基于Google Earth Engine云计算的城市群生态质量长时序动态监测——以粤港澳大湾区为例

人类活动对生态环境的影响日益强烈，及时动态地监测生态现状及其变化信息对城市生态的管理和保护以及可持续发展具有重大意义。遥感生态指数（RSEI）是一种客观、快速和简便的生态质量监测和评价技术，已被广泛应用于生态学研究领域，但在进行大范围长时间监测时往往面临云遮挡和拼接困难的问题。因此，本文基于Google Earth Engine （GEE）平台，对1988-2018年来粤港澳大湾区共3530景Landsat遥感影像进行批量去云，采用中值合成法逐年计算绿度、湿度、干度和热度等遥感指标并利用主成分分析法构建遥感生态指数，评价了该区域近30年生态质量的时空变化。该方法改善了遥感生态指数在大范围长时序监测中数据缺失和拼接困难等问题，增加了时间序列的可比性。研究表明：（1）遥感生态指数能够较好地表征粤港澳大湾区的生态质量，其中绿度和湿度指标与其呈正相关，干度和热度指标与其呈负相关；（2）时间上，三十年间粤港澳大湾区生态质量呈"上升-下降-上升-下降"的波动下降趋势，空间上，生态质量具有明显的空间异质性，主要呈现西北和东北部高和中部低的状态。重度和中度退化区主要集中在区域中部，总体改善区域主要位于西部和东北部，基本不变区域主要包括北部区域以及香港，轻度退化区分布较为分散；（3）基于GEE云计算的图像处理可以较好的改善遥感数据缺失、色差和时间不一致等问题，极大的提高影像处理的效率，扩展了遥感生态指数在大范围长时间序列生态监测中的应用。研究结果可以为提升遥感生态指数适用范围和准确度提供参考，并为快速城市化背景下生态保护和土地管理提供理论依据。

Dynamic monitoring of long time series of ecological quality in urban agglomerations using Google Earth Engine cloud computing: A case study of the Guangdong-Hong Kong-Macao Greater Bay Area, China

The impact of human activities on the ecological environment is becoming more intense, and timely dynamic monitoring of the ecological status and change information is of great significance to urban ecological management, protection and sustainable development. The remote sensing-based ecological index (i.e. RSEI) is an objective, fast and simple ecological quality monitoring and evaluation method and has been widely applied in the field of ecological studies. But it often faces the problems of cloud occlusion and difficulty in mosaic when conducting large-scale and long-term monitoring. Thus, in this study, we collected 3530 Landsat images from 1988 to 2018 over the Guangdong-Hong Kong-Macao Greater Bay Area (GBA) and removed their clouds on Google Earth Engine(GEE) platform. The median value composite was adopted to calculate remote sensing indicators such as greenness, wetness, dryness and heat year by year and the RSEI was constructed using principal component analysis to evaluate the temporal and spatial changes of ecological quality in the region over the past 30 years. The method ameliorates RSEI in terms of the problems of data missing and image mosaic for large-scale and long-term monitoring, and increases the comparability of time series. The results show that: (1) the RSEI can better characterize the ecological quality of the GBA, in which greenness and wetness indicators have a positive correlation with RSEI, while dryness and heat indicators are negatively correlated with it. (2) In perspective of time, the ecological quality of GBA has shown a fluctuating downward trend of "up-down-up-down" in the past 30 years. In perspective of space, the ecological quality has presented obviously spatial heterogeneity, mainly showing the state of high in the northwest and northeast and low in the middle. The severe and moderately degraded areas are mainly concentrated in the central area, and the overall improvement areas are mainly located in the west and northeast. The unchanging areas mainly include the northern area and the Hong Kong, and the lightly degraded areas are scattered. (3) When using the RSEI, image processing of median composite based on the GEE cloud computing can better improve the problems of remote sensing image data missing, chromatic aberration and time inconsistency, greatly improve the efficiency of image processing, and extend the remote sensing ecological index in a large-scale and long-term sequence of ecology monitoring application. The research results can provide a reference for improving the scope and accuracy of the RSEI, and provide a theoretical basis for ecological protection and land management in the context of rapid urbanization.