利用AI Earth平台和Sentinel-2 MSI数据进行江苏省重点湖库蓝藻水华与水生植物时空监测

作者	单位
姜明岑	中国环境监测总站, 国家环境保护环境监测质量控制重点实验室, 北京 100012
李文攀	中国环境监测总站, 国家环境保护环境监测质量控制重点实验室, 北京 100012
解鑫	中国环境监测总站, 国家环境保护环境监测质量控制重点实验室, 北京 100012
张虎军	江苏省无锡环境监测中心, 江苏无锡 214121
宋挺*	江苏省无锡环境监测中心, 江苏无锡 214121

通讯作者:宋挺* 江苏省无锡环境监测中心, 江苏无锡 214121

摘要点击次数: 390

全文下载次数: 782

中文摘要:

蓝藻水华已成为湖泊面临的严重环境挑战之一,对全球公共卫生及水生态系统构成重大威胁。水生植物在淡水生态系统中扮演着基础营养角色,是生物多样性的重要组成部分,对维持生态平衡起着关键作用。结合阿里巴巴达摩院AI Earth平台和欧洲航天局Sentinel-2 MSI数据,基于物候学原理开发了一套适用于中小型湖库的蓝藻水华和水生植物自动识别业务系统,并对江苏省12个存在蓝藻水华风险的湖库的蓝藻水华和水生植物时空分布变化进行了分析。结果表明,长江以南湖库的蓝藻水华强度普遍高于长江以北。除元荡以外,长江以南大部分湖库近年来的蓝藻水华强度整体呈现下降趋势,尤其是在2023年。相比之下,长江以北湖库的蓝藻水华强度虽然总体较低,但高邮湖和骆马湖在2023年却有明显增强。水生植物方面,除了云龙湖水库、澄湖、淀山湖和元荡等水生植物原本较少的湖库,以及石梁河水库和实施了生态湿地修复的长荡湖外,其余6个水生植物丰富的湖库均表现出水生植被面积下降的趋势,并在2023年降至近年来的最低点。研究结果可为湖库管理和生态恢复提供重要的科学依据,有助于推动水体健康监控工作的长期开展和持续改进。

英文摘要:

Cyanobacterial blooms have emerged as a major environmental challenge for lakes,posing significant risks to global public health and aquatic ecosystems.Aquatic plants play a fundamental trophic role in freshwater ecosystems,serving as vital components of biodiversity and maintaining ecological stability.To address this issue,this sutdy developed an automated monitoring system based on phenological principles for identifying cyanobacterial blooms and aquatic plants in small and medium-sized lakes and reservoirs by integrating the AI Earth platform from Alibaba DAMO Academy with Sentinel-2 MSI data from the European Space Agency.This system was applied to assess the spatiotemporal dynamics of these phenomena in 12 lakes and reservoirs in Jiangsu Province identified as being at risk of cyanobacterial blooms.The results showed that bloom intensity was generally higher in lakes south of the Yangtze River than those north of it.Except for Yuandang,most southern lakes exhibited an overall declining trend in bloom intensity,particularly in 2023.In contrast,although bloom intensity in northern lakes was generally lower,Gaoyou Lake and Luoma Lake showed a notable increase in 2023.Regarding aquatic plants,six lakes with historically high vegetation cover experienced a consistent reduction in vegetated area,reaching their lowest levels in recent years by 2023.This trend excluded lakes with historically sparse aquatic plants (Yunlong Lake Reservoir,Cheng Lake,Dianshan Lake,and Yuandang),as well as Shilianghe Reservoir,and Changdang Lake,which has undergone ecological wetland restoration.These findings provide robust scientific evidence for lake management and ecological restoration strategies,facilitating the long-term advancement and continuous refinement of freshwater ecosystem health monitoring initiatives.

查看全文查看/发表评论下载PDF阅读器