渤海湾叶绿素a浓度的遥感反演模型及其应用研究

发布时间：2018-05-20 17:09

本文选题：渤海湾 + 叶绿素a　；参考：《中国地质大学(北京)》2015年硕士论文

【摘要】：过去几十年中,在我国经济持续快速发展的同时,受气候和人类活动的双重影响,我国沿海海域的环境和空间结构发生了很大的变化,海洋环境的服务和生态平衡功能受到极大影响。其中,渤海湾由于其半封闭的地理位置,海水循环自净能力低于我国其他海域,因此监测渤海湾水质状况对合理评估其生态环境、资源开发和利用的程度具有重要意义。测定渤海湾叶绿素a浓度,对于评价其水质健康状况、水产资源分布及污染程度等具有指示作用。应用遥感技术可以弥补传统调查手段的不足,实现大范围、长时间连续的水质参数监测。尽管现在已有成形的方法可以作用于海洋叶绿素a浓度的反演,可是这些模型可能不适用于渤海湾的浑浊水体,因此,需要建立针对渤海湾海域的监测模型。本文基于渤海湾水体的光谱特征研究,利用实测叶绿素a浓度数据和准同步的多光谱ETM影像,构建了具有渤海湾区域特色的叶绿素a浓度的遥感监测反演模型。依据定量遥感的反演需求,针对ETM影像进行了一系列预处理工作,完成了精确提取遥感影像信息的前期准备。通过分析ETM影像各波段及波段组合与水体叶绿素a浓度的相关性,选取了敏感性较强的波段组合作为变量,运用逐步回归分析方法,探寻这些变量与叶绿素a浓度之间的定量关系,建立了R2为0.864、RMSE为0.957的统计模型。为了进一步提高反演精度,运用人工神经网络技术,通过对比不同结构的网络的训练结果,确定了一个三层结构的BP神经网络模型,其中ETM影像1-4波段反射率作为输入节点,叶绿素a浓度实测值作为输出节点,隐含层包含8个节点,该模型的R2为0.956、RMSE为0.856。结果表明,BP神经网络模型具有更好的拟合效果,为利用遥感技术快速、准确地监测水域叶绿素a浓度提供了可靠的基础和方法。应用构建的统计模型反演获得了2007年至2010年的叶绿素a浓度分布图,用以研究渤海湾叶绿素a浓度的空间分布特征和时间变化规律。通过分析发现,渤海湾叶绿素a浓度分布大致呈现南部稍高于北部,近岸高于远岸,且随离岸距离增加而减小的空间特征。其叶绿素a浓度年季差异不明显,由于降水、水温等因素影响,春夏季总体高于秋季。
[Abstract]:In the past few decades, with the sustained and rapid economic development of our country, the environment and spatial structure of our coastal waters have undergone great changes due to the dual influence of climate and human activities. The service and ecological balance functions of the marine environment are greatly affected. Because of its semi-closed geographical location, the self-purification capacity of seawater circulation in Bohai Bay is lower than that in other sea areas in China. Therefore, monitoring the water quality of Bohai Bay is of great significance to evaluate its ecological environment, exploitation and utilization of resources. The determination of chlorophyll a concentration in Bohai Bay can be used to evaluate the health of water quality, distribution of aquatic resources and the degree of pollution. The application of remote sensing technology can make up for the deficiency of traditional investigation methods and realize the monitoring of water quality parameters in a wide range and for a long time. Although the existing methods can be used to invert the concentration of chlorophyll a in the ocean, these models may not be applicable to the muddy waters of the Bohai Bay. Therefore, it is necessary to establish a monitoring model for the Bohai Bay. Based on the spectral characteristics of the Bohai Bay water body and the measured chlorophyll a concentration data and quasi-synchronous multispectral ETM image, a remote sensing monitoring inversion model of chlorophyll a concentration in the Bohai Bay region is established in this paper. According to the demand of quantitative remote sensing inversion, a series of pre-processing work was carried out for ETM images, and the preparation for accurate extraction of remote sensing image information was completed. By analyzing the correlation between each band and band combination of ETM image and the concentration of chlorophyll a in water body, the sensitive band combination was selected as the variable, and the stepwise regression analysis method was used. To explore the quantitative relationship between these variables and the concentration of chlorophyll a, a statistical model with R2 of 0.864 and RMSE of 0.957 was established. In order to further improve the inversion accuracy, a three-layer BP neural network model is established by comparing the training results of different structures using artificial neural network technology, in which the 1-4 band reflectivity of ETM image is used as the input node. The measured value of chlorophyll a concentration is used as the output node, and the hidden layer contains 8 nodes, and the R2 of the model is 0.956% RMSE (0.856 6). The results show that the BP neural network model has better fitting effect and provides a reliable basis and method for rapid and accurate monitoring of chlorophyll a concentration in water using remote sensing technology. The statistical model was used to invert the chlorophyll a concentration distribution from 2007 to 2010, which was used to study the spatial distribution characteristics and temporal variation of chlorophyll a concentration in Bohai Bay. It is found that the distribution of chlorophyll a concentration in the Bohai Bay is slightly higher in the south than in the north, and higher in the nearshore than in the far shore, and decreases with the increase of offshore distance. The annual and seasonal differences of chlorophyll a concentration were not obvious. Due to the influence of precipitation and water temperature, the concentration of chlorophyll a in spring and summer was generally higher than that in autumn.
【学位授予单位】：中国地质大学(北京)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：X87;X834

【参考文献】