摘要
依赖地下水生态系统(GDEs)在维持生态平衡与提供生态服务方面发挥着重要作用,但在环境变化下GDEs正面临日益严重的威胁.GDEs的识别评价是对其进行监测、管理和保护的基础.综述了GDEs的发展历程、分类划定和识别评价的研究方法,总结了GDEs识别评价模型框架,探讨了以GIS为典型工具的地理空间技术与Landsat、Sentinel、MODIS等提供的遥感数据在GDEs识别中的应用,提出了目前研究的局限性与未来发展方向.当前,GDEs的识别评价研究从使用水位监测、同位素解析等单一判定方法发展到水文地质数据、生态指标、气象数据等多源数据整合的综合性方法,结果表征也由传统分类算法转变为运用耦合机器学习、随机森林模型等.然而,如何准确识别GDEs空间分布、高精度监测其动态变化及完善适应性管理策略仍是亟待解决的问题.未来应推动GDEs识别评价方法体系的理论创新,在整合多源数据上实现方法创新,在探究深度学习算法的运用和人工智能融合上实现技术创新,以期提高GDEs的管理水平和保护修复效率.
Groundwater Dependent Ecosystems(GDEs)have played a vital role in maintaining ecological balance and providing ecological services,but they are facing increasingly severe threats under environmental changes.The identification and assessment of GDEs have been the basis for their monitoring,management,and protection.In this paper,the development,classification,delineation,and research methods for the identification and assessment of GDEs were reviewed.The framework of GDE identification and assessment models was summarized.The application of geospatial technologies,with GIS as a typical tool,and remote-sensing data provided by Landsat,Sentinel,MODIS,etc.,in the identification of GDEs was explored.The current research limitations and future directions were put forward.The research on the identification and assessment of GDEs has evolved from using single determination methods,such as water level monitoring and isotopic analysis,to a comprehensive approach that integrates multiple data sources,including hydrogeological data,ecological indicators,and meteorological data.The representation of results has also shifted from traditional classification algorithms to the use of machine-learning-coupled algorithms,such as random forest models.However,how to accurately identify the spatial distribution of GDEs,precisely monitor their dynamic changes,and improve adaptive management strategies remain urgent issues to be resolved.In the future,theoretical innovation in the methodological system for GDE identification and assessment should be promoted.Innovation in methods should be achieved by integrating multi-source data,and technological innovation should be realized by exploring the application of deep-learning algorithms and the integration of artificial intelligence,so as to enhance the management level and efficiency of protection and restoration of GDEs.
作者
叶志豪
黄龙浩
周欣
郑明霞
何连生
孟睿
张晓宇
周强
YE Zhi-hao;HUANG Long-hao;ZHOU Xin;ZHENG Ming-xia;HE Lian-sheng;MENG Rui;ZHANG Xiao-yu;ZHOU Qiang(Chinese Research Academy of Environmental Sciences,Beijing 100012,China;The College of River and Ocean Engineering,Chongqing Jiaotong University,Chongqing 400000,China)
出处
《中国环境科学》
北大核心
2025年第8期4491-4500,共10页
China Environmental Science
基金
科技部国家重点研发计划(2022YFC3703104)。
