大熊猫国家公园四川片区生态质量时空演变及模拟预测OACHSSCD
Spatiotemporal evolution and simulation prediction of ecological quality in the Sichuan Area of the Giant Panda National Park
大熊猫国家公园是我国长江上游区生物多样性保护的生态安全屏障,科学评估其生态质量现状、揭示其演变规律并预测未来趋势,对实现国家公园的精细化管理和可持续发展至关重要.基于 2000-2023 年的多源遥感、气象及社会经济数据,构建了生态质量评价指标,采用空间自相关和地理探测器模型,解析了生态质量的时空动态与驱动机制,对比了贝叶斯神经网络等4 种模型的反演效果,对 2030 年不同情景下的生态质量进行了模拟预测.结果表明:(1)2000-2023 年,研究区生态质量整体显著向好,优良等级面积占比从 77.49%提升至 87.78%,空间上呈现"西北低、东南高"的强聚集格局.(2)土地利用生态指数、高程和气温是主导生态质量空间分异的核心因子,各因子交互作用均表现为增强效应,尤其当人口、GDP 等社会经济因子与高程等自然因子叠加时,其影响力被显著放大.(3)阻尼最小二乘法(LM)神经网络模型因其预测结果更为稳健,基于该模型,生态保护优先情景下,2030 年优等生态区面积将进一步增加 5.34%;自然发展情景下则面临轻微退化风险,优等区面积下降 1.86%;而经济发展优先情景将导致优等区面积锐减 53.24%,引发严重生态退化.研究认为,大熊猫国家公园的生态基底由自然地理格局决定,但人类活动通过与自然本底的交互作用对生态系统产生影响.研究成果可为国家公园的精细化管理和适应性保护策略制定提供科学依据.
The Giant Panda National Park serves as an ecological security barrier for biodiversity conservation in the upper reaches of the Yangtze River in China.A scientific assessment of its current ecological status,along with an analysis of its evolutionary patterns and future trends,is crucial for achieving refined management and sustainable development of the national park.This study utilized multi-source remote sensing,meteorological,and socio-economic data from 2000 to 2023 to establish an ecological quality evaluation index.Spatial autocorrelation and geographical detector models were employed to analyze the spatiotemporal dynamics of ecological quality and the driving mechanisms behind these changes.Additionally,the performance of four models—Bayesian Neural Network,BP Neural Network,GA-BP Neural Network,and LM Neural Network—was compared to assess their ability to simulate ecological quality.The study also included simulations for 2030,forecasting ecological quality under different scenarios,providing insights into the future ecological trends and their implications.The performance of four models,including the Bayesian neural network,was compared,and ecological quality under different scenarios for 2030 was simulated.The results indicated that:(1)From 2000 to 2023,the overall ecological quality of the study area showed significant improvement,with the proportion of areas rated as good or excellent increasing from 77.49%to 87.78%.Spatially,it exhibited a strong clustering pattern characterized by"lower values in the northwest and higher values in the southeast."(2)Land use ecological index,elevation,and temperature were identified as the core factors driving the spatial differentiation of ecological quality.All factor interactions demonstrated enhancing effects,particularly when socio-economic factors such as population and GDP interacted with natural factors like elevation,which significantly amplified their influence.(3)The LM neural network model demonstrated more robust predictive performance.Based on this model,under an ecological protection priority scenario,the area of excellent ecological zones was projected to increase by 5.34%by 2030;under a natural development scenario,a slight risk of degradation appeared,with the area of excellent zones decreasing by 1.86%;while under an economic development priority scenario,the area of excellent zones showed a sharp decrease of 53.24%,indicating severe ecological degradation.The study concluded that while the ecological foundation of the Giant Panda National Park was determined by its natural geographical patterns,human activities influenced the ecosystem through interactions with these natural conditions.These findings provided a scientific basis for formulating refined management and adaptive conservation strategies for the national park.
王瑶;向明顺;邬真妮;李建华
成都理工大学地理与规划学院,成都 610059成都理工大学地理与规划学院,成都 610059||成都理工大学青藏高原及其东缘人文地理研究中心,成都 610059成都理工大学地理与规划学院,成都 610059成都理工大学地理与规划学院,成都 610059
生态质量驱动机制机器学习情景模拟大熊猫国家公园
ecological qualitydriving mechanismmachine learningscenario simulationGiant Panda National Park
《生态学报》 2026 (8)
3999-4016,18
国家自然科学基金项目(42471229)
评论