关键词: 谷歌地球引擎, 生态环境质量, 遥感生态指数, 时空演变趋势, 恐龙谷地区

Abstract: Objective　Mountain ecosystems play an important role in maintaining regional ecological security and promoting sustainable development. The Dinosaur Valley region, located in Lufeng City, Chuxiong Yi Autonomous Prefecture, Yunnan Province, is a typical mountainous area of the Central Yunnan Plateau with complex topography and diverse geomorphological characteristics. As an important part of the Yunnan Lufeng Dinosaur National Geopark, this region has significant geological heritage and scientific education value, as well as important ecological protection functions. In recent years, under the joint influence of climate change and human activities, the eco-environmental quality of the region is facing increasing pressure. However, systematic studies on the dynamic evolution of eco-environmental quality in this area remain limited. Therefore, this study utilizes remote sensing monitoring to analyze the spatiotemporal variations and future trends of eco-environmental quality in the Dinosaur Valley region, in order to provide a scientific basis for regional ecological protection, territorial spatial planning, refined ecological management, and sustainable development of this region. Methods　Based on the Google Earth Engine (GEE) cloud computing platform, this study used Landsat series remote sensing images from 2000 to 2020 to conduct long-term eco-environmental quality analysis and comprehensive evaluation of the Dinosaur Valley region. Considering that in complex mountainous areas, the traditional normalized difference vegetation index (NDVI) is easily affected by terrain shadow, slope aspect, and surface relief, the normalized difference mountain vegetation index (NDMVI) was introduced to replace the traditional NDVI in this work. Then, an improved remote sensing ecological index (RSEI) model was constructed by integrating ecological indicators including greenness, wetness, heat, and dryness, and Principal component analysis (PCA) was used to objectively determine the weights of ecological indicators and reduce the subjectivity caused by artificial weighting. In addition, Theil-Sen median trend analysis, coefficient of variation analysis, and the Hurst exponent were employed to comprehensively evaluate the temporal variation, spatial differentiation, fluctuation characteristics, and future sustainability of eco-environmental quality. It is worth noting that the future trend identified by the Hurst exponent mainly reflects the inertia characteristics based on historical variation trends. Results and Discussion　The results show that the eco-environmental quality of the Dinosaur Valley region improved overall from 2000 to 2020. The mean RSEI increased from 0.55 to 0.63, indicating an overall improvement in the regional ecological condition during the study period. Spatially, areas with high eco-environmental quality were mainly distributed in mountainous, valley, and hilly regions with relatively high vegetation coverage and weaker human interference, whereas areas with low eco-environmental quality were mostly located in local regions with more frequent human activities and relatively low vegetation coverage. This indicates that vegetation coverage, terrain conditions, and human activity intensity jointly affect the spatial pattern of eco-environmental quality in the study area. Compared with the traditional NDVI, NDMVI can more accurately reflect the actual vegetation conditions in complex mountainous terrain and reduce the interference caused by terrain shadow, slope, and aspect differences. The average contribution rate of the first principal component of the improved RSEI model reached 82.33%, indicating that the model can effectively integrate greenness, wetness, heat, and dryness information and has good applicability and explanatory ability. Meanwhile, the introduction of NDMVI significantly improved the spatial consistency of vegetation indices in complex terrain, showing better performance than the traditional NDVI. The eco-environmental quality of the study area showed relatively small interannual fluctuation, with an average coefficient of variation of 0.10836, suggesting that the overall ecological condition remained relatively stable from 2000 to 2020, although some local areas still showed certain instability. Trend analysis further showed that the proportion of improved areas reached 91.3%, while degraded areas accounted for 7.82%, indicating that ecological improvement was dominant in the region during the study period. According to the Hurst exponent, the proportions of areas experiencing continuous improvement and degradation in the future were 62.67% and 29.44%, respectively. This indicates that, under the inertia of historical trends, the future eco-environmental quality in the Dinosaur Valley region will still be dominated by improvement, but some areas may face the risk of transitioning from current improvement to degradation or continued degradation. Conclusions　The improved RSEI model based on the GEE platform and NDMVI is suitable for eco-environmental quality assessment in complex mountainous areas such as the Dinosaur Valley region. From 2000 to 2020, the eco-environmental quality of the study area generally showed an improving trend, with the improvement area far exceeding the degradation area. However, the risk of local degradation still exist, especially in areas with relatively frequent human activities and low vegetation coverage. Therefore, dynamic monitoring and refined management of key risk areas should be strengthened in the future. Human interference in ecological space should be reasonably controlled, and ecological restoration measures such as vegetation restoration and soil and water conservation should be continuously promoted. The results of this study reveal the spatiotemporal evolution characteristics and future trends of eco-environmental quality in the Dinosaur Valley region, and can provide scientific support for regional ecological protection, ecological restoration, territorial spatial planning, and sustainable development in the region.

Key words: Google Earth Engine, eco-environmental quality, remote sensing ecological index, spatiotemporal evolution trend, Dinosaur Valley region