J.; Ouyang, Z.Y.; Shi, P.; Fan, J.; Wang, X.D.; Zheng, H.; Zhao, W.; Wu, F. Current Condition and Protection Strategies of Qinghai-Tibet Plateau Ecological Security Barrier. Bull. Chin. Acad. Sci. 2021, 36, 1298–1306. [Google Scholar]
  15. Yang, Y.; Wu, L.; Lin, Q.; Yuan, M.; Xu, D.; Yu, H.; Hu, Y.; Duan, J.; Li, X.; He, Z.; et al. Responses of the functional structure of soil microbial community to livestock grazing in the Tibetan alpine grassland. Glob. Chang. Biol. 2013, 19, 637–648. [Google Scholar] [CrossRef] [PubMed]
  16. Huang, K.; Zhang, Y.; Zhu, J.; Liu, Y.; Zu, J.; Zhang, J. The Influences of Climate Change and Human Activities on Vegetation Dynamics in the Qinghai-Tibet Plateau. Remote Sens. 2016, 8, 876. [Google Scholar] [CrossRef]
  17. Lehnert, L.W.; Wesche, K.; Trachte, K.; Reudenbach, C.; Bendix, J. Climate variability rather than overstocking causes recent large scale cover changes of Tibetan pastures. Sci. Rep. 2016, 6, 24367. [Google Scholar] [CrossRef] [PubMed]
  18. Xiong, Q.; Xiao, Y.; Liang, P.; Li, L.; Zhang, L.; Li, T.; Pan, K.; Liu, C. Trends in climate change and human interventions indicate grassland productivity on the Qinghai–Tibetan Plateau from 1980 to 2015. Ecol. Indic. 2021, 129, 108010. [Google Scholar] [CrossRef]
  19. Wei, Y.; Lu, H.; Wang, J.; Wang, X.; Sun, J. Dual Influence of Climate Change and Anthropogenic Activities on the Spatiotemporal Vegetation Dynamics Over the Qinghai-Tibetan Plateau From 1981 to 2015. Earth’s Future 2022, 10, e2021EF002566. [Google Scholar] [CrossRef]
  20. Fensholt, R.; Proud, S.R. Evaluation of Earth Observation based global long term vegetation trends—Comparing GIMMS and MODIS global NDVI time series. Remote Sens. Environ. 2012, 119, 131–147. [Google Scholar] [CrossRef]
  21. Wang, X.; Xiao, J.; Li, X.; Cheng, G.; Ma, M.; Che, T.; Dai, L.; Wang, S.; Wu, J. No Consistent Evidence for Advancing or Delaying Trends in Spring Phenology on the Tibetan Plateau. J. Geophys. Res. Biogeosci. 2017, 122, 3288–3305. [Google Scholar] [CrossRef]
  22. Wang, H.; Zhan, J.; Wang, C.; Liu, W.; Yang, Z.; Liu, H.; Bai, C. Greening or browning? The macro variation and drivers of different vegetation types on the Qinghai-Tibetan Plateau from 2000 to 2021. Front. Plant Sci. 2022, 13, 1045290. [Google Scholar] [CrossRef]
  23. Ran, Q.; Hao, Y.; Xia, A.; Liu, W.; Hu, R.; Cui, X.; Xue, K.; Song, X.; Xu, C.; Ding, B.; et al. Quantitative Assessment of the Impact of Physical and Anthropogenic Factors on Vegetation Spatial-Temporal Variation in Northern Tibet. Remote Sens. 2019, 11, 1183. [Google Scholar] [CrossRef]
  24. Wang, Y.; Lv, W.; Xue, K.; Wang, S.; Zhang, L.; Hu, R.; Zeng, H.; Xu, X.; Li, Y.; Jiang, L.; et al. Grassland changes and adaptive management on the Qinghai–Tibetan Plateau. Nat. Rev. Earth Environ. 2022, 3, 668–683. [Google Scholar] [CrossRef]
  25. Lou, P.; Wu, T.; Yang, S.; Wu, X.; Chen, J.; Zhu, X.; Chen, J.; Lin, X.; Li, R.; Shang, C.; et al. Deep learning reveals rapid vegetation greening in changing climate from 1988 to 2018 on the Qinghai-Tibet Plateau. Ecol. Indic. 2023, 148, 110020. [Google Scholar] [CrossRef]
  26. Bai, Y.; Guo, C.; Degen, A.A.; Ahmad, A.A.; Wang, W.; Zhang, T.; Li, W.; Ma, L.; Huang, M.; Zeng, H.; et al. Climate warming benefits alpine vegetation growth in Three-River Headwater Region, China. Sci. Total Environ. 2020, 742, 140574. [Google Scholar] [CrossRef]
  27. Kong, R.; Zhang, Z.; Zhang, F.; Tian, J.; Chang, J.; Jiang, S.; Zhu, B.; Chen, X. Increasing carbon storage in subtropical forests over the Yangtze River basin and its relations to the major ecological projects. Sci. Total Environ. 2020, 709, 136163. [Google Scholar] [CrossRef]
  28. Liu, E.; Xiao, X.; Shao, H.; Yang, X.; Zhang, Y.; Yang, Y. Climate Change and Livestock Management Drove Extensive Vegetation Recovery in the Qinghai-Tibet Plateau. Remote Sens. 2021, 13, 4808. [Google Scholar] [CrossRef]
  29. Zhu, B.; Zhang, Z.; Tian, J.; Kong, R.; Chen, X. Increasing Negative Impacts of Climatic Change and Anthropogenic Activities on Vegetation Variation on the Qinghai–Tibet Plateau during 1982–2019. Remote Sens. 2022, 14, 4735. [Google Scholar] [CrossRef]
  30. MCA. Brief Book of Administrative Divisions of the People’s Republic of China 2015; Ministry of Civil Affairs of the People’s Republic of China; China Cartographic Publishing House: Beijing, China, 2015. [Google Scholar]
  31. Cui, Q.H.; Jiang, Z.G.; Liu, J.K.; Su, J.P. A Review of the cause of rangeland degradation on Qinghai-Tibet Plateau. Pratacult. Sci. 2007, 24, 7. [Google Scholar]
  32. Vermote, E. NOAA Climate Data Record (CDR) of Normalized Difference Vegetation Index (NDVI); Version 5; NOAA National Centers for Environmental Information: Boulder, CO, USA, 2019. [Google Scholar] [CrossRef]
  33. Yang, K.; He, J. China Meteorological Forcing Dataset (1979–2018); National Tibetan Plateau Data Center: Beijing, China, 2019. [Google Scholar] [CrossRef]
  34. Yue, S.; Pilon, P.; Cavadias, G. Power of the Mann–Kendall and Spearman’s rho tests for detecting monotonic trends in hydrological series. J. Hydrol. 2002, 259, 254–271. [Google Scholar] [CrossRef]
  35. Li, L.; Zhang, Y.; Liu, L.; Wu, J.; Li, S.; Zhang, H.; Zhang, B.; Ding, M.; Wang, Z.; Paudel, B. Current challenges in distinguishing climatic and anthropogenic contributions to alpine grassland variation on the Tibetan Plateau. Ecol. Evol. 2018, 8, 5949–5963. [Google Scholar] [CrossRef]
  36. Evans, J.; Geerken, R. Discrimination between climate and human-induced dryland degradation. J. Arid Environ. 2004, 57, 535–554. [Google Scholar] [CrossRef]
  37. Sun, W.; Song, X.; Mu, X.; Gao, P.; Wang, F.; Zhao, G. Spatiotemporal vegetation cover variations associated with climate change and ecological restoration in the Loess Plateau. Agric. For. Meteorol. 2015, 209–210, 87–99. [Google Scholar] [CrossRef]
  38. Ma, M.; Wang, Q.; Liu, R.; Zhao, Y.; Zhang, D. Effects of climate change and human activities on vegetation coverage change in northern China considering extreme climate and time-lag and -accumulation effects. China. Sci. Total Environ. 2022, 860, 160527. [Google Scholar] [CrossRef]
  39. Pan, T.; Zou, X.; Liu, Y.; Wu, S.; He, G. Contributions of climatic and non-climatic drivers to grassland variations on the Tibetan Plateau. Ecol. Eng. 2017, 108, 307–317. [Google Scholar] [CrossRef]
  40. Zhu, Z.; Piao, S.; Myneni, R.B.; Huang, M.; Zeng, Z.; Canadell, J.G.; Ciais, P.; Sitch, S.; Friedlingstein, P.; Arneth, A.; et al. Greening of the Earth and its drivers. Nat. Clim. Chang. 2016, 6, 791. [Google Scholar] [CrossRef]
  41. Sun, J.; Liu, M.; Fu, B.; Kemp, D.; Zhao, W.; Liu, G.; Han, G.; Wilkes, A.; Lu, X.; Chen, Y.; et al. Reconsidering the efficiency of grazing exclusion using fences on the Tibetan Plateau. Sci. Bull. 2020, 65, 1405–1414. [Google Scholar] [CrossRef]
  42. Zhou, H.; Zhou, L.; Zhao, X.; Liu, W.; Li, Y.; Gu, S.; Zhou, X. Stability of alpine meadow ecosystem on the Qinghai-Tibetan Plateau. Chin. Sci. Bull. 2006, 51, 320–327. [Google Scholar] [CrossRef]
  43. Ganjurjav, H.; Gao, Q.; Zhang, W.; Liang, Y.; Li, Y.; Cao, X.; Wan, Y.; Li, Y.; Danjiu, L. Effects of Warming on CO2 Fluxes in an Alpine Meadow Ecosystem on the Central Qinghai–Tibetan Plateau. PLoS ONE 2015, 10, e0132044. [Google Scholar] [CrossRef]
  44. Huang, S.W.; Li, X.S.; Wu, B.F.; Pei, L. The Distribution and Drivers of Land Degradation in the Three-North Shelter Forest Region of China during 1982–2006. Acta Geogr. Sin. 2012, 67, 589–598. [Google Scholar]
  45. Ding, J.; Yang, T.; Zhao, Y.; Liu, D.; Wang, X.; Yao, Y.; Peng, S.; Wang, T.; Piao, S. Increasingly Important Role of Atmospheric Aridity on Tibetan Alpine Grasslands. Geophys. Res. Lett. 2018, 45, 2852–2859. [Google Scholar] [CrossRef]
  46. Sun, J.; Hou, G.; Liu, M.; Fu, G.; Zhan, T.; Zhou, H.; Tsunekawa, A.; Haregeweyn, N. Effects of climatic and grazing changes on desertification of alpine grasslands, Northern Tibet. Ecol. Indic. 2019, 107, 105647. [Google Scholar] [CrossRef]
  47. Wu, J.; Wang, X. Effect of enclosure ages on community characters and biomass of the degraded alpine steppe at the northern Tibet. Acta Agrestia Sin. 2017, 25, 261–266. [Google Scholar]
  48. Liu, Y.; Lu, C. Quantifying Grass Coverage Trends to Identify the Hot Plots of Grassland Degradation in the Tibetan Plateau during 2000–2019. Int. J. Environ. Res. Public Health 2021, 18, 416. [Google Scholar] [CrossRef] [PubMed]
Remotesensing 15 04988 g001
Figure 1. The vegetation type (a) and DEM (b) on the QTP in China.
Figure 1. The vegetation type (a) and DEM (b) on the QTP in China.
Remotesensing 15 04988 g001
Remotesensing 15 04988 g002
Figure 2. (a,b) Spatial distributions of correlation coefficient NOAA CDR AVHRR NDVI and MODIS NDVI, NOAA CDR AVHRR NDVI, and GIMMS3g NDVI.
Figure 2. (a,b) Spatial distributions of correlation coefficient NOAA CDR AVHRR NDVI and MODIS NDVI, NOAA CDR AVHRR NDVI, and GIMMS3g NDVI.
Remotesensing 15 04988 g002
Remotesensing 15 04988 g003
Figure 3. Annual NDVI (a) and NDVI in each season (b) from 1982 to 2020. The spring is March to May, the summer is June to August, the autumn is September to November, and the winter is December to February in the next year.
Figure 3. Annual NDVI (a) and NDVI in each season (b) from 1982 to 2020. The spring is March to May, the summer is June to August, the autumn is September to November, and the winter is December to February in the next year.
Remotesensing 15 04988 g003
Remotesensing 15 04988 g004
Figure 4. Spatial distribution of interannual NDVI trends during 1982–2020 on the QTP. The upper panel shows NDVI trend for each 1° longitude bin, right panel shows the longitudinal gradient.
Figure 4. Spatial distribution of interannual NDVI trends during 1982–2020 on the QTP. The upper panel shows NDVI trend for each 1° longitude bin, right panel shows the longitudinal gradient.
Remotesensing 15 04988 g004
Remotesensing 15 04988 g005
Figure 5. The statistical trends of vegetation in different zones: (a) eco-climatic zones; (b) vegetation land covers (agriculture: all kinds of cropland, forest: broad-leaved and needle-leaved tree cover, grassland: mosaic herbaceous covered grassland and grassland, wetland: shrub or herbaceous covered water, settlement: urban, sparse vegetation: liches and sparse vegetation, bare: bare areas, water: water, other forest: tree cover with flooded); and (c) 500 m interval altitude zones, as well as statistical proportions of areas for different regions (d–f).
Figure 5. The statistical trends of vegetation in different zones: (a) eco-climatic zones; (b) vegetation land covers (agriculture: all kinds of cropland, forest: broad-leaved and needle-leaved tree cover, grassland: mosaic herbaceous covered grassland and grassland, wetland: shrub or herbaceous covered water, settlement: urban, sparse vegetation: liches and sparse vegetation, bare: bare areas, water: water, other forest: tree cover with flooded); and (c) 500 m interval altitude zones, as well as statistical proportions of areas for different regions (d–f).
Remotesensing 15 04988 g005
Remotesensing 15 04988 g006
Figure 6. (a,b) Spatial distribution of interannual N D V I h a and N D V I c c trends during 1982–2020. (c,d) Spatial distribution of relative contribution rates in HA and CC to the annual NDVI change on the QTP.
Figure 6. (a,b) Spatial distribution of interannual N D V I h a and N D V I c c trends during 1982–2020. (c,d) Spatial distribution of relative contribution rates in HA and CC to the annual NDVI change on the QTP.
Remotesensing 15 04988 g006
Remotesensing 15 04988 g007
Figure 7. (a,b) Contributions and contribution rates of CC and HA to NDVI change at the city scale.
Figure 7. (a,b) Contributions and contribution rates of CC and HA to NDVI change at the city scale.
Remotesensing 15 04988 g007
Remotesensing 15 04988 g008
Figure 8. (a,b) Temporal trends during 1979–2018 of mean annual temperature and annual precipitation. (c,d) Spatial distributions of correlation coefficient between NDVI and mean annual temperature and annual precipitation on the QTP.
Figure 8. (a,b) Temporal trends during 1979–2018 of mean annual temperature and annual precipitation. (c,d) Spatial distributions of correlation coefficient between NDVI and mean annual temperature and annual precipitation on the QTP.
Remotesensing 15 04988 g008
Remotesensing 15 04988 g009
Figure 9. Spatial distribution of interannual NDVI trends during 2000–2020 on the Naqu, from NOAA CDR AVHRR NDVI (a) and MODIS NDVI (b). The smaller plots at the top right of the figure show the slopes of the annual changes in NDVI for the same region.
Figure 9. Spatial distribution of interannual NDVI trends during 2000–2020 on the Naqu, from NOAA CDR AVHRR NDVI (a) and MODIS NDVI (b). The smaller plots at the top right of the figure show the slopes of the annual changes in NDVI for the same region.
Remotesensing 15 04988 g009
Remotesensing 15 04988 g010
Figure 10. (a,b) GDP and population density over the QTP in 2015. (c,d) Their increase rate during 1995–2019.
Figure 10. (a,b) GDP and population density over the QTP in 2015. (c,d) Their increase rate during 1995–2019.
Remotesensing 15 04988 g010
Remotesensing 15 04988 g011
Figure 11. (a) The population and the production of three industries over the QTP during 1970–2020. NDVI, livestock population, temperature, and precipitation of the Tibet (b,c) and Qinghai provinces (e,f) over time series. (d) The overview of the administrative boundary between Tibet and Qinghai provinces.
Figure 11. (a) The population and the production of three industries over the QTP during 1970–2020. NDVI, livestock population, temperature, and precipitation of the Tibet (b,c) and Qinghai provinces (e,f) over time series. (d) The overview of the administrative boundary between Tibet and Qinghai provinces.
Remotesensing 15 04988 g011
Table 1. Overview of the AVHRR bands.
Table 1. Overview of the AVHRR bands.
BandResolutionWavelengthMain Application
11.09 km0.58–0.68Daytime images, vegetation, snow and ice, climate
21.09 km0.725–1.00Daytime imagery, vegetation, water/road boundaries, agricultural estimate
3A1.09 km1.58–1.64Daytime images, soil moisture, cloud and snow discrimination, drought monitoring
3B1.09 km3.55–3.93Subsurface elevation, nighttime cloud charts, forest fires, volcanic activity
41.09 km10.30–11.30Diurnal images, sea and surface temperature, soil moisture
51.09 km11.50–12.50Diurnal images, sea and surface temperature, soil moisture
Table 2. Different eco-climatic zones shown in Figure 1a
Table 2. Different eco-climatic zones shown in Figure 1a
Temperature RegionArid/HumidEco-Climate Zone
H1. Plateau sub-cold RegionB. Semi-humid regionH1B1 Guoluo-Nagqu alpine shrub meadow zone
C. Semi-arid regionH1C1 South Qinghai alpine steppe zone
H1C2 Qutang alpine steppe zone
H1D1 Kunlun Mountain alpine desert zone
H2. Plateau temperate regionD. Arid regionH2AB1 West Sichuan alpine coniferous forest
AB. Humid/Sub-humid regionH2C1 East Tibet-Qilian alpine coniferous/steppe
C. Semi-arid regionH2C2 South Tibet alpine shrub steppe zone
H2D1 Qaidam Basin desert zone
H2D3 Ngari mountains desert zone
H3. Subtropical broadleaf evergreen forest regionH2D2 Kunlun Mountain north desert zone
VA5: Yunnan Plateau, VA6: Southern flank of the Eastern Himalaya. Data source: China’s Eco-geographical Region, Map 2008 Major.
Table 3. Calculation of the relative contribution rates of CC and HA to NDVI change.
Table 3. Calculation of the relative contribution rates of CC and HA to NDVI change.
FNDVIFCCFHARelative Contribution Rate (%)
CCHA
>0>0>0 F c c F N D V I × 100 % F H A F N D V I × 100 %
>0<0100%0
<0>00100
<0<0<0 F c c F N D V I × 100 % F H A F N D V I × 100 %
>0<00100%
<0>0100%0
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Liu, Y.; Xie, Y.; Guo, Z.; Xi, G. Effects of Climate Variability and Human Activities on Vegetation Dynamics across the Qinghai–Tibet Plateau from 1982 to 2020. Remote Sens. 2023, 15, 4988. https://doi.org/10.3390/rs15204988

AMA Style

Liu Y, Xie Y, Guo Z, Xi G. Effects of Climate Variability and Human Activities on Vegetation Dynamics across the Qinghai–Tibet Plateau from 1982 to 2020. Remote Sensing. 2023; 15(20):4988. https://doi.org/10.3390/rs15204988

Chicago/Turabian Style

Liu, Yiyang, Yaowen Xie, Zecheng Guo, and Guilin Xi. 2023. "Effects of Climate Variability and Human Activities on Vegetation Dynamics across the Qinghai–Tibet Plateau from 1982 to 2020" Remote Sensing 15, no. 20: 4988. https://doi.org/10.3390/rs15204988

APA Style

Liu, Y., Xie, Y., Guo, Z., & Xi, G. (2023). Effects of Climate Variability and Human Activities on Vegetation Dynamics across the Qinghai–Tibet Plateau from 1982 to 2020. Remote Sensing, 15(20), 4988. https://doi.org/10.3390/rs15204988

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

Article Metrics

Back to TopTop