湖泊水质参数遥感监测研究进展

Research Progress on Remote Sensing Monitoring of Lake Water Quality Parameters

湖泊是水圈的重要组成,也是重要的淡水资源之一.湖泊的水环境日益遭到污染,监测湖泊水质动态变化对生态环境保护具有重要意义.遥感技术为湖泊水质监测提供技术支持,克服了人工采样成本低和时效性差的特点,在湖泊水质监测中得到了广泛的应用.通过遥感可以监测的指标包括悬浮物、叶绿素a、有色可溶有机物、溶解氧和透明度等.虽然遥感为湖泊水质监测打开新思路,但是仍然存在问题:(1)遥感接受到的信号受到大气的影响,而大气校正算法不够成熟,需要更精确的大气校正来控制大气影响,消除大气和光照等对水体反射的影响;(2)受季节和空间制约,不同湖泊表面光学特性存在差异以及水中生物光学特性的差异性,利用有限实测数据建立的模型缺乏可移植性;(3)水生植物和外力作用(风力、鱼群等)的干扰,造成实测和模型估测之间的误差,而且数据的同步性难以保证,导致模型可靠性下降;(4)实测数据和遥感数据的时空尺度不匹配,难以捕捉动态精细的水质变化.因此,未来结合多源数据以及其他水文模型,深入研究水体中各组分的光学特性,优化反演算法,发展对实测数据依赖程度低的反演算法,构建可迁移的反演模型,打破模型地域局限,推动湖泊水体污染监测预警业务化发展.

Lakes are an important component of the hydrosphere and are important freshwater resources.The water environment of lakes has become increasingly polluted,and monitoring the dynamic changes in lake water quality is of great significance for ecological environment protection.Remote sensing can provide technical support for lake water quality monitoring,overcoming the low-cost and poor timeliness characteristics of manual sampling;thus,it has been widely used in lake water quality monitoring.Remote sensing can be used to monitor many indicators,including suspended solids,chlorophyll a,soluble organic matter,dissolved oxygen,transparency,etc.Although remote sensing provides a new method for lake water quality monitoring,there are still some problems in practical application.(1)It is difficult to accurately retrieve the component changes in different substances in lake water from the signals received by remote sensing,resulting in the limitation of remote sensing accuracy.Atmospheric correction can eliminate the images reflected by factors such as atmosphere and light.In order to improve the accuracy of water quality monitoring,higher-accuracy atmospheric correction algorithms are needed.However,at present the atmospheric correction algorithm is not mature enough and lacks the portability between sensors.Therefore,atmospheric correction is still a difficult problem of remote sensing retrieval in lake water quality.(2)Due to seasonal and spatial constraints,there are differences in surface optical properties of different lakes and biological optical properties,resulting in changes in remote sensing reflectance.There is also a lack of portability of the model established by using limited measured data.(3)The interference of aquatic plants and external forces(wind,fish,etc.)causes the error between the measured data and the model estimation,and the synchronization of the data is difficult to guarantee,which will introduce large errors to the lake water quality model,resulting in a decrease in the reliability of the model.(4)The spatio-temporal scale of measured data and remote sensing data do not match,and it is difficult to capture dynamic and fine changes in water quality.More precise observation of water quality is needed that can capture rapid changes in lakes.However,it is still challenging to obtain real-time measured data that meet the requirements.Therefore,it is necessary to further understand the spectral characteristics of water quality parameters,combining multi-source data and other hydrological models.In the future,a retrieval algorithm with low dependence will be developed,and migrated models will be constructed to break the regional limitations of the model.Additionally,remote sensing promotes the operational development and early warning for the water quality of lakes.