沿海水域污水遥感监测方法研究

发布时间：2018-03-10 02:06

本文选题：污水监测　切入点：遥感　出处：《大连海事大学》2016年硕士论文　论文类型：学位论文

【摘要】：海洋是地球上生命的摇篮,蕴藏着巨大的能源,世界上各个国家的发展都离不开海洋。随着我国日新月异地发展,我国沿海和内河区域船舶溢油污染、压载水和船舶生活垃圾任意排放量与日俱增,陆源污染物的海洋排放量持续增加,我国沿海水域海水污染问题变得日益严峻,不仅影响着人们的日常生活和身体健康,也制约了我国航运业的发展。为了规范船舶操作、保护海洋环境,IMO制定了一些船舶防污染公约,比如《1954年油污公约》(OIL54)、《MARPOL73/78公约》、《(2001年控制船舶防污底系统国际公约》(AFS Convention 2004)、《国际安全与环保拆船公约》等。我国拥有漫长的海岸线,保护海洋环境防治海水污染必须引起我们的重视。IMO制定的防治海洋污染公约的实施将对我国的航运业产生巨大的影响。传统的海水水质监测方法效率较低,无法获得大范围海水的水质状况,寻找一种更加精确、简便的沿海海域污水监测模型,对海水水质进行监测是一项十分重要的工作。随着全球遥感理论不断革新,海洋污水遥感监测技术逐渐由定性遥感向定量遥感转变。将遥感技术应用于海洋水质监测领域,不仅可以对某一区域进行长期监测,从而对海水污染趋势进行预测,而且节约了污水监测成本。对海水水质进行监测同样符合IMO船舶防污染相关公约的精神。本文对我国沿海水域污水遥感监测方法进行了深入的理论研究,污水遥感监测分为定性监测点和定量监测。论文利用HJ1A/1B遥感卫星数据对我国香港海域海水水质进行了定量反演,同时利用NOAA系列卫星数据对2010年8月2日河北秦皇岛老龙头海域船舶排放的污水团的扩散及漂移状况进行了解译分析。在反演香港海域海水水质时,以反演叶绿素浓度指标为例,首先从我国资源卫星应用中心获取2012年12月26日香港海域的HJ1A/1B影像资料,其次从香港环保署获取2012年12月香港海域76个水质监测点的叶绿素浓度现场数据。本文利用三种方法对香港海域叶绿素浓度进行了分析：①用Pearson相关性分析法分析了100多种波段组合,并选取Pearson相关性系数最大的波段组合对叶绿素浓度进行多元线性回归分析；②将HJ1A/1B遥感数据四个波段数据作为输入,建立BP神经网络模型对香港海域叶绿素浓度进行反演；③建立RBF神经网络模型对叶绿素浓度进行反演。研究结果表明：多元线性回归模型和RBF神经网络模型反演结果相对误差较大,约为0.67；BP神经网络反演结果误差较小,相对误差约为0.38。2010年8月2日秦皇岛老龙头海域发生海水污染事故,给当地养殖户造成了巨大损失,本文利用遥感图像增强的相关方法,对NOAA系列遥感影像进行人工目视解译,最终发现海水污染源头及污水漂移扩散规律。
[Abstract]:The ocean is the cradle of life on the earth and contains huge energy resources. The development of every country in the world can not be separated from the ocean. With the rapid development of our country, the oil spill pollution from ships in the coastal and inland areas of our country, The amount of ballast water and domestic garbage discharged from ships is increasing with each passing day, and the marine discharge of land-based pollutants is increasing continuously. The problem of sea water pollution in coastal waters of our country becomes more and more serious, which not only affects the daily life and health of people, but also affects the health of people. It also restricts the development of the shipping industry in China. In order to regulate the operation of ships, IMO has formulated a number of anti-pollution conventions on ships for the protection of the marine environment. For example, the 1954 Oil pollution Convention, the MARPOL73/78 Convention, the 2001 International Convention for the Control of Anti-fouling bottom Systems on ships, the International Convention on the Safety and Environmental Protection of ships, etc. China has a long coastline. To protect the marine environment and prevent sea water pollution, we must pay attention to the implementation of the convention on the prevention and control of marine pollution made by IMO, which will have a great impact on the shipping industry of our country. The traditional methods for monitoring the water quality of sea water are less efficient. It is a very important task to find a more accurate and simple model for monitoring the sewerage water quality in coastal waters, which is unable to obtain the water quality of a wide range of seawater, and it is a very important task to monitor the water quality of the sea water. With the innovation of the theory of global remote sensing, The remote sensing monitoring technology of marine sewage is gradually changing from qualitative remote sensing to quantitative remote sensing. Applying remote sensing technology to the field of ocean water quality monitoring can not only monitor a certain area for a long time, so as to predict the trend of sea water pollution. Moreover, the cost of sewage monitoring is saved. The monitoring of sea water quality is also in line with the spirit of IMO ship pollution Prevention Convention. This paper makes a deep theoretical study on the remote sensing monitoring method of sewage in coastal waters of our country. The remote sensing monitoring of sewage is divided into qualitative monitoring points and quantitative monitoring points. The paper uses HJ1A/1B remote sensing satellite data to carry out quantitative inversion of seawater quality in Hong Kong sea area of China. At the same time, using the NOAA series satellite data, the paper interprets and analyzes the dispersion and drift of sewage pellets discharged by vessels in the Laotou Sea area of Qinhuangdao, Hebei Province in August 2nd 2010. When retrieving the seawater quality in Hong Kong waters, Taking the index of chlorophyll concentration inversion as an example, the HJ1A/1B image data of Hong Kong sea area in December 26th 2012 were obtained from the China Resource Satellite Application Center. Secondly, the field data of chlorophyll concentration in 76 water quality monitoring sites in Hong Kong waters in December 2012 were obtained from the Hong Kong Environmental Protection Agency. The chlorophyll concentration in Hong Kong waters was analyzed by using three methods: 1 and Pearson correlation analysis. More than 100 band combinations, The band combination with the largest correlation coefficient of Pearson was selected to carry out multivariate linear regression analysis on chlorophyll concentration. The four bands of HJ1A/1B remote sensing data were used as input. A BP neural network model was established to invert chlorophyll concentration in Hong Kong Sea area. (3) RBF neural network model was established to invert chlorophyll concentration. The results show that multiple linear regression model and RBF neural network model are used to invert chlorophyll concentration. The relative error of fruit is large, The inversion error of BP neural network is about 0.67 and the relative error is about 0.38. In August 2nd, 2010, the sea water pollution accident occurred in the Laotou sea area of Qinhuangdao, which caused huge losses to the local farmers. In this paper, the method of remote sensing image enhancement is used in this paper. The artificial visual interpretation of NOAA series remote sensing images was carried out, and the source of seawater pollution and the rule of sewage drift diffusion were found.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：X87

【相似文献】