Academic Journal of Science and Technology ISSN: 2771-3032 | Vol. 16, No. 1, 2025 19 Spatiotemporal Variation of Vegetation Coverage and Its Driving Factors in the Beijing‐Tianjin‐Hebei Region Based on the Google Earth Engine Platform Yujun Sun NORTH CHINA UNIVERSITY OF SCIENCE AND TECHNOLOGY, Tangshan 063210, China Abstract: Fractional Vegetation Cover (FVC) is an important indicator for evaluating the quality of regional ecological environments and sustainable development. As a core region of China's socio-economic development, the Beijing-Tianjin-Hebei area experiences vegetation dynamics that are significantly influenced by both intensive human activities and climate change. This study aims to reveal the spatiotemporal patterns of vegetation coverage in the Beijing-Tianjin-Hebei region from 2000 to 2023, and to identify the major climatic driving factors, with the goal of providing scientific support for regional ecological conservation and coordinated development policymaking. On this basis, a combination of methods including the Theil-Sen (Sen’s) slope estimator, Mann-Kendall (MK) test, Hurst exponent analysis, and partial correlation analysis was comprehensively applied. An in-depth analysis was conducted on the spatiotemporal dynamics of FVC, its future persistence, and its relationships with temperature and precipitation. The results indicate that: (1) Spatially, the FVC in the Beijing-Tianjin-Hebei region exhibits a distinct "northwest-high, southeast-low" distribution pattern. Temporally, the region experienced three phases: rapid increase (2000–2010), significant decline (2010–2015), and restorative growth (2015–2023).(2) Vegetation improvement areas (47.90%) and degradation areas (41.68%) show strong spatial heterogeneity, and up to 59.83% of the region exhibits anti-persistent characteristics in vegetation dynamics.(3) Precipitation serves as a positive driving factor for improved FVC, while the influence of temperature exhibits spatial heterogeneity. The results of this study highlight the need for future regional ecological management to shift from solely pursuing increases in vegetation coverage toward enhancing ecosystem quality and resilience, and to develop adaptive strategies to address potential ecological risks. Keywords: Fractional Vegetation Cover; Spatiotemporal Dynamics; Beijing-Tianjin-Hebei; Google Earth Engine. 1. Introduction As a core component of terrestrial ecosystems, vegetation serves as a vital link connecting the soil, atmosphere, and hydrosphere, playing a crucial role in regulating the global carbon balance, conserving water resources, preventing soil erosion, maintaining biodiversity, and mitigating climate change [1]. Fractional Vegetation Cover (FVC) is a key parameter for quantifying surface vegetation conditions and assessing macro-scale ecological quality. Under the dual pressures of global climate change and intensive human activities, vegetation cover has become an “indicator” of ecosystem response and resilience at the regional scale [2]. Conducting long-term, large-scale precise monitoring and attribution analysis of FVC is of significant theoretical and practical importance for understanding ecosystem evolution, evaluating the effectiveness of ecological projects, and formulating regional sustainable development strategies[3]. The Beijing-Tianjin-Hebei region is one of the most densely populated areas in China, with the fastest urbanization process and the most intense human- environment interactions[4].Rapid industrialization and urban expansion have long posed a series of severe ecological and environmental challenges to the region, including water scarcity, land degradation, and air pollution.To address these issues and establish an ecological security barrier for the capital [5], China has initiated and continuously advanced multiple large-scale ecological restoration projects in the region since the early 21st century, including the "Three- North" Shelter Forest Program, management of sandstorm sources in Beijing and Tianjin, greening of the Taihang Mountains, and the conversion of farmland to forest and grassland [6–9].Against the complex backdrop of simultaneous development pressures and ecological conservation, the evolution of vegetation coverage and the effectiveness of ecological projects in the Beijing-Tianjin- Hebei region have become focal points in regional ecology, geography, and environmental science. Traditional vegetation remote sensing studies are often constrained by bottlenecks in data acquisition, storage, and processing, making it difficult to efficiently perform long- term, large-scale dynamic analyses. In recent years, geospatial big data cloud platforms such as Google Earth Engine (GEE), with their vast online archives of remote sensing imagery (e.g., MODIS, Landsat) and powerful parallel computing capabilities, have provided revolutionary technical support for long-term analysis of ecological parameters [10].Currently, numerous studies have utilized the GEE cloud platform; for example, Song Menglai et al. [11] calculated annual FVC from Landsat imagery spanning 1990 to 2020 using GEE. Xu Mengchen et al. [12] used the GEE platform, integrating Landsat surface reflectance data and MODIS EVI products, to estimate the spatiotemporal dynamics of FVC in the Yangtze River Delta.In contrast, local computation of FVC is less efficient; for instance, Lu Junjing et al. [13] estimated FVC locally using only eight Landsat images from 1985 to 2020.These examples demonstrate that the GEE cloud platform enables researchers to rapidly and efficiently conduct in-depth studies of vegetation dynamics at regional and even global scales. Therefore, this study relies on the GEE cloud platform, utilizing MODIS NDVI remote sensing data from 2000 to 20 2023, and applies the pixel dichotomy model to estimate vegetation coverage in the Beijing-Tianjin-Hebei region. Subsequently, multiple methods including Sen’s slope, Mann- Kendall test, Hurst exponent, and partial correlation analysis are comprehensively applied to analyze the spatiotemporal evolution and driving forces of FVC. This research aims to: (1) characterize the spatial distribution of FVC in the Beijing- Tianjin-Hebei area during the last 24 years; (2) detect regions with significant vegetation changes and forecast the sustainability of future vegetation trends; and (3) examine how temperature and precipitation as major climatic factors drive changes in vegetation coverage. 2. Overview of the Study Area The study area is the Beijing-Tianjin-Hebei region, which administratively includes the two municipalities of Beijing and Tianjin as well as the entire territory of Hebei Province (Fig. 1).The region lies between 113°27′E and 119°50′E longitude and 36°03′N and 42°40′N latitude, covering a total area of approximately 218,000 square kilometers. As the core of China’s “Capital Economic Circle,” the ecological status of the Beijing-Tianjin-Hebei region holds significant national strategic importance. The topography of the region generally exhibits a stepped pattern sloping from the northwest to the southeast. The northwest consists of mountainous and hilly areas surrounded by the Yanshan and Taihang Mountains, characterized by high elevation and complex terrain, serving as a crucial water conservation zone and ecological barrier. The southeast is dominated by the vast North China Plain, featuring flat and open terrain, which is the main population and urban concentration area as well as the primary agricultural production zone. The Beijing-Tianjin-Hebei region experiences a typical temperate monsoon climate, marked by hot and rainy summers and cold, dry winters. Under the influence of human activities, the land use patterns in the region have become highly complex. As shown in Figure 1, the main land cover types include forests, shrubs, and grasslands in the northwestern mountainous areas; croplands and water bodies in the southeastern plains; and impervious surfaces radiating outward from the cores of Beijing and Tianjin. Figure 1. Overview of the Study Area 3. Materials and Methods 3.1. Data Sources and Processing The data sources used in this study are listed in Table 1. The NDVI dataset was obtained from the MOD13Q1 product— a 16-day composite vegetation index generated from MODIS Terra daily surface reflectance—accessed through the GEE platform. The spatial resolution of this dataset is 250 meters. To minimize disturbances caused by clouds, atmospheric conditions, and solar elevation angle, this study employed the Maximum Value Composite (MVC) approach to process the 16-day NDVI data annually, generating yearly NDVI datasets from 2000 to 2023.The precipitation and temperature datasets were sourced from the National Tibetan Plateau Scientific Data Center [14][15], and their monthly records from 2000 to 2023 were processed into annual sums and means, respectively. Land cover data were obtained from the 30-meter resolution Chinese Annual Land Cover Dataset (CLCD) developed by Huang Xin’s team at Wuhan University [16].All raster datasets were resampled to a spatial resolution of 250 meters using bilinear interpolation and standardized to the WGS-84 geographic coordinate system. 21 Table 1. Data Sources Name Data Source Spatial Resolution Temporal Resolution MOD13Q1 NDVI GEE Cloud Platform 250m 16 days CLCD GEE Cloud Platform 30m Annual Temperature National Tibetan Plateau Data Center 1000m Monthly Precipitation National Tibetan Plateau Data Center 1000m Monthly 3.2. FVC Calculation Method This study estimates fractional vegetation cover (FVC) using the pixel dichotomy model. The model assumes that the surface information of each pixel is a linear combination of pure vegetation and pure bare soil components, and that FVC is linearly related to the NDVI value of the pixel. The calculation formula is as follows: 𝐹 NDVI NDVI NDVI NDVI (1) In the formula, FVC denotes the fractional vegetation cover; NDVI represents the normalized difference vegetation index of the pixel; NDVIV and NDVIS correspond to the NDVI values of pure vegetation and pure bare soil pixels, respectively. In this study, to enhance the regional applicability and temporal stability of the model, NDVIV and NDVIS are not set as fixed empirical values. Instead, they are dynamically determined from the cumulative frequency histogram of all annual NDVI images from 2000 to 2023, with the 95th and 5th percentiles of NDVI values used as their respective approximations. To ensure the physical validity of FVC values, pixel values exceeding 1 are set to 1, and those below 0 are set to 0. 3.3. Methods for Spatiotemporal Evolution and Future Trend Analysis This study estimates FVC using the pixel dichotomy model. The model assumes that the surface characteristics of each pixel are a linear combination of pure vegetation and pure bare soil, and that the FVC is linearly related to the NDVI value of the pixel. The calculation formula is as follows: slope median FVC FVC 𝑗 𝑖 ∀𝑗 𝑖 (2) In the formula, FVC_i and FVC_j represent the FVC values for years i and j respectively (i < j); Median denotes the median function; slope indicates the trend. The Mann-Kendall significance test is characterized by not requiring the data to follow a normal distribution and being robust to missing values and outliers. For a sequence Xt=x1， x2，⋯xn, the test statistic S is defined as: 𝑆     sgn 𝑥 𝑥 (3) In the formula, n denotes the length of the time series; sgn is the sign function. The test statistic Z is used to perform the trend test, with the specific procedure as follows: 𝑍 ⎩ ⎪ ⎨ ⎪ ⎧ 𝑆 1 var 𝑆 𝑆 0 0 𝑆 0 𝑆 1 var 𝑆 𝑆 0 (4) In the formula, the calculation of var(S) is given by: var 𝑆 𝑛 𝑛 1 2𝑛 5 18 (5) In this study, the significance level is set at 𝛼=0.05, corresponding to a critical value Z1-α/2=±1.96. When the absolute value of Z exceeds 1.96, the trend passes the significance test at the 95% confidence level. The Hurst exponent is a statistic that quantitatively describes the long-term correlation or persistence of a time series [19]. This study uses the Hurst exponent to quantitatively characterize the persistence of fractional vegetation cover in the Beijing-Tianjin-Hebei region over a long temporal sequence, and combines it with the slope trend to investigate the future trajectory of vegetation cover. The Hurst exponent HHH is interpreted as follows: when 00.0001 >1.96 Significant Increase 20.38 >0.0001 -1.96~1.96 Non-significant Increase 27.52 -0.0001~0.0001 ~ Nearly Stable 10.42 <-0.0001 -1.96~1.96 Significant Decrease 19.43 <-0.0001 <-1.96 Non-significant Decrease 22.25 4.4. Future Characteristics of FVC Change Figure 5 illustrates that the future spatial pattern of FVC change in the Beijing-Tianjin-Hebei area reveals an overall lack of persistence in vegetation trends. According to the statistics in Table 5, areas exhibiting anti-persistence (H < 0.5) account for as much as 59.83%, including 32.23% classified as “increase-to-decrease” and 27.60% as “decrease-to- increase” regions. In contrast, persistent regions (H > 0.5), namely “continuous increase” and “continuous decrease,” together account for only 29.95% of the total area. The largest category, “increase-to-decrease” (32.23%), is widely distributed across the Yanshan and Taihang mountain ranges. Although these areas have achieved notable ecological restoration over the past two decades, they face a significant risk of future degradation, potentially due to the diminishing marginal effects of ecological projects, shifts in climatic conditions, or increased human disturbances. Additionally, the second-largest category, “decrease-to-increase” (27.60%), is primarily concentrated in the southeastern plains, where vegetation degradation was historically severe, particularly in peri-urban expansion zones. This trend may be attributed to the implementation of urban greening policies and industrial restructuring, which allowed certain lands to recover. The spatial distribution of these categories carries significant implications as an early warning for regional ecological management and policy-making. 25 Figure 5. Hurst Index-Based Spatial Pattern of Future Vegetation Change in the Beijing-Tianjin-Hebei Area during 2000 to 2023 Table 5. Hurst Index-Based Classification of Fractional Vegetation Cover in the Beijing-Tianjin-Hebei Area from 2000 to 2023 slope HI Future Change Trends Percentage of Area (%) >0.0001 0~0.5 Increase-to-Decrease 32.23 >0.0001 0.5~1 Continuous Increase 15.67 -0.0001~0.0001 0~0.5 Unpredictable 9.45 -0.0001~0.0001 0.5~1 Continuous Stability 0.96 <-0.0001 0~0.5 Decrease-to-Increase 27.60 <-0.0001 0.5~1 Continuous Decrease 14.28 4.5. Driving Forces of FVC Figure 6 illustrates that rainfall acts as a broadly significant positive driver throughout the study region, with the majority of areas displaying positive correlations, especially in the Yanshan and Taihang mountain ranges to the northwest and the North China Plain to the southeast. The significance test map (Fig. 6, right) shows that most of the positively correlated regions achieve significance at p < 0.05, suggesting that enhanced rainfall substantially facilitates vegetation development within the study area. This outcome aligns with the natural geographic context of the Beijing-Tianjin-Hebei region as a transitional zone between semi-humid and semi- arid climates, where water availability is a critical limiting factor for vegetation growth and ecosystem productivity; thus, abundant precipitation effectively alleviates water stress and positively influences FVC. In contrast, the effect of temperature (Fig. 7) displays a complex spatial heterogeneity. In the northwest regions of the Yanshan and Taihang Mountains, there is generally a significant negative correlation between temperature and FVC, indicating that in these higher-altitude and relatively cooler mountainous areas, increased temperature inhibits vegetation growth. The primary reason may be that rising temperatures intensify soil evaporation and plant transpiration, leading to more severe moisture deficits under limited precipitation supply; this water stress negatively outweighs any positive effects from improved thermal conditions. In contrast, temperature and fractional vegetation cover in the flat southeastern North China Plain mainly exhibit a positive relationship, which may be attributed to moderate temperature increases satisfying thermal requirements for crops and other vegetation, lengthening the growing season and thereby enhancing vegetation growth and cover. 26 Figure 6. Spatial Distribution of Partial Correlation between Precipitation and FVC Figure 7. Spatial Distribution of Partial Correlation between Temperature and FVC 5. Conclusion This study conducted a spatiotemporal evolution and driving factor analysis of FVC in the Beijing-Tianjin-Hebei region from 2000 to 2023. The main conclusions are as follows: (1) Vegetation cover in the study area exhibits a spatial pattern of "high in the northwest and low in the southeast," with high-coverage zones concentrated mainly in the Yanshan and Taihang Mountains in the northwest; (2) The temporal evolution shows three phases, including rapid growth from 2000 to 2010, a sharp decline from 2010 to 2015, and a recovery after 2015. Nearly 60% of the vegetation change trends display anti-persistence; (3) Increased precipitation acts as a favorable driver for vegetation growth in the region, while rising temperature promotes vegetation growth in the southeastern plains with relatively sufficient moisture and heat conditions but inhibits vegetation growth in the water-limited northwestern mountainous areas. 27 6. Discussion The results of this study provide strong evidence for the significant achievements of national ecological projects such as the "Three-North Shelter Forest Program" and the "Beijing-Tianjin Sandstorm Source Control." However, the study also reveals the continuous degradation of vegetation in the southeastern plains, highlighting the encroachment of ecological space amid rapid urbanization in China. Future ecological management strategies in the Beijing-Tianjin- Hebei region should be adjusted with targeted measures. In the northwestern mountainous areas, the policy focus should shift from merely "increasing greenery" to "maintaining vitality," by implementing scientific forest management and tending to enhance the health and resilience of forest ecosystems, enabling them to better adapt to future climate change. For the southeastern plains, the core policy should emphasize "preserving greenery," through delineating and strictly enforcing ecological protection redlines, integrating urban green spaces, parks, and river corridors to form ecological networks, and vigorously promoting green infrastructure such as "sponge cities" to mitigate the negative impacts of urbanization on ecosystems and achieve harmonious coexistence between humans and nature. Furthermore, although this study reveals the dynamics of vegetation changes in the Beijing-Tianjin-Hebei region, some limitations remain. The partial correlation analysis used for driving factor identification can only uncover linear relationships, whereas in reality, the influences of driving factors are often complex and nonlinear. Additionally, this research mainly focuses on climatic factors, while the impacts of human activities and their interactions with natural factors require further investigation to provide a more comprehensive and in-depth assessment of the actual state and sustainability of regional ecosystems. References [1] T. Yu, P. Yang, X. Deng, et al. Fanjingshan long-term vegetation coverage change trend and comparative analysis, Bulletin of Surveying and Mapping, (2023) No.10, p.1-6. (In Chinese) [2] X.N. Zhou, R.J. Ba, C.Z. Xiao, et al. Soil characteristics and spatiotemporal changes of vegetation in the lower section of the arid valley area of the upper Minjiang River based on geological construction, Modern Geology, Vol.38 (2024) No.3, p.660-673. (In Chinese) [3] W.Y. Li, A.J. Lan, Z.M. Fan, et al. Dynamic evaluation of ecological environment quality and its driving factors in the ecological functional zone of Central Guizhou cities, Progress in Geophysics, Vol.39 (2024) No.5, p.1749-1762. (In Chinese) [4] Z.H. Liu, Q.L. Yuan, C. Li, et al. Synergistic effect of low- carbon economy and high-quality development under the “dual-carbon” strategy: A case study of Beijing-Tianjin-Hebei urban agglomeration, Environmental Science, Vol.45 (2024) No.11, p.6301-6312. (In Chinese) [5] S.L. Yu, S.L. Wang. Strict law enforcement and supervision to create an ecological security barrier in the capital, China Environmental Monitoring, (2019) No.11, p.48-51. (In Chinese) [6] S.D. Li, D.Q. Feng. World-renowned ecological project— China’s “Three-North Shelterbelt Program”, Zhejiang Forestry, (2021) No.9, p.9-11. (In Chinese) [7] J.Y. Wang, Y.Y. Zhao, J. Geng, et al. Supply and demand matching relationship of ecosystem services in the Beijing- Tianjin sandstorm source control project area, Transactions of the Chinese Society of Agricultural Engineering, Vol.41 (2025) No.4, p.279-287. (In Chinese) [8] J. Shi. Thoughts and countermeasure suggestions on promoting the greening of Taihang Mountains in Hebei, National Afforestation, (2023) No.11, p.68-70. (In Chinese) [9] Y.Q. Ma, J.H. Li, W. Cao, et al. Estimation of carbon sequestration potential and contribution of returning farmland to forest and grassland in Beijing-Tianjin-Hebei region, Acta Geographica Sinica, Vol.79 (2024) No.3, p.732-746. (In Chinese) [10] H. Tamiminia, B. Salehi, M. Mahdianpari, et al. Google Earth Engine for geo-big data applications: A meta-analysis and systematic review, ISPRS Journal of Photogrammetry and Remote Sensing, Vol.164 (2020), p.152-170. [11] M.L. Song, H.T. Chen, H. Ding, et al. Spatiotemporal evolution characteristics and influencing factors of vegetation coverage in Tianjin from 1990 to 2020, Research of Soil and Water Conservation, Vol.30 (2023) No.1, p.154-163. (In Chinese) [12] L.C. Xu, L. Deng, J. Chen, et al. Spatiotemporal changes and characteristic analysis of vegetation coverage in the Yangtze River Delta based on GEE, Beijing Surveying and Mapping, Vol.38 (2024) No.3, p.387-392. (In Chinese) [13] J.J. Lu, L.G. Sun, L. Zuo, et al. Spatiotemporal evolution characteristics and influencing factors of vegetation coverage based on functional zoning in Beijing-Tianjin-Hebei region, Remote Sensing for Natural Resources, Vol.36 (2024) No.4, p.242-253. (In Chinese) [14] S.Z. Peng. China 1km monthly precipitation dataset (1901– 2024), National Tibetan Plateau Data Center (2020). (In Chinese) [15] S.Z. Peng. China 1km monthly average temperature dataset (1901–2023), National Tibetan Plateau Data Center (2019). (In Chinese) [16] J. Yang, X. Huang. The 30 m annual land cover dataset and its dynamics in China from 1990 to 2019, Earth System Science Data, Vol.13 (2021) No.8, p.3907-3925. [17] Y.T. Liu, L. Wang, X.H. Li, et al. Spatiotemporal variation and influencing factors of vegetation coverage in Southwest China from 2000 to 2020, Plateau Meteorology, Vol.43 (2024) No.1, p.264-276. (In Chinese) [18] D.L. Wang, A.X. Su, W.P. Liu. Comparative study on several methods of vegetation coverage change trend analysis, Anhui Agricultural Science, Vol.47 (2019) No.5, p.10-14. (In Chinese) [19] M.L. Deng, F.G. Jiang, Y. Long, et al. Dynamic change and spatiotemporal pattern analysis of vegetation NDVI in Shenmu City, Journal of Central South University of Forestry and Technology, Vol.43 (2023) No.7, p.109-119. (In Chinese)