Wang Si-Meng, Qin Bo-Qiang
School of Geography and Ocean Sciences, Nanjing University, Nanjing 210023, China.
State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing 210008, China.
Huan Jing Ke Xue. 2023 Mar 8;44(3):1228-1243. doi: 10.13227/j.hjkx.202203285.
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.
湖泊是水圈的重要组成部分，是重要的淡水资源。湖泊水环境日益受到污染，监测湖泊水质动态变化对生态环境保护具有重要意义。遥感可为湖泊水质监测提供技术支持，克服人工采样成本高、时效性差的特点；因此，它已被广泛应用于湖泊水质监测。遥感可用于监测多种指标，包括悬浮固体、叶绿素a、可溶性有机物、溶解氧、透明度等。虽然遥感为湖泊水质监测提供了一种新方法，但在实际应用中仍存在一些问题。① 从遥感接收到的信号中难以准确反演湖水中不同物质的成分变化，导致遥感精度受限。大气校正可以消除大气和光等因素反射的图像。为提高水质监测精度，需要更高精度的大气校正算法。然而，目前大气校正算法还不够成熟，缺乏传感器之间的可移植性。因此，大气校正是湖泊水质遥感反演中的一个难题。② 由于季节和空间限制，不同湖泊的表面光学特性和生物光学特性存在差异，导致遥感反射率发生变化。利用有限实测数据建立的模型也缺乏可移植性。③ 水生植物和外力（风、鱼等）的干扰导致实测数据与模型估计之间存在误差，且数据同步难以保证，这会给湖泊水质模型带来较大误差，导致模型可靠性下降。④ 实测数据与遥感数据的时空尺度不匹配，难以捕捉水质的动态精细变化。需要更精确的水质观测来捕捉湖泊中的快速变化。然而，获取满足要求的实时实测数据仍具有挑战性。因此，有必要进一步了解水质参数的光谱特征，结合多源数据和其他水文模型。未来，将开发低依赖性的反演算法，构建迁移模型以打破模型的区域限制。此外，遥感推动了湖泊水质的业务化发展和预警。
