基于Google Earth Engine平台的三江源地区生态环境质量动态监测与分析

三江源是中国陆地生态系统最脆弱和敏感的区域之一,一旦遭到破坏则不可逆转。受三江源地区多云等不利气象条件的影响,很难获取大范围尺度上季相一致的、无云的Landsat遥感影像。本文利用Google Earth Engine平台,对1990-2015年的相同季节的3766景Landsat影像进行像元级融合并重构最小云量影像集,借助GEE的并行云端计算,快速得到了能够反映生态环境质量的遥感生态指数（RSEI）,对三江源地区的生态环境质量进行了评价与监测。三江源时空变化与差异分析表明：1990-2000年生态环境质量呈快速下降状态,RSEI平均值从0.588下降到了0.505,生态环境质量变化以轻度恶化为主;2000-2015年生态环境质量下降速度变缓,并于2015年呈现变好态势,生态环境质量变化以不变为主,且轻度恶化面积大幅减少;该地区的生态环境状况呈现出空间分异,自西向东,生态状况变差。基于GEE平台在三江源地区的实验结果表明,GEE可以作为大区域范围的生态环境质量评价与监测的计算平台。

Monitoring and Assessment of the Eco-Environment Quality in the Sanjiangyuan Region based on Google Earth Engine

The Sanjiangyuanregion is known in China forits fragile and sensitive terrestrial ecosystem. Degradation of its ecosystem is often irreversible, to which remote sensing based monitoring has much to offer. However, due to the cloudy and other adverse meteorological conditions there, obtaining cloudless Landsat imagery of the same season in the large region remains a big challenge. In this study, based on the Google Earth Engine platform (GEE), an image composite method was applied and 3766 tiles of Landsat TM historical images were employed to generate the same seasonal clear imagery with the lowest cloud possible composited at the pixel level. With the help of parallel cloud computing ability in GEE, a remote sensing ecological index (RSEI) was calculated directly and efficiently in GEE to reflect regional eco-environment quality. Specifically, we monitored and assessed the eco-environment quality of the Sanjiangyuan region in Qinghai Province during 1990-2015. Results show that: (1) During 1990-2000, the eco-environment quality dropped quickly (RSEI average value dropped quickly from 0.588 to 0.505), and his region suffered from mainly mild degradation; (2) During 2000-2015, degradation of the eco-environment quality slowed down toward stabilization and the eco-environment quality showed upgrade tide from 2015, and the areas of mild degradation significantly decreased. Moreover, the eco-environment quality in this region showed a spatial gradient of west-to-east degradation. Our findings provide comprehensive information for improving the eco-environment of the Sanjiangyuan region, and demonstrate the potential of using GEE for monitoring and assessing eco-environment quality at large scales.