MODIS辅助深圳湾HJ CCD影像悬浮泥沙浓度反演研究

发布时间：2018-04-29 14:44

本文选题：深圳湾 + MODIS　；参考：《华中师范大学》2015年硕士论文

【摘要】：近岸河口海湾水体较浑浊,属于Ⅱ类水体。悬浮泥沙浓度是水体的最重要的三大水质参数之一,对海岸带冲淤变化进程以及对水体生态环境变化都具有重要意义。传统的船舶采样方法既耗时又费力,对于动态变化极大的河口海湾监测十分不利。遥感以其大面积高效率长时序获取数据的优势,在水环境研究领域应用越来越广泛。本文以经过验证的MODIS悬浮泥沙浓度(Suspended Sediment Concentration, SSC)反演结果为基础,建立HJ-1A/1B卫星CCD影像的反演模型,开展深圳湾的悬浮泥沙浓度定量反演研究,验证了在不精确大气校正条件下的参数反演的可行性,对于推动环境卫星的应用发展具有积极意义。本文的主要贡献如下：(1) MODIS影像SSC反演模型针对MODIS/Terra影像250m空间分辨率数据,利用SeaDAS version 6.4官方软件包进行大气校正获得遥感反射率。结合2007、2008年两年2个站位的同步实测悬浮泥沙浓度数据建立深圳湾的红光/绿光波段比值敏感因子的反演模型。其中的指数模型,R2为0.80,均方根误差为5.06,相对误差为15%,能够满足反演精度要求。(2)基于MODIS的HJ-1A/1B卫星CCD悬浮泥沙浓度RC、SSCI反演试验利用经过精确的瑞利校正HJ-1A/1B卫星CCD数据得出遥感反射率,以与HJ-1A/1B卫星CCD数据准同步的2009-2012年的指数模型提取的MODIS数据为基础,进行悬浮泥沙浓度RC(Rayleigh Correction)方法反演试验。HJ-1A/1B卫星CCD数据经过精确的瑞利散射校正得出遥感反射率,以基线原理分析CCD红光通道指数在不同浑浊程度的Ⅱ类水体中悬浮泥沙浓度所表现的基线高度变化,达到有效解决浑浊水体传统水色遥感大气校正剔除气溶胶多次散射信号困难的问题,即悬浮泥沙浓度指数(Suspended Sediment Concentration Index, SSCI)方法。利用该方法计算的HJ-1A/1B卫星CCD遥感反射率数据,以与HJ-1A/1B卫星CCD数据准同步的2009-2012年指数模型提取的MODIS数据为基础,进行悬浮泥沙浓度指数SSCI反演试验。(3) HJ-1A/1B影像RC和SSCI悬浮泥沙浓度反演以及精度验证。建立MODIS数据辅助的基于RC方法的HJ-1A/1B影像单波段、波段比值、单个CCD和4个CCD的深圳湾悬浮泥沙浓度指数反演模型,以及建立MODIS数据辅助的基于SSCI方法的单波段的单个CCD和4个CCD的深圳湾悬浮泥沙浓度指数反演模型,并进行精度验证和评价。模型检验表明,对于任意选取的CCD传感器数据,其RC模型和SSCI模型都有较高的决定系数R2和较低的均方根误差RMSE (R20.76,RMSE≤11.78),模型能够满足反演要求。其中,基于MODIS辅助的HJ-1A/1B卫星CCD传感器SSCI红光波段的指数模型最适合反演深圳湾水域的悬浮泥沙浓度,模型反演的悬浮泥沙产品表明,该方法能为MODIS退役之后的以HJ-1A/1B卫星CCD数据源的悬浮泥沙浓度延续观测具有积极意义,也可以为其他缺乏同步实测数据的传感器水环境参数反演提供借鉴。
[Abstract]:The water body of the coastal estuary is cloudy and belongs to the class II water body. The suspended sediment concentration is one of the most important three water quality parameters of the water body. It is of great significance to the process of erosion and deposition in the coastal zone and the changes in the ecological environment of the water body. The traditional method of ship sampling is not only time-consuming and laborious, but also for the monitoring of the dynamic changing estuarine Bay. Remote sensing is more and more widely used in the field of water environment research with its large area and high efficiency and long time sequence. Based on the verified MODIS suspended sediment concentration (Suspended Sediment Concentration, SSC) inversion results, the inversion model of HJ-1A/1B satellite CCD image is established, and the suspended mud in the Shenzhen Bay is carried out. The quantitative inversion of sand concentration has proved the feasibility of parameter inversion under the imprecise atmospheric correction. It is of great significance to promote the application and development of environmental satellites. The main contributions of this paper are as follows: (1) the SSC inversion model of MODIS image is aimed at the spatial resolution data of MODIS/Terra image 250m, and uses the official soft soft of SeaDAS version. An inversion model of the red / green band ratio sensitivity factor of Shenzhen Bay is established with the simultaneous measurement of suspended sediment concentration data from 2 stations in 20072008 years and two years. The index model, R2 is 0.80, the root mean square error is 5.06, and the phase pair error is 15%, which can satisfy the requirement of the inversion accuracy (2). MODIS based HJ-1A/1B satellite CCD suspended sediment concentration RC, SSCI inversion experiment using accurate Rayleigh correction HJ-1A/1B satellite CCD data to obtain remote sensing reflectance, based on the MODIS data extracted from the 2009-2012 year exponential model of the HJ-1A/1B satellite CCD data, the suspended sediment concentration RC (Rayleigh) method The inversion test.HJ-1A/1B satellite CCD data is corrected by accurate Rayleigh scattering, and the reflectance of remote sensing is obtained. The baseline principle is used to analyze the baseline height variation of the suspended sediment concentration of the CCD red channel index in the class II water body of different turbidity degree, and to effectively solve the traditional water color remote sensing atmosphere correction and eliminate the aerosol more than the aerosol. The problem of the secondary scattering signal is difficult, that is, the Suspended Sediment Concentration Index (SSCI) method. The HJ-1A/1B satellite CCD reflectance data calculated by this method is based on the MODIS data extracted from the 2009-2012 year exponential model of the HJ-1A/1B satellite CCD data, and the suspended sediment concentration index is carried out. SSCI inversion test. (3) inversion of suspended sediment concentration in HJ-1A/1B image RC and SSCI and accuracy verification. Establish RC based RC based HJ-1A/1B image single band, band ratio, single CCD and 4 CCD in Shenzhen Bay suspended sediment concentration index model, as well as single band based on SSCI method aided by building MODIS data A single CCD and 4 CCD in Shenzhen Bay suspended sediment concentration index inversion model is used to verify and evaluate the accuracy of the model. The model test shows that the RC model and SSCI model have higher determination coefficient R2 and lower mean square root error RMSE (R20.76, RMSE < 11.78) for any selected CCD sensor data, and the model can meet the inversion requirements. The exponential model of the red light band of the HJ-1A/1B satellite CCD sensor based on MODIS is most suitable for the inversion of the suspended sediment concentration in the waters of Shenzhen Bay. The suspended sediment products from the model show that this method can be of positive significance for the continuous observation of the suspended sediment concentration of the HJ-1A/1B satellite CCD data source after the decommissioning of MODIS. It can be used for reference for other water environmental parameters inversion without synchronous measured data.

【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P715.7

【相似文献】