Objective This study systematically reveals the spatiotemporal evolution characteristics of fractional vegetation cover (FVC) and its driving factors in the Guangdong-Hong Kong-Macao Greater Bay Area, thereby providing a scientific basis for regional ecological environment management and sustainable development. Methods Based on the Google Earth Engine (GEE) platform, the MODIS MOD13Q1 dataset was used, and the pixel-based binary model was employed to retrieve FVC during the growing seasons (April to September each year) from 2000 to 2023. Combining Theil-Sen median trend analysis, Mann-Kendall tests, coefficient of variation (CV), and spatial autocorrelation analysis, this study analyzed the spatiotemporal variation characteristics of FVC. Additionally, based on the Light Gradient Boosting Machine (LightGBM) model and the Shapley Additive Explanations (SHAP) method, the primary driving factors and their threshold effects were explored. Results (1) From 2000 to 2023, the multi-year average FVC during the growing season in the Guangdong-Hong Kong-Macao Greater Bay Area was 0.67, with an overall significant upward trend (annual growth rate of 0.001 1/a, p < 0.05). The spatial distribution pattern was characterized by “low in the central region and high in the peripheral areas”. The FVC variation trends were predominantly improving (55.8%), among which 22.8% of the region reached a significantly improved level. (2) Land surface temperature, population density, and land use type were the primary driving factors affecting FVC. Among them, land surface temperature had the highest average SHAP value (>0.06), and the average SHAP values of population density and land use type both exceeded 0.05, with a cumulative contribution rate of 54.23%. After land surface temperature and population density exceeded the thresholds of 18.14 ℃ and 4 119.26 people/km², respectively, their positive promoting effects on FVC gradually weakened and turned into negative inhibitory effects. Conclusion Over the past 24 years, the evolution of FVC in the Guangdong-Hong Kong-Macao Greater Bay Area has exhibited significant spatial heterogeneity and nonlinear driving characteristics. The thermal environment and human activities are the core constraints restricting regional vegetation growth.
变异系数(Coefficient of Variation, CV)又称标准离差率,能够表征变量的相对稳定性或波动程度[14]。公式如下:
式中:σ为FVC的标准差;FVC i 为第i年植被覆盖度;μ为FVC的多年均值。CV值越大表明植被波动越剧烈;反之则趋于稳定。采用自然断点法,将其分为极低波动(CV<0.05)、低波动(0.05≤CV<0.10)、中等波动(0.10≤CV<0.15)、高波动(0.15≤CV<0.20)及极高波动(CV≥0.20)5个等级。
NiuZ N, HeH L, ZhangM Y, et al. Assessment and attribution analysis of terrestrial ecosystem quality in China from 2003 to 2020[J]. Acta Geographica Sinica, 2025,80(7):1901-1919.
XieC F, ShuT, LiuY, et al. Evaluating the health of Liangzi Lake riparian wetland ecosystem based on vegetation biotic integrity[J]. Resources and Environment in the Yangtze Basin, 2015,24(8):1387-1394.
ChenY H, LiX B, ShiP J. Landscape analysis of vegetation cover change based on remote sensing: a case study of Haidian District, Beijing[J]. Acta Ecologica Sinica, 2002,22(10):1581-1586.
[7]
AndreattaD, GianelleD, ScottonM, et al. Estimating grassland vegetation cover with remote sensing: a comparison between Landsat-8, Sentinel-2 and PlanetScope imagery[J]. Ecological Indicators, 2022,141:109102.
YangS K, LiuJ, WeiR, et al. Spatiotemporal characteristics and causes of vegetation coverage during the growing season in the Upper Yangtze River Basin[J]. Resources and Environment of the Yangtze River Basin, 2022,31(7):1523-1533.
RenY H, FengY, ChenW K, et al. Characteristics and driving factors of vegetation coverage changes in Longquan Mountain Urban Forest Park, Chengdu[J]. Chinese Journal of Applied Ecology, 2025,36(7):2103-2113.
ZhaoC X, LiuX R, ZhuJ, et al. Spatiotemporal trends and driving factors of vegetation coverage in Jiangsu Province from 2013 to 2022[J]. Bulletin of Soil and Water Conservation, 2025,45(3):397-406.
WangR Q, JiangX H, NieT. Spatiotemporal variation and driving force analysis of NDVI in the Pearl River Basin based on Geographic Detector[J]. People of the Pearl River, 2022,43(7):61-73.
ChenW Y, XiaL H, XuG L, et al. Dynamic variation of NDVI and its influencing factors in the Pearl River Basin from 2000 to 2020[J]. Ecology and Environmental Sciences, 2022,31(7):1306-1316.
ZhaoJ C, ZhangS H, YinX X, et al. Analysis of vegetation cover change and its influencing factors in the Guangdong-Hong Kong-Macao Greater Bay Area[J]. Surveying and Mapping Science, 2022,47(3):75-84.
FengX H, ZengZ L. Natural driving forces of vegetation cover characteristics and change trends in the Guangdong-Hong Kong-Macao Greater Bay Area [J]. Ecology and Environmental Sciences, 2022,31(9):1713-1724.
LiuD D, PanP, FuJ, et al. Spatiotemporal characteristics and driving factors of vegetation cover in Southern Jiangxi from 2000 to 2020[J]. Chinese Journal of Applied Ecology, 2023,34(11):2919-2928.
[24]
WangQ, ZhangQ P, ZhouW. Grassland coverage changes and analysis of the driving forces in Maqu County[J]. Physics Procedia, 2012,33:1292-1297.
ZhengY J, LiuJ H, LiuY, et al. Analysis of vegetation cover changes and driving factors in Ordos City from 2000 to 2018[J]. Environmental Science Research, 2022,35(11):2458-2468.
[27]
徐建华.现代地理学中的数学方法[M].3版.北京:高等教育出版社,2017.
[28]
XuJ H. Mathematical methods in modern geography[M].3rd ed. Beijing: Higher Education Press, 2017. ()
ChenW, WangZ, ZhaoH M, et al. Research on temporal and spatial variation of fractional vegetation cover in Golden Camellia National Nature Reserve using linear fusion methods[J]. Bulletin of Surveying and Mapping, 2021(11):1-6.
YuanL H, JiangW G, ShenW M, et al. Spatiotemporal changes of vegetation cover in the Yellow River Basin from 2000 to 2010[J]. Acta Ecologica Sinica, 2013,33(24):7798-7806.
RenM, WangX M, LiuL, et al. Analysis of the spatial and temporal changes and influencing factors of environmental regulation efficiency in coastal city clusters of China[J]. Geographic Sciences, 2019,39(7):1119-1128.
[35]
LIIIPearson K.. On lines and planes of closest fit to systems of points in space[J]. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science, 1901,2(11):559-572.
[36]
DormannC F, ElithJ, BacherS, et al. Collinearity: a review of methods to deal with it and a simulation study evaluating their performance[J]. Ecography, 2013,36(1):27-46.
[37]
YuanQ Q, ShenH F, LiT W, et al. Deep learning in environmental remote sensing: achievements and challenges[J]. Remote Sensing of Environment, 2020,241:111716.
[38]
SunY L, YangY L, ZhangL, et al. The relative roles of climate variations and human activities in vegetation change in North China[J]. Physics and Chemistry Earth, Parts A/B/C,2015,87/88:67-78.
[39]
BentéjacC, CsörgőA, Martínez-MuñozG. A comparative analysis of gradient boosting algorithms[J]. Artificial Intelligence Review, 2021, 54(3): 1937-1967.
YuanH, TangD W, LiuH, et al. Dynamics evolution and driving factors interaction effects of ecosystem services in the Wuling Mountain Area[J]. Environmental Science, 2026,47(1):431-443.