Next Article in Journal
Optimized Plant Diversity and Carbon Storage for Priority Protection Areas in China
Next Article in Special Issue
Estimation of Quercus Biomass in Shangri-La Based on GEDI Spaceborne Lidar Data
Previous Article in Journal
TSBA-YOLO: An Improved Tea Diseases Detection Model Based on Attention Mechanisms and Feature Fusion
Previous Article in Special Issue
Automatic Detection and Classification of Dead Nematode-Infested Pine Wood in Stages Based on YOLO v4 and GoogLeNet
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Forests
Volume 14
Issue 3
10.3390/f14030620
forests-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

Spatio-Temporal Variation and Climatic Driving Factors of Vegetation Coverage in the Yellow River Basin from 2001 to 2020 Based on kNDVI

by
Xuejuan Feng
,
Jia Tian
*,
Yingxuan Wang
,
Jingjing Wu
,
Jie Liu
,
Qian Ya
and
Zishuo Li
School of Agriculture, Ningxia University, Yinchuan 750021, China
*
Author to whom correspondence should be addressed.
Forests 2023, 14(3), 620; https://doi.org/10.3390/f14030620
Submission received: 10 February 2023 / Revised: 15 March 2023 / Accepted: 18 March 2023 / Published: 20 March 2023
(This article belongs to the Special Issue Forestry Remote Sensing: Biomass, Changes and Ecology)
Browse Figures
Review Reports Versions Notes

Abstract

:
The Yellow River Basin (YRB) is a fundamental ecological barrier in China and is one of the regions where the ecological environment is relatively fragile. Studying the spatio-temporal variations in vegetation coverage in the YRB and their driving factors through a long-time-series vegetation dataset is of great significance to eco-environmental construction and sustainable development in the YRB. In this study, we sought to characterize the spatio-temporal variation in vegetation coverage and its climatic driving factors in the YRB from 2001 to 2020 by constructing a new kernel normalized difference vegetation index (kNDVI) dataset based on MOD13 A1 V6 data from the Google Earth Engine (GEE) platform. Using Theil–Sen median trend analysis, the Mann–Kendall test, and the Hurst exponent, we investigated the spatio-temporal variation characteristics and future development trends of the vegetation coverage. The climatic driving factors of vegetation coverage in the YRB were obtained via partial correlation analysis and complex correlation analysis of the associations between kNDVI and both temperature and precipitation. The results reveal the following: The spatial distribution pattern of kNDVI in the YRB showed that vegetation coverage was high in the southeast and low in the northwest. Vegetation coverage fluctuated from 2001 to 2020, with a main significant trend of increasing growth at a rate of 0.0995/5a. The response of vegetation to climatic factors was strong in the YRB, with a stronger response to precipitation than to temperature. Additionally, the main driving factors of vegetation coverage in the YRB were found to be non-climatic factors, which were mainly distributed in Henan, southern Shaanxi, Shanxi, western Inner Mongolia, Ningxia, and eastern Gansu. The areas driven by climatic factors were mainly distributed in northern Shaanxi, Shandong, Qinghai, western Gansu, northeastern Inner Mongolia, and Sichuan. Our findings have implications for ecosystem restoration and sustainable development in the YRB.

1. Introduction

Vegetation serves as the basis for terrestrial ecosystems and is a crucial element in the carbon cycle and water regulation [1]. Vegetation growth not only depends on natural factors but is also sensitive to anthropogenic factors [2], so it is crucial to study the patterns in spatio-temporal variation and the driving factors of the vegetation growth. The Yellow River Basin (YRB) is a crucial ecological zone and one of the most significant grain-producing regions in China. However, the YRB has a fragile ecosystem and suffers significant environmental issues [3,4]. Affected by natural factors such as topography, precipitation, and land-use/cover change (LUCC), areas with soil and water loss account for 62% of the YRB, also impacting on the degradation of vegetation [5]. Some human activities, such as overgrazing and water extraction, also make the vegetation sensitive [6]. Moreover, several studies have shown that biotic and abiotic disturbances can obtain varying degrees of fluctuation in the vegetation coverage [7,8]. The driving factors of vegetation coverage in the basin show obvious heterogeneity in different topographical units, with vegetation coverage decreasing in some regions [9]. In conclusion, the restoration and management of vegetation in the YRB is related to natural and social factors. Therefore, it is crucial to monitor the vegetation dynamics and driving factors in the YRB. The findings of this study can serve as a theoretical foundation for managing vegetation, soil erosion, and ecological restoration in the YRB.
Some traditional research methods used to study vegetation coverage, such as ecological footprint and index evaluation methods, have certain disadvantages. Because of limitations due to terrain and other natural conditions, obtaining data is difficult, time-consuming, and inefficient [10,11,12]. With the rapid development of remote-sensing technology, studies on the large-scale and long-term vegetation dataset based on the Google Earth Engine platform (GEE) have become a hot topic in recent years [13,14]. Vegetation indexes (VIs) can be used to study the coverage and growth status of vegetation [15,16]. VIs are convenient for large-scale vegetation research and provide a corresponding theoretical basis for ecological restoration and vegetation management. As the two most commonly used VIs, studies based on the normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI) have made some achievements in characterizing the spatio-temporal variation of vegetation coverage in China and even globally [17,18,19,20]. However, the NDVI has some defects itself, the most prominent being errors in the processing of atmospheric noise, soil background, and saturation [21]. There have been attempts to compensate for these problems with the EVI by using information from other bands, but the problem of saturation still remains. Thus, CampsValls et al. [22] introduced a new vegetation index called the kernel normalized difference vegetation index (kNDVI) in 2021. Different from the advantages of improving VIs before, kNDVI was proposed by using the principle of machine learning and applying the kernel-method theory to the extraction and calculation of NDVI [23,24]. CampsValls et al. evaluated and compared the performance of the kNDVI with the NDVI and near-infrared reflectance vegetation (NIRv). According to their findings, kNDVI performed better than NDVI and NIRv in all applications, including forestlands, biomes, and climatic zones. The kNDVI is better at handling saturation effects, complex phenological cycles, and seasonal variations, together with solving the mixed-pixel issue.
Moreover, the kNDVI has been used as a suitable VI to gradually represent the state of vegetation coverage on natural and agricultural systems [25,26,27,28]. Previous studies on dynamic changes in vegetation have found that the vegetation coverage in most areas of the YRB shows a tendency of significant increase. However, the vegetation coverage in some regions still shows a tendency to degrade [29,30,31]. Many scholars have analyzed the coupling relationship of vegetation coverage with environmental factors in the YRB [32]. However, there have been few studies on the driving factors of vegetation, particularly regarding their distribution. In addition, most studies on vegetation coverage in the YRB are based on the NDVI and EVI [33,34], and few are based on the kNDVI at present. Thus, the spatial distribution of vegetation coverage in the YRB and its influencing mechanism remain unclear.
In this study, we sought to determine whether the kNDVI can accurately reflect spatio-temporal variation and its driving factors in the YRB by establishing a kNDVI dataset using MOD13 A1 V6 data from 2001 to 2020 on the GEE platform. Using the methods of Theil–Sen median trend analysis (Theil–Sen slope analysis), the Mann–Kendall (MK) test, and the Hurst exponent, we analyzed the spatial and temporal variation of vegetation coverage in the YRB and its future development trends. Partial correlation analysis and complex correlation analysis were used to study the climatic driving factors of vegetation coverage in the YRB. Our research conclusions can provide constructive recommendations for sustainable development, ecological protection, and restoration in the YRB.

2. Study Area

The Yellow River, which originates from the Bayan Har Mountains and merges into the Bohai Sea, is the second-longest river in China. It passes through the provinces of Qinghai, Sichuan, Gansu, Ningxia, Inner Mongolia, Shaanxi, Shanxi, Henan, and Shandong as it flows from west to east (Figure 1a). The YRB is situated between 95°53′ E–119°12′ E and 32°9′ N–41°50′ N, with a basin area of 7.5 × 105 km2 [35]. This region includes four geomorphic units that span the Inner Mongolian Plateau, the Huang-Huai-Hai Plain, the Tibetan Plateau, and the Loess Plateau [36]. The climate in most areas of the YRB is semi-arid or arid, with less than 450 mm of yearly precipitation on average and little natural water supply. The geography of the YRB region is variable, with higher elevations in the west and lower elevations in the east (Figure 1b). The average annual temperature in the YRB ranges from −3.5 to 15 °C [37,38]. Grassland, cropland, woodland, barren, and sparse vegetation are the major land use types in the YRB (Figure 1c). However, with the increasingly serious environmental problems and the overexploitation of water resources, the YRB has become one of the most vulnerable areas of China’s ecological environment [39].

3. Materials and Methods

3.1. Data and Data Collection

The MOD13 A1 V6 dataset (https://lpdaac.usgs.gov/products/mod13a1v006/, accessed on 3 July 2022) was used for vegetation coverage data in the YRB, with 500 m spatial resolution and 16-day temporal resolution. Through calculating NDVI data on the GEE platform pixel-by-pixel, we finally acquired 23 kNDVI images per year from 2001 to 2020. The annual median kNDVI pixels in the images were used to represent the overall condition of the vegetation in the related year during 2001–2020. The formula for calculating kNDVI is as follows:
k N D V I = t a n h N I R − R e d 2 σ 2
where σ is a length-scale parameter that can be adjusted to capture the NDVI’s nonlinear sensitivity to vegetation density. With the generalization σ = 0.5(NIR + Red), the formula for kNDVI is as follows:
k N D V I = t a n h ⁡ N D V I 2
The temperature data utilized in this research were from a 1 km monthly mean temperature dataset for China covering the period 1901–2020 [40], while the precipitation data were from a 1 km monthly precipitation dataset for China covering the same time period [41]. Both the temperature and precipitation data were supplied by the China Tibetan Plateau Data Science Center (http://data.tpdc.ac.cn/zh-hans/, accessed on 1 August 2022), with 1000 m spatial resolution and 30-day temporal resolution. We calculated the average temperature and total precipitation from 2001 to 2020 on the PIE-Engine platform (PIE) (https://engine.piesat.cn/, accessed on 4 August 2022), and 20 images of temperature and 20 images of precipitation from 2001 to 2020 were obtained. Finally, these images were uploaded onto the GEE platform. The data for the vector boundaries came from the National Earth System Science Data Center (http://www.geodata.cn/, accessed on 9 October 2022). The elevation data were obtained from CMR SEARCH (https://cmr.earthdata.nasa.gov/, accessed on 11 October 2022), with a spatial resolution of 30 m. The land-cover data were from WorldCover (https://esa-worldcover.org/, accessed on 11 October 2022), with a spatial resolution of 10 m.

3.2. Methods

Based on the kNDVI and climatic data, the patterns in vegetation variation were determined using Theil–Sen slope analysis and the MK test method. Then, the Hurst exponent enabled us to evaluate the sustainability of vegetation dynamics in the region. Finally, the partial correlation analysis and complex correlation analysis methods were used to analyze the response of vegetation to climatic parameters and their driving factors (Figure 2).

3.2.1. Spatial–Temporal Variation and Future Trend Analysis

A combination of Theil–Sen slope analysis and MK testing is gradually used to analyze long-term vegetation sequences that reflect trends in each pixel in a time series [42,43]. Theil–Sen slope analysis is highly computationally efficient and insensitive to measurement error and discrete data [44]. Hence, Theil–Sen slope analysis was applied to detect the trend in the kNDVI in the YRB during the period 2001–2020 at the pixel level. MK testing has the advantage of not requiring samples to conform to a specific distribution and is free of outliers’ interference, and a linear trend does not necessarily require essentiality [45]. The test method described has gained broad usage in assessing the significance of trends in long-term data series. We chose the MK test to measure the significance of the vegetation trend.
By combining the trend of the kNDVI with the Hurst exponent, future trends in vegetation can be predicted [46]. In this study, the Hurst exponent values (H) were split into three circumstances: when H > 0.5, the kNDVI time series sustainable development exhibits the same pattern as the kNDVI time series in the future; when H = 0.5, the kNDVI time series is a stochastic series with no sustainability; and when H < 0.5, the kNDVI time series is anti-sustainable, representing the opposite trend of the kNDVI time series in the future [29]. The SkNDVI and ZS formulas are as follows:
S k N D V I = M e d i a n ⁡ k N D V I j − k N D V I i j − i , 2001 ≤ i < j ≤ 2020
Z s = S − 1 v a r ⁡ ( S ) , S > 0 ; 0 , S = 0 S + 1 v a r ⁡ ( S ) , S < 0 ;
Where ,   S = ∑ i = 1 n − 1   ∑ j = i + 1 n   s g n ⁡ x j − x i
V a r ⁡ ( S ) = n ( n − 1 ) ( 2 n + 5 ) − ∑ i = 1 m   t i t i − 1 2 t i + 5 18
where, SkNDVI signifies the slope value as determined by the Theil–Sen slope analysis; kNDVIi and kNDVIj are the kNDVI values of the pixels i and j, respectively; the ZS statistic is valued in the range (−∞, +∞); Z represents the standardized test statistic; sgn stands for a sign function; n is the duration of the kNDVI time series; m is the number of recurring data sets in the data; and ti is the number of repeats in the extent i. A given significance level, |ZS| > u1−α/2, indicates that the time series exhibits a significant change at the level α. In general, the value of α is 0.05. In this study, we chose α = 0.05, it indicates that we used a confidence level of 0.05 to determine the significance of the kNDVI trend at the pixel level.

3.2.2. Driving Factor Analysis

Correlation analysis is frequently used in the investigation of vegetation coverage driving factors. In this study, the level of closeness between the kNDVI and climatic factors (average temperature and total precipitation) was investigated using partial correlation analysis in the YRB from 2001 to 2020. The magnitude of the value for the partial correlation coefficient (PCC) can reflect the degree of influence that a climatic factor has on vegetation, and the significance was tested using a t-test. The simultaneous effect of both temperature and precipitation on the kNDVI were studied via complex correlation analysis. Values of the complex correlation coefficient (CCC) were estimated to evaluate the relationship between the kNDVI and both temperature and precipitation, wherein the significance was assessed via the F-test [47].
Because of the spatial heterogeneity of the PCC and CCC between the kNDVI and climatic factors, the spatial distribution of climatic driving factors for vegetation change in the YRB was regionalized and summarized on the basis of the t-test and F-test. To guarantee maximal consistency, spatial non-repeatability, and regional continuity for every classification, pixels that passed the F-test at α = 0.05 were extracted for additional climate-driven regionalization, and other pixels were regarded as being influenced by non-climate factors. Additionally, the t-test results between the kNDVI and each climatic parameters were used to categorize the climate-driven factors into three groups. This approach of partitioning has been widely used for the driving factors of vegetation coverage [48,49]. The basis for classification is shown in Table 1.

4. Results

4.1. Spatio-Temporal Characteristics of the kNDVI

4.1.1. The Spatial Distribution Characteristics of Vegetation Coverage

The characteristics of the median kNDVI’s spatial distribution during the course of the study’s 20-year period are shown in Figure 3a, and were calculated using the median kNDVI data for the YRB from each year (2001–2020). As shown in this figure, the spatial distribution of kNDVI in the YRB indicates that vegetation coverage was high in the southeast, while it was low in the northwest. The average kNDVI in the whole basin was 0.082, and the kNDVI value ranged from 0 to 0.53 for the whole basin. The average kNDVI values in the upper, middle, and lower reaches of the YRB were 0.047, 0.12, and 0.15, respectively. In comparison to the upper reaches, the kNDVI value was comparatively greater in the middle and lower reaches of the YRB. Using the natural breakpoint method, the results of the statistical regimentations of the median kNDVI values and the proportions of each category over the 20 years of the YRB are shown in Figure 3b. The regions with kNDVI values less than 0.03 for the largest proportion of the YRB, which were classified as low vegetation coverage, were mainly distributed in some fragile regions such as the Kubuqi Desert, Mu Us Sandy Land, and some areas of the Loess Plateau. The regions with kNDVI values between 0.03 and 0.07 were deemed to have a median amount of vegetation and were mainly distributed in eastern Inner Mongolia, central and western Gansu province, northern Shaanxi province, northern Shanxi province, southern Ningxia province, and some regions of Qinghai province. Areas with kNDVI values ranging from 0.07 to 0.17 were classified as having high vegetation coverage, while those with values greater than 0.17 were classified as having extremely high vegetation coverage. These areas were mainly distributed in Henan province, Shandong province, southern Shaanxi province, and central and southern Shanxi province.

4.1.2. Temporal Variation Characteristics of the Vegetation Coverage

To study the temporal variation characteristics of kNDVI in the YRB, the annual median kNDVI pixels in the images were used to represent the overall condition of the vegetation in the related year during 2001–2020. As shown in Figure 4, the annual distribution of the kNDVI values is shown in the box plot, while the line chart shows inter-annual variation in the kNDVI in the YRB from 2001 to 2020. The kNDVI value in the YRB indicates a trend of fluctuating upward at a rate of 0.0995/5a. The range of kNDVI values in individual years shows wide variation. The annual kNDVI in the YRB fluctuated from 0.034 to 0.077. The highest value was 0.076 in 2020, while the minimum value was 0.034 in 2001. The total value of kNDVI increased by 0.042 over the 20 years. The overall growth of vegetation coverage in the basin shows an increasing trend from 2001 to 2020.

4.1.3. Spatial Variation Characteristics of Vegetation Coverage

The trends in the spatial distribution of kNDVI variation in the YRB from 2001 to 2020 were analyzed for each pixel using Theil–Sen slope analysis and MK testing. Since pixels with a SkNDVI of 0 strictly did not exist, we consulted other scholars’ categorization techniques and established the following classifications in accordance with the actual SkNDVI conditions [29,50]: pixels with −0.0005 < SkNDVI < 0.0005 were categorized as areas of stable vegetation, pixels with SkNDVI ≥ 0.0005 were categorized as areas of vegetation-increased, and pixels with SkNDVI ≤ −0.0005 were categorized as areas of vegetation degradation. We used the MK test to evaluate the significance of kNDVI trends at the confidence level of 0.05. Variations were defined as significant if the test results (ZS) were greater than 1.96 or less than −1.96 and insignificant if −1.96 < ZS < 1.96. By combining the results of Theil–Sen slope analysis and the MK test, we obtained the spatial distribution of inter-annual kNDVI trend variations in the YRB at the pixel scale. As shown in Table 2, the results were recategorized into five categories, and each category’s area proportion was calculated: the areas with improved vegetation coverage accounted for 75.89%, the areas with stable vegetation coverage accounted for 22.05%, and the areas with degraded vegetation coverage accounted for 2.06%.
As shown in Figure 5 for the YRB from 2001 to 2020, in comparison to areas with decreased vegetation coverage, the areas with increased vegetation coverage were noticeably greater. The areas with improved vegetation coverage were mainly distributed in Shaanxi, Shanxi, Henan, Shandong, northern Qinghai province, Gansu province, central and southern Ningxia, Sichuan, and eastern and southern Inner Mongolia. The areas with vegetation stabilization were mainly distributed in Ningxia Plain, Hetao Plain in Inner Mongolia, and western Qinghai province. A small percentage of areas showed vegetation degradation and were mainly distributed in Hetao Plain in Inner Mongolia, Taiyuan Basin in Shanxi province, Guanzhong Basin in Shaanxi province, northwest Henan province, central Qinghai province, eastern Gansu province, southern Shaanxi province, and southern Shanxi province.

4.1.4. Future Development Trends of the Vegetation Coverage in the YRB

As shown in Figure 6a, the average H value of the kNDVI in the YRB was 0.656. The areas with an H value less than 0.5 accounted for 4.59%, and the areas with a H value greater than 0.5 accounted for 95.41%. The results show that the kNDVI in the YRB has strong sustainability. To demonstrate the variability in trends and the sustainability of vegetation coverage, the kNDVI variation in trends was overlaid with the H value to get the coupled information of variation in trends and sustainability. As shown in Figure 6b, the coupling consequences were split into six scenarios, where the category of undetermined future trends included any combination of unsustainability and kNDVI trends from 2001 to 2020 in areas with H values less than 0.5.
The spatial distribution of future trends for kNDVI in the YRB is displayed in Figure 6b. The areas presenting “sustainability and improvement” accounted for 73.20% of the total and were majorly distributed in the middle and lower reaches of the YRB, and some regions in the upper reaches of the YRB such as Gansu province, Sichuan province, northern Qinghai province, southern Ningxia, and southeastern Inner Mongolia. The areas in the “sustainability and stability” category accounted for 20.30% and were mainly distributed in northern Ningxia, northwestern Inner Mongolia, and western Qinghai province. Only 1.92% of the areas in the whole basin showed “sustainability and degradation”, of which 0.71% showed “sustainability and severe degradation”. These areas were mainly distributed in Hetao Plain in Inner Mongolia, Guanzhong Basin in Shaanxi province, Taiyuan Basin in Shanxi province, central Henan province, central and western Qinghai province, southern Shaanxi province, and southern Shanxi province. The areas with undetermined variation in trends in the future accounted for 4.58%, and were mainly distributed in Qinghai province, Sichuan province, southern and eastern Gansu province, northern Inner Mongolia, southern Shaanxi province, southern Shaanxi province, and Henan province.

4.2. Driving Factors of Vegetation Coverage

4.2.1. Spatio-Temporal Variation Characteristics of the Meteorological Factors

In order to investigate the temporal variation characteristics of the climatic factors in the YRB from 2001 to 2020, the temperature and precipitation of pixels in the YRB in the images of average temperature and total precipitation were taken to denote the entire situation of the climate in the related year. The average temperature and average precipitation from 2001 to 2020 were used to represent the spatial distribution characteristics of the climate in the YRB. As suggested in Figure 7a,b, temperature and precipitation in the YRB exhibited a fluctuating increasing tendency. Annual average temperature increased at a rate of approximately 0.058 °C/5a, and annual total precipitation increased at a rate of 14.725 mm/5a. Consequently, the YRB’s climate exhibited a particular warm and humid trend. In order to clearly illustrate the spatial distribution of temperature and precipitation, the natural breakpoint method was used to divide temperature and precipitation into 10 categories. As shown in Figure 7c,d, the average temperature and total precipitation in the YRB were 6.66 °C and 480.80 mm from 2001 to 2020. The temperature ranged from −18.57 to 15.96 °C and progressively increased from west to east. The precipitation ranged from 110 to 902 mm and progressively increased from northwest to southeast. Both the temperature and precipitation in the YRB had significant spatial heterogeneity.

4.2.2. Partial Correlation Analysis of Relationship between kNDVI and Climatic Factors

The spatial distribution of PCC between kNDVI and climatic factors was classified using the breakpoint method. During 2001–2020, the PCC between kNDVI and temperature was in the range of −1 to 1, with an average value of 0.15 (Figure 8a). Positively correlated pixels (76.41%) were more universal than negatively correlated pixels (23.59%), while “significant positive correlation” and “significant negative correlation” pixels accounted for 12.87% and 0.61%, respectively (Figure 8b). As shown, the spatial distribution patterns of the PCC between kNDVI and temperature exhibited marked heterogeneity. The significantly positively correlated pixels were mainly concentrated in Shandong province, Shanxi province, eastern Inner Mongolia, eastern Henan province, northeastern Shaanxi province, Qinghai province, and Sichuan province. The negatively correlated pixels were mainly distributed in eastern Inner Mongolia, central and southern Shaanxi, central Ningxia, Henan, Qinghai, and eastern Gansu, while the significantly negatively correlated pixels were mainly concentrated in central and southern Shaanxi province and central Ningxia.
The PCC between kNDVI and precipitation was also in the range of −1 to 1, with an average value of 0.21 (Figure 8c). Positively correlated pixels (80.53%) were more universal than negatively correlated pixels (19.47%). As shown in Figure 8d, the significantly positively correlated pixels were mainly distributed in Shaanxi province, central and eastern Qinghai province, Sichuan province, central and western Gansu province, central Ningxia, and eastern and southeastern Inner Mongolia, while the significantly negatively correlated and negatively correlated pixels were mainly concentrated in the west of Qinghai province, Hetao Plain of Inner Mongolia, south of Shaanxi province, Taiyuan Basin of Shaanxi province, and other small regions. Overall, the responses of vegetation to climatic factors was strong in the YRB, with a stronger response to precipitation than to temperature.

4.2.3. Analysis of Driving Factors of Vegetation Coverage

The CCC between the kNDVI and climatic factors was classified using the breakpoint method and is visualized in Figure 9a, while Figure 9b depicts the significance level of the CCC. The CCC ranged from 0 to 1 in the YRB, and its average value was 0.38. The pixels of the CCC that passed the 0.05 significance test accounted for only 17.40%.
Based on the PCC and CCC between the kNDVI and climatic factors during the period 2001–2020, we found that there was spatial and temporal heterogeneity in the YRB, and we regionalized the vegetation coverage rules for climatic driving factors. The spatial distribution of the driving factors is shown in Figure 9c. As shown, only 17.40% of the vegetation coverage in the YRB was driven by climate factors, while 82.60% was driven by non-climate factors. Of the areas driven by climate factors, 6.78% were found to be driven by precipitation and were distributed in eastern Qinghai province, western and northern Gansu province, Sichuan province, central Ningxia, southern and eastern Inner Mongolia, northern Shaanxi province, and other small areas. The temperature-driven regions accounted for 10.02% and were mainly distributed in eastern and northern Qinghai province, eastern Inner Mongolia, northeastern Shaanxi province, northern and central Shaanxi province, and Shandong province. The smallest region, driven by both temperature and precipitation, accounted for only 0.60%, and areas were mainly distributed in Qinghai province, Gansu province, and Shanxi province. In general, we found that vegetation in the YRB was mainly driven by non-climatic factors and was widely distributed in the provinces of the YRB.

5. Discussion

5.1. Spatio-Temporal Change in kNDVI

In this study, we analyzed the temporal and spatial variation of the kNDVI in the YRB from 2001 to 2020. The spatial distribution of the kNDVI indicated that vegetation coverage was high in the southeast of the YRB (Figure 3a), which was due to the humid climate and abundant precipitation, and the types of vegetation were mainly crops in addition to trees and shrubs [51]. The ecological environment was found to be relatively fragile in the northwest areas, so the kNDVI was lower. These places were mainly restricted by environmental factors. For example, Inner Mongolia and Ningxia have relatively little precipitation, while Qinghai has a high altitude and insufficient hydrothermal conditions that are not conducive to large-scale vegetation growth [52]. Compared with existing findings of previous studies, the vegetation coverage data generally still show a stable and increasing trend on the whole (Figure 4) [53]. The vegetation coverage with “significant improvement” represented the largest proportion of the possible classifications in the YRB, and the future trend in the vegetation coverage with “sustainability and obvious improvement” classification also represented the largest proportion (Figure 5 and Figure 6b). These consequences are consistent with those of Yuan et al. [29]. However, the proportion of the vegetation coverage that showed “sustainability and severe degradation” had decreased significantly, indicating that vegetation coverage in the YRB has significantly increased in recent years. This is mainly thanks to the implementation of forestry engineering projects, agricultural irrigation and water conservancy technology, the policy of returning farmland to forest and grassland, and other projects in recent years [54,55,56]. From this viewpoint, the Grain for Green Project, the implementation of the Chinese government’s Natural Forest Protection Project, and related forest protection policies can be deemed a success. According to data on the National Forestry and Grassland Administration (http://www.forestry.gov.cn/, accessed on 5 October 2022), the construction of plantation forests in the basin, with the aim of improving the ecological environment in the YRB, is still advancing year by year. However, there were still some regions showing “sustainability and degradation” or “undetermined variation in trends in the future”. Natural, human, or other factors have a significant negative impact on the vegetation at these locations [57]. For example, because of the high altitude, the temperature was lower in Qinghai province than other provinces in the YRB, so the environmental conditions in these areas were not conducive to the growth of trees and shrubs [58]. As can be seen from the land-use remote sensing image (Figure 1c), vegetation coverage is mainly dominated by grassland. Other studies have also discovered that ecological restoration can be promoted by adopting measures such as grassland cultivation and artificial afforestation, in addition to mountain closures for forest and grassland cultivation [59], so vegetation management should be strengthened in these areas using specific measures to promote ecological restoration. Generally speaking, compared to the upper reaches, the vegetation coverage in the middle and lower reaches of the YRB is more stable.

5.2. Driving Factor Analysis of kNDVI

Analysis of the driving factors of vegetation coverage showed that non-climatic factors overwhelmingly dominate in impacting on vegetation coverage in the YRB (Figure 9c). In contrast, the dominant factors of kNDVI in different regions varied. Existing research indicates that the main driving factors of vegetation coverage include human activities, precipitation and temperature [60], and in areas with less human activity, changes in vegetation coverage are mostly caused by meteorological factors [61]. From the spatial distribution analysis of driving factors, it was found that the distribution of climatic factors in the YRB showed significant spatial heterogeneity. Most of the regions found to be driven by precipitation were located in arid or semi-arid regions with less precipitation. The regions driven by temperature have relatively low temperatures. The regions driven by both precipitation and temperature, however, were characterized by high elevation; for example, Qinghai province and western Gansu province. This is because elevation and terrain can affect the regulation of temperature, humidity, and light, which indirectly influence vegetation coverage [62]; this agrees with the finding reached by Deng et al. [63]. Afforestation should be combined with climate characteristics, and corresponding cold- and drought-prevention measures should be formulated in areas affected by climate factors. Linking climate-driven regional vegetation management to rainfall and temperature as much as possible can provide a theoretical basis for the improvement and management of vegetation. However, the local climate, hydrology, and other factors, such as the impact of excessive tree planting, soil moisture transpiration, and the reduction of water resources in the created area, which do not favor improving vegetation coverage, should be taken into consideration during the construction process [5].

5.3. Limitations and Future Work

In this study, we analyzed the variation trends in the vegetation coverage, predicted the sustainability of vegetation dynamics, and explored the spatial distribution of vegetation coverage driven by climatic factors in the YRB. These findings can offer information for efficient monitoring of vegetation change as well as a theoretical foundation for vegetation preservation and restoration in the YRB. However, there were some limitations. Firstly, we were able to quantify future vegetation trends by superimposing the results of the kNDVI trend and H value. However, the Hurst exponent was unable to foresee the duration of the sustainability of vegetation dynamics; thus it is necessary to develop techniques to ascertain the length of the trend. In addition, the annual median value of the kNDVI was low in individual years, so the disturbance years and disturbance areas of vegetation coverage from 2001 to 2020 could be calculated to further analyze the driving factors of vegetation coverage. Finally, As climatic driving parameters of vegetation coverage, we solely selected temperature and precipitation. According to some studies, the growth of vegetation is significantly influenced by climatic conditions on a broad scale and also controlled by terrain on a relatively small scale [64]. Some other natural factors such as soil moisture, soil type, runoff, and evaporation also have an effect on vegetation coverage [65,66]. Therefore, future research should take into account these natural aspects that were not taken into account in this study. Moreover, vegetation coverage driven by non-climatic factors occupied the biggest proportion of the YRB and the intricate process of vegetation coverage over a long time series that is susceptible to non-climatic influences including urban expansion, project construction, grazing, and LUCC [49]. Hence, future research should aim to incorporate these human-induced factors into the analysis for a clearer understanding of the spatial distribution of the various non-climate driving factors.

6. Conclusions

In this study, the kernel normalized difference vegetation index (kNDVI) was calculated using MOD13 A1 V6 data on the GEE platform. The annual median kNDVI from 2001 to 2020 was used to assess the overall vegetation conditions in each year. The spatial and temporal characteristics of the kNDVI in the YRB were evaluated, and its response to climatic factors and driving factors were studied using time series data in respect of temperature and precipitation.
The results show that vegetation coverage in the YRB was high in the southeast and low in the northwest, with significant spatial heterogeneity. The annual median kNDVI fluctuated from 0.034 to 0.077 from 2001 to 2020, showing an upward trend with a rate of 0.0995/5a. Improvement was observed in 75.89% of the vegetation coverage, 22.05% remained unchanged, and 2.06% showed degradation. We found that 95.41% of the vegetation coverage was sustainable, with 73.20% showing “sustainability and improvement” and 6.50% showing “sustainability and degradation” or “unpredictable future trend”. Vegetation responses to climatic factors were strong in the YRB, with a stronger response to precipitation than to temperature. At the 0.05 confidence level, 82.60% of the vegetation coverage was found to be driven by non-climatic factors, primarily distributed in Henan, southern Shaanxi, Shanxi, western Inner Mongolia, Ningxia, and eastern Gansu, while 17.40% was found to be driven by climatic factors, primarily in northern Shaanxi, Shandong, Qinghai, western Gansu, northeast Inner Mongolia, and Sichuan.

Author Contributions

Conceptualization, X.F. and J.T.; methodology, X.F. and Y.W.; validation, X.F., J.T. and Y.W.; formal analysis, X.F. and J.W.; investigation, X.F., J.L. and Z.L.; resources, X.F., J.T. and Q.Y.; data curation, X.F. and Y.W.; writing—original draft preparation, X.F.; writing—review and editing, X.F. and J.T.; visualization, X.F.; supervision, J.T.; project administration, J.T.; funding acquisition, J.T. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Natural Science Foundation of China (Project No. 31960330).

Data Availability Statement

The data presented in this study are available on reasonable request from the corresponding author.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Chen, G.; Huang, Y.; Chen, J.; Wang, Y. Spatiotemporal Variation of Vegetation Net Primary Productivity and Its Responses to Climate Change in the Huainan Coal Mining Area. J. Indian Soc. Remote Sens. 2019, 47, 1905–1916. [Google Scholar] [CrossRef]
  2. Zheng, K.; Tan, L.; Sun, Y.; Wu, Y.; Duan, Z.; Xu, Y.; Gao, C. Impacts of climate change and anthropogenic activities on vegetation change: Evidence from typical areas in China. Ecol. Indic. 2021, 126, 107648. [Google Scholar] [CrossRef]
  3. Zhang, X.; Wang, G.; Xue, B.; Yinglan, A. Changes in vegetation cover and its influencing factors in the inner Mongolia reach of the yellow river basin from 2001 to 2018. Environ. Res. 2022, 215, 114253. [Google Scholar] [CrossRef] [PubMed]
  4. Zhan, C.; Liang, C.; Zhao, L.; Jiang, S.; Niu, K.; Zhang, Y. Drought-related cumulative and time-lag effects on vegetation dynamics across the Yellow River Basin, China. Ecol. Indic. 2022, 143, 109409. [Google Scholar] [CrossRef]
  5. Zhang, W.; Wang, L.; Xiang, F.; Qin, W.; Jiang, W. Vegetation dynamics and the relations with climate change at multiple time scales in the Yangtze River and Yellow River Basin, China. Ecol. Indic. 2020, 110, 105892. [Google Scholar] [CrossRef]
  6. Hua, W.; Chen, H.; Zhou, L.; Xie, Z.; Qin, M.; Li, X.; Ma, H.; Huang, Q.; Sun, S. Observational quantification of climatic and human influences on vegetation greening in China. Remote Sens. 2017, 9, 425. [Google Scholar] [CrossRef] [Green Version]
  7. Ye, J.; Wang, N.; Sun, M.; Liu, Q.; Ding, N.; Li, M. A New Method for the Rapid Determination of Fire Disturbance Events Using GEE and the VCT Algorithm—A Case Study in Southwestern and Northeastern China. Remote Sens. 2023, 15, 413. [Google Scholar] [CrossRef]
  8. Stefanidis, S.; Alexandridis, V.; Mallinis, G. A cloud-based mapping approach for assessing spatiotemporal changes in erosion dynamics due to biotic and abiotic disturbances in a Mediterranean Peri-Urban forest. Catena 2022, 218, 106564. [Google Scholar] [CrossRef]
  9. Ren, Y.; Liu, J.; Liu, S.; Wang, Z.; Liu, T.; Shalamzari, M.J. Effects of climate change on vegetation growth in the Yellow River Basin from 2000 to 2019. Remote Sens. 2022, 14, 687. [Google Scholar] [CrossRef]
  10. Tangen, B.A.; Bansal, S.; Jones, S.; Dixon, C.S.; Nahlik, A.M.; DeKeyser, E.S.; Hargiss, C.L.; Mushet, D.M. Using a vegetation index to assess wetland condition in the Prairie Pothole Region of North America. Front. Environ. Sci. 2022, 10, 889170. [Google Scholar] [CrossRef]
  11. Eva, I.; Stephanie, H.; Rasmus, F.; Michael, C. Drought footprint on European ecosystems between 1999 and 2010 assessed by remotely sensed vegetation phenology and productivity. Glob. Chang. Biol. 2014, 20, 581–593. [Google Scholar]
  12. Hideyuki, N.; Mahito, K.; Sin, M.; Midori, O. Assessing the impact of storm surge flooding on coastal pine forests using a vegetation index. Landsc. Ecol. Eng. 2022, 19, 151–159. [Google Scholar]
  13. Liao, W.; Nie, X.; Zhang, Z. Interval association of remote sensing ecological index in China based on concept lattice. Environ. Sci. Pollut. Res. Int. 2022, 29, 34194–34208. [Google Scholar] [CrossRef]
  14. Xia, W.; Xiaojie, Y.; Changzheng, J.; Wei, D. Dynamic monitoring and analysis of factors influencing ecological environment quality in northern Anhui, China, based on the Google Earth Engine. Sci. Rep. 2022, 12, 20307. [Google Scholar]
  15. Faour, G.; Mhawej, M.; Nasrallah, A. Global trends analysis of the main vegetation types throughout the past four decades. Appl. Geogr. 2018, 97, 184–195. [Google Scholar] [CrossRef]
  16. Tian, H.; Ji, X.; Zhang, F. Spatiotemporal Variations of Vegetation Net Primary Productivity and Its Response to Meteorological Factors Across the Yellow River Basin During the Period 1981–2020. Front. Environ. Sci. 2022, 10, 1981–2020. [Google Scholar] [CrossRef]
  17. Alam, S.S.; Qingmin, M. Use time series NDVI and EVI to develop dynamic crop growth metrics for yield modeling. Ecol. Indic. 2020, 121, 107124. [Google Scholar]
  18. Zhao, W.; Luo, T.; Wei, H.; Alamu; Zhang, L. Relative impact of climate change and grazing on NDVI changes in grassland in the Mt. Qomolangma Nature Reserve and Adjacent Regions during 2000–2018. Diversity 2022, 14, 171. [Google Scholar] [CrossRef]
  19. Xiong, Y.; Wang, H. Spatial relationships between NDVI and topographic factors at multiple scales in a watershed of the Minjiang River, China. Ecol. Inform. 2022, 69, 101617. [Google Scholar] [CrossRef]
  20. Jin, H.; Chen, X.; Wang, Y.; Zhong, R.; Zhao, T.; Liu, Z.; Tu, X. Spatio-temporal distribution of NDVI and its influencing factors in China. J. Hydrol. 2021, 603, 127129. [Google Scholar] [CrossRef]
  21. Alencar, A.; Shimbo, J.Z.; Lenti, F.; Marques, C.B.; Zimbres, B.; Rosa, M.; Arruda, V.; Castro, I.; Ribeiro, J.P.F.M.; Varela, V.; et al. Mapping three decades of changes in the Brazilian Savanna Native Vegetation using Landsat data processed in the Google Earth Engine platform. Remote Sens. 2020, 12, 924. [Google Scholar] [CrossRef] [Green Version]
  22. Camps-Valls, G.; Campos-Taberner, M.; Moreno-Martínez, Á.; Walther, S.; Duveiller, G.; Cescatti, A.; Mahecha, M.D.; Muñoz-Marí, J.; García-Haro, F.J.; Guanter, L.; et al. A unified vegetation index for quantifying the terrestrial biosphere. Sci. Adv. 2021, 7, eabc7447. [Google Scholar] [CrossRef] [PubMed]
  23. Luis, R.Á.J.; Manel, M.R.; Jordi, M.M.; Gustau, C.V. Digital Signal Processing with Kernel Methods; John Wiley & Sons, Ltd.: Hoboken, NJ, USA, 2018. [Google Scholar]
  24. Gustavo, C.V.; Lorenzo, B. Kernel Methods for Remote Sensing Data Analysis; John Wiley & Sons, Ltd.: New York, NY, USA, 2009. [Google Scholar]
  25. Gensheimer, J.; Turner, A.J.; Köhler, P.; Frankenberg, C.; Chen, J. A convolutional neural network for spatial downscaling of satellite-based solar-induced chlorophyll fluorescence (SIFnet). Biogeosciences 2022, 19, 1777–1793. [Google Scholar] [CrossRef]
  26. Forzieri, G.; Dakos, V.; McDowell, N.G.; Ramdane, A.; Cescatti, A. Emerging signals of declining forest resilience under climate change. Nature 2022, 608, 534–539. [Google Scholar] [CrossRef] [PubMed]
  27. Liu, Y.; Zhao, W.; Chen, S.; Ye, T. Mapping crop rotation by using deeply synergistic optical and SAR time series. Remote Sens. 2021, 13, 4160. [Google Scholar] [CrossRef]
  28. Wang, X.; Biederman, J.A.; Knowles, J.F.; Scott, R.L.; Turner, A.J.; Dannenberg, M.P.; Köhler, P.; Frankenberg, C.; Litvak, M.E.; Flerchinger, G.N.; et al. Satellite solar-induced chlorophyll fluorescence and near-infrared reflectance capture complementary aspects of dryland vegetation productivity dynamics. Remote Sens. Environ. 2022, 270, 112858. [Google Scholar] [CrossRef]
  29. Jiang, W.; Yuan, L.; Wang, W.; Cao, R.; Zhang, Y.; Shen, W. Spatio-temporal analysis of vegetation variation in the Yellow River Basin. Ecol. Indic. 2015, 51, 117–126. [Google Scholar] [CrossRef]
  30. Cheng, Y.; Zhang, L.; Zhang, Z.; Li, X.; Wang, H.; Xi, X. Spatiotemporal variation and influence factors of vegetation cover in the Yellow River Basin (1982–2021) based on GIMMS NDVI and MOD13A1. Water-Sui. 2022, 14, 3274. [Google Scholar] [CrossRef]
  31. Han, Z.; Wu, Q.; Lai, R.; Hyder, S.S.E.; Hou, D.; Hu, C. Spatio-temporal variations of vegetation cover and its influence on surface air temperature change over the Yellow River Basin, China. J. Water Clim. Chang. 2022, 13, 3239–3252. [Google Scholar] [CrossRef]
  32. Wang, H.; He, L.; Yin, J.; Yu, Z.; Liu, S.; Yan, D. Effects of effective precipitation and accumulated temperature on the terrestrial EVI (Enhanced Vegetation Index) in the Yellow River Basin, China. Atmosphere 2022, 13, 1555. [Google Scholar] [CrossRef]
  33. Han, J.; Zhang, X.; Wang, J.; Zhai, J. Geographic exploration of the driving forces of the NDVI spatial differentiation in the Upper Yellow River Basin from 2000 to 2020. Sustainability 2023, 15, 1922. [Google Scholar] [CrossRef]
  34. Zhang, Y.; Ye, A. Spatial and temporal variations in vegetation coverage observed using AVHRR GIMMS and Terra MODIS data in the mainland of China. Int. J. Remote. Sens. 2020, 41, 4238–4268. [Google Scholar] [CrossRef]
  35. Liu, C.; Zhang, X.; Wang, T.; Chen, G.; Zhu, K.; Wang, Q.; Wang, J. Detection of vegetation coverage changes in the Yellow River Basin from 2003 to 2020. Ecol. Indic. 2022, 138, 108818. [Google Scholar] [CrossRef]
  36. Zhou, X.; Yang, Y.; Sheng, Z.; Zhang, Y. Reconstructed natural runoff helps to quantify the relationship between upstream water use and downstream water scarcity in China’s river basins. Hydrol. Earth Syst. Sc. 2019, 23, 2491–2505. [Google Scholar] [CrossRef] [Green Version]
  37. Sun, P. The effect of a small initial distortion of the basic flow on the subcritical transition in plane Poiseuille flow. Q. Appl. Math. 2001, 59, 667–699. [Google Scholar] [CrossRef] [Green Version]
  38. Wang, X.; Shi, S.; Zhao, X.; Hu, Z.; Hou, M.; Xu, L. Detecting spatially non-stationary between vegetation and related factors in the Yellow River Basin from 1986 to 2021 using multiscale geographically weighted regression based on Landsat. Remote Sens. 2022, 14, 6276. [Google Scholar] [CrossRef]
  39. Yang, Z.; Tian, J.; Su, W.; Wu, J.; Liu, J.; Liu, W.; Guo, R. Analysis of ecological environmental quality change in the Yellow River Basin using the remote sensing based ecological index. Sustainability 2022, 14, 10726. [Google Scholar] [CrossRef]
  40. Peng, S. 1-km Monthly Mean Temperature Dataset for China (1901–2020); National Tibetan Plateau Data Center: Beijing, China, 2019. [Google Scholar]
  41. Peng, S. 1-km Monthly Precipitation Dataset for China (1901–2020); National Tibetan Plateau Data Center: Beijing, China, 2020. [Google Scholar]
  42. Zhang, X.; Liu, Y.; Chen, X.; Long, L.; Su, Y.; Yu, X.; Zhang, H.; Chen, Y.; An, S. Analysis of spatial and temporal changes of vegetation cover and its driving forces in the Huainan mining area. Environ. Sci. Pollut. Res. 2022, 29, 60117–60132. [Google Scholar] [CrossRef]
  43. Fensholt, R.; Langanke, T.; Rasmussen, K.; Reenberg, A.; Prince, S.D.; Tucker, C.; Scholes, R.J.; Le, Q.B.; Bondeau, A.; Eastman, R.; et al. Greenness in semi-arid areas across the globe 1981–2007—An Earth Observing Satellite based analysis of trends and drivers. Remote Sens. Environ. 2012, 121, 144–158. [Google Scholar] [CrossRef]
  44. Liu, Z.; Wang, H.; Li, N.; Zhu, J.; Pan, Z.; Qin, F. Spatial and temporal characteristics and driving forces of vegetation changes in the Huaihe River Basin from 2003 to 2018. Sustainability 2020, 12, 2198. [Google Scholar] [CrossRef] [Green Version]
  45. Tošić, I. Spatial and temporal variability of winter and summer precipitation over Serbia and Montenegro. Theor. Appl. Climatol. 2004, 77, 47–56. [Google Scholar] [CrossRef]
  46. Zuo, Y.; Li, Y.; He, K.; Wen, Y. Temporal and spatial variation characteristics of vegetation coverage and quantitative analysis of its potential driving forces in the Qilian Mountains, China, 2000–2020. Ecol. Indic. 2022, 143, 109429. [Google Scholar] [CrossRef]
  47. Zhu, Y.; Zhang, S.; Luo, P.; Su, F.; Sun, B.; Liang, L.; Guo, J.; Yang, R. Assessing ecohydrological factors variations and their relationships at different spatio-temporal scales in semiarid area, northwestern China. Adv. Space Res. 2021, 67, 2368–2381. [Google Scholar] [CrossRef]
  48. Li, M.; Yin, L.; Zhang, Y.; Su, X.; Liu, G.; Wang, X.; Au, Y.; Wu, X. Spatio-temporal dynamics of fractional vegetation coverage based on MODIS-EVI and its driving factors in Southwest China. Acta Ecol. Sin. 2021, 41, 1138–1147. [Google Scholar]
  49. Xu, B.; Li, J.; Luo, Z.; Wu, J.; Liu, Y.; Yang, H.; Pei, X. Analyzing the Spatiotemporal Vegetation Dynamics and Their Responses to Climate Change along the Ya’an–Linzhi Section of the Sichuan–Tibet Railway. Remote Sens. 2022, 14, 3584. [Google Scholar] [CrossRef]
  50. Wang, J.; Fan, Y.; Yang, Y.; Zhang, L.; Zhang, Y.; Li, S.; Wei, Y. Spatial-temporal evolution characteristics and driving force analysis of NDVI in the Minjiang River Basin, China, from 2001 to 2020. Water 2022, 14, 2923. [Google Scholar] [CrossRef]
  51. Li, Q.; Cao, Y.; Miao, S. Spatio-temporal variation in vegetation coverage and its response to climate factors in the Yellow River Basin, China. Acta Ecol. Sin. 2022, 42, 4041–4054. [Google Scholar]
  52. Liu, H.; Song, X.; Wen, W.; Jia, Q.; Zhu, D. Quantitative effects of climate change on vegetation dynamics in alpine grassland of Qinghai-Tibet Plateau in a county. Atmosphere 2022, 13, 324. [Google Scholar] [CrossRef]
  53. Jian, S.; Zhang, Q.; Wang, H. Spatial–temporal trends in and attribution analysis of vegetation change in the Yellow River Basin, China. Remote Sens. 2022, 14, 4607. [Google Scholar] [CrossRef]
  54. Li, Z.; Sun, X.; Huang, Z.; Zhang, X.; Wang, Z.; Li, S.; Zheng, W.; Zhai, B. Changes in nutrient balance, environmental effects, and green development after returning farmland to forests: A case study in Ningxia, China. Sci. Total Environ. 2020, 735, 139370. [Google Scholar] [CrossRef]
  55. Zheng, K.; Wei, J.; Pei, J.; Cheng, H.; Zhang, X.; Huang, F.; Li, F.; Ye, J. Impacts of climate change and human activities on grassland vegetation variation in the Chinese Loess Plateau. Sci. Total Environ. 2019, 660, 236–244. [Google Scholar] [CrossRef] [PubMed]
  56. Song, C.; Liu, Q.; Song, J.; Jiang, Z.; Lu, Z.; Chen, Y. Land use efficiency in the Yellow River Basin in the background of China’s economic transformation: Spatial-temporal characteristics and influencing factors. Land 2022, 11, 2306. [Google Scholar] [CrossRef]
  57. Jiang, W.; Niu, Z.; Wang, L.; Yao, R.; Gui, X.; Xiang, F.; Ji, Y. Impacts of drought and climatic factors on vegetation dynamics in the Yellow River Basin and Yangtze River Basin, China. Remote Sens. 2022, 14, 930. [Google Scholar] [CrossRef]
  58. Chen, K.; Han, Y.; Cao, S.; Ma, J.; Cao, G.; Lu, H. The study of vegetation carbon storage in Qinghai Lake Valley based on remote sensing and CASA Model. Procedia Environ. Sci. 2011, 10, 1568–1574. [Google Scholar]
  59. Zhao, A.; Zhang, A.; Lu, C.; Wang, D.; Wang, H.; Liu, H. Spatiotemporal variation of vegetation coverage before and after implementation of Grain for Green Program in Loess Plateau, China. Ecol. Eng. 2017, 104, 13–22. [Google Scholar] [CrossRef]
  60. Li, S.; Liang, W.; Fu, B.; Lü, Y.; Fu, S.; Wang, S.; Su, H. Vegetation changes in recent large-scale ecological restoration projects and subsequent impact on water resources in China’s Loess Plateau. Sci. Total Environ. 2016, 569–570, 1032–1039. [Google Scholar] [CrossRef]
  61. Liu, C.; Liu, B.; Zhao, W.; Zhu, Z. Temporal and spatial variability of water use efficiency of vegetation and its response to precipitation and temperature in Heihe River Basin. Acta Ecol. Sin. 2020, 40, 1–12. [Google Scholar]
  62. Wu, M.; He, H.; Zong, S.; Tan, X.; Du, H.; Zhao, D.; Liu, K.; Liang, Y. Topographic Controls on Vegetation Changes in Alpine Tundra of the Changbai Mountains. Forests 2018, 9, 756. [Google Scholar] [CrossRef] [Green Version]
  63. Deng, X.; Wu, L.; He, C.; Shao, H. Study on Spatiotemporal Variation Pattern of Vegetation Coverage on Qinghai–Tibet Plateau and the Analysis of Its Climate Driving Factors. Int. J. Env. Res. Public Health 2022, 19, 8836. [Google Scholar] [CrossRef]
  64. 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] [Green Version]
  65. Florinsky, I.V.; Kuryakova, G.A. Influence of topography on some vegetation cover properties. Catena 1996, 27, 123–141. [Google Scholar] [CrossRef]
  66. Qiu, Y.; Fu, B.; Wang, J.; Chen, L. Spatial variability of soil moisture content and its relation to environmental indices in a semi-arid gully catchment of the Loess Plateau, China. J. Arid Environ. 2001, 49, 723–750. [Google Scholar] [CrossRef] [Green Version]
Forests 14 00620 g001 550
Figure 1. The study area in the YRB: (a) location in China, (b) elevation, and (c) land-cover class.
Figure 1. The study area in the YRB: (a) location in China, (b) elevation, and (c) land-cover class.
Forests 14 00620 g001
Forests 14 00620 g002 550
Figure 2. Flow chart of the research.
Figure 2. Flow chart of the research.
Forests 14 00620 g002
Forests 14 00620 g003 550
Figure 3. Spatial distribution of the vegetation coverage: (a) spatial distribution of median kNDVI from 2001 to 2020 in the YRB and (b) the proportions of each classification.
Figure 3. Spatial distribution of the vegetation coverage: (a) spatial distribution of median kNDVI from 2001 to 2020 in the YRB and (b) the proportions of each classification.
Forests 14 00620 g003
Forests 14 00620 g004 550
Figure 4. Temporal variation of median kNDVI in the YRB from 2001 to 2020. The left axis represents the kNDVI value of the box plot, and the right axis represents the median kNDVI value.
Figure 4. Temporal variation of median kNDVI in the YRB from 2001 to 2020. The left axis represents the kNDVI value of the box plot, and the right axis represents the median kNDVI value.
Forests 14 00620 g004
Forests 14 00620 g005 550
Figure 5. Trends of inter-annual kNDVI change in the YRB from 2001 to 2020.
Figure 5. Trends of inter-annual kNDVI change in the YRB from 2001 to 2020.
Forests 14 00620 g005
Forests 14 00620 g006 550
Figure 6. Spatial distribution of Hurst exponent and future trends in vegetation coverage: (a) spatial distribution of Hurst exponent and (b) spatial distribution of the kNDVI future development trends based on kNDVI trends and sustainability.
Figure 6. Spatial distribution of Hurst exponent and future trends in vegetation coverage: (a) spatial distribution of Hurst exponent and (b) spatial distribution of the kNDVI future development trends based on kNDVI trends and sustainability.
Forests 14 00620 g006
Forests 14 00620 g007 550
Figure 7. Trends and spatial distribution of temperature and precipitation in the YRB from 2001 to 2020: (a) trends of average temperature, (b) trends of total precipitation, (c) spatial distribution of temperature, and (d) spatial distribution of precipitation.
Figure 7. Trends and spatial distribution of temperature and precipitation in the YRB from 2001 to 2020: (a) trends of average temperature, (b) trends of total precipitation, (c) spatial distribution of temperature, and (d) spatial distribution of precipitation.
Forests 14 00620 g007
Forests 14 00620 g008 550
Figure 8. Spatial distribution of partial correlation coefficient and significance level between kNDVI and climatic factors: (a) partial correlation coefficient between kNDVI and temperature, (b) significance degree of partial correlation coefficient between kNDVI and temperature, (c) partial correlation coefficient between kNDVI and precipitation, and (d) significance degree of partial correlation coefficient between kNDVI and precipitation. RP·kNDVI-T refers to the partial correlation coefficient between kNDVI and temperature; RT·kNDVI-P refers to the partial correlation coefficient between kNDVI and precipitation.
Figure 8. Spatial distribution of partial correlation coefficient and significance level between kNDVI and climatic factors: (a) partial correlation coefficient between kNDVI and temperature, (b) significance degree of partial correlation coefficient between kNDVI and temperature, (c) partial correlation coefficient between kNDVI and precipitation, and (d) significance degree of partial correlation coefficient between kNDVI and precipitation. RP·kNDVI-T refers to the partial correlation coefficient between kNDVI and temperature; RT·kNDVI-P refers to the partial correlation coefficient between kNDVI and precipitation.
Forests 14 00620 g008
Forests 14 00620 g009 550
Figure 9. Analysis of driving factors of kNDVI: (a) spatial pattern of complex correlation coefficient between kNDVI and climate factors, (b) significance level of the complex correlation coefficient, and (c) driving factors of the kNDVI in the YRB from 2001 to 2020.
Figure 9. Analysis of driving factors of kNDVI: (a) spatial pattern of complex correlation coefficient between kNDVI and climate factors, (b) significance level of the complex correlation coefficient, and (c) driving factors of the kNDVI in the YRB from 2001 to 2020.
Forests 14 00620 g009
Table
Table 1. Basis for the classification of climatic driving factors of dynamic change in the kNDVI.
Table 1. Basis for the classification of climatic driving factors of dynamic change in the kNDVI.
Type of Driving FactorClassification Basis
RT·kNDVI-PRP·kNDVI-TRkNDVI-T-P
Driven by precipitation|t| > t0.05 F > F0.05
Driven by temperature |t| > t0.05F > F0.05
Driven by both temperature and precipitation|t| < t0.05|t| < t0.05F > F0.05
Driven by non-climate factors F < F0.05
Note: RT·kNDVI-P denotes the partial correlation coefficient value between kNDVI and precipitation, RP·kNDVI-T denotes the partial correlation coefficient value between kNDVI and temperature, RkNDVI-T-P denotes the complex correlation coefficient value between kNDVI and both temperature and precipitation, t0.05 indicates that the correlation, as determined by t-test, was significant at the 0.05 level, F0.05 indicates that the correlation, as determined by F-test, was significant at the 0.05 level.
Table
Table 2. Statistical analysis results of kNDVI trends.
Table 2. Statistical analysis results of kNDVI trends.
SkNDVIZS ValuekNDVI TrendsArea Percentage/%
≥0.0005≥1.96Significantly improved65.67
≥0.0005−1.96–1.96Slightly improved10.22
−0.0005–0.0005−1.96–1.96Stable22.05
≤−0.0005−1.96–1.96Slightly degraded1.31
≤−0.0005≤−1.96Severely degraded0.75
Note: The pixels of SKNDVI between −0.0005 and 0.0005 and ZS ≤ −1.96 or ZS ≥ 1.96 were categorized as areas of stable vegetation.
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

Feng, X.; Tian, J.; Wang, Y.; Wu, J.; Liu, J.; Ya, Q.; Li, Z. Spatio-Temporal Variation and Climatic Driving Factors of Vegetation Coverage in the Yellow River Basin from 2001 to 2020 Based on kNDVI. Forests 2023, 14, 620. https://doi.org/10.3390/f14030620

AMA Style

Feng X, Tian J, Wang Y, Wu J, Liu J, Ya Q, Li Z. Spatio-Temporal Variation and Climatic Driving Factors of Vegetation Coverage in the Yellow River Basin from 2001 to 2020 Based on kNDVI. Forests. 2023; 14(3):620. https://doi.org/10.3390/f14030620

Chicago/Turabian Style

Feng, Xuejuan, Jia Tian, Yingxuan Wang, Jingjing Wu, Jie Liu, Qian Ya, and Zishuo Li. 2023. "Spatio-Temporal Variation and Climatic Driving Factors of Vegetation Coverage in the Yellow River Basin from 2001 to 2020 Based on kNDVI" Forests 14, no. 3: 620. https://doi.org/10.3390/f14030620

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