【作者】 李颖；

【导师】 蒋建军；

【作者基本信息】 南京师范大学 ， 地图学与地理信息系统， 2011， 硕士

【摘要】 内陆水体的总悬浮物浓度是其重要的水质参数,也是水质评价重要指标之一。与传统监测方法相比,基于遥感水质参数估算的方法具有效益高、面积大、时效好、可取得连续数据等优点。目前基于卫星数据的总悬浮物浓度估算模型已有大量研究,由于水质参数随着自然环境的变化,随时都在改变,野外实验实际采样往往需要一段时间,卫星数据获取是某一瞬间,因此用实测水质数据与卫星数据建立估算模型时,难以实现卫星数据与实测数据同步。本研究尝试针对这一问题进行改进。论文以太湖为研究区,采用太湖的实测光谱数据与水体总悬浮物浓度数据,通过ASD实测光谱数据模拟环境1号卫星通道,构建太湖水体总悬浮物浓度估算模型；通过地面定标方法,确定实测光谱估算参数与环境1号卫星所获得的参数之间的修正关系；最终实现基于环境1号卫星多光谱数据的太湖总悬浮物浓度估算模型构建。环境1号卫星数据采用B星CCD2传感器数据。通过研究得出以下结论：(1)确立了太湖水体总悬浮物浓度估算适宜波段太湖水体呈现典型内陆二类水体的光学特性。通过水体各组分吸收与散射特性的分析,同时结合样点实测光谱分析,确定太湖总悬浮物浓度估算敏感波段在600nm～700nm。(2)构建了基于ASD实测光谱模拟波段的总悬浮物浓度估算模型通过ASD实测光谱模拟环境1号卫星4个通道(HJ1、HJ2、HJ3、HJ4),得到ASD光谱模拟波段(ASD1、ASD2、ASD3和ASD4),分析了波段和波段组合与总悬浮物浓度相关性,相关性较好的分别有ASD3、ASD4、ASD4*ASD4/ASD1、ASD3*ASD4/(ASD 1+ASD2)。分别以这些参数为自变量建立了指数、线性、对数、多项式、乘幂等形式的模型,通过对比分析,最终选择最优化的模型为SS=2004*ASD3-34.95, SS是总悬浮物浓度。(3)构建了基于环境1号卫星的太湖总悬浮物浓度估算模型选择地表相对一致的地块作为定标区,通过定标方法确立了ASD模拟波段与环境1号卫星参数关系为ASD3=0.284*HJ3-0.014。从而构建了基于环境1号卫星的总悬浮物浓度估算模型为：SS=569.5*HJ3-63.9。模型计算的平均相对误差在0.25左右,与直接用卫星数据和地面实测水质数据进行建模(传统方法)的计算结果相比,精度有所提高。(4)太湖总悬浮物浓度时空分异特征首先在时间尺度上秋冬季节总悬浮物浓度含量较高,从高到低依次：冬>秋>春、夏,尤其在1、12月份总悬浮物浓度明显偏高。在空间分布上西南区域高多于东北区域,中心区域相对较高,在新塘港、小梅口、新港口、梅梁湖等区域也相对较高,东太湖偏低。总体来讲,太湖总悬浮物浓度分布较之其他水质参数规律性不太明显。更多还原

【Abstract】 Total suspended solids(TSS) concentration is one of the most important indexes for inland water quality assessment and water environment evaluation. Compare to conventional measurements, remote sensing in estimating water indexes has excellencies of high efficiency, large area, effects, and it can get the continuous data. Currently there are many models of estimating the total suspended solids based on satellite data.However, due to water quality parameters with the natural environment, changing all the time. Actual sampling always need a long time, but Satellite data acquisition is a moment. During building retrieval model, the satellite data and water quality data are non-synchronization. This paper attempts to improve this problem.Taihu lake was selected as study area. First, we use the measured spectral synchronization data (ASD spectrometer data) to simulate HJ-1 multiply Spectral channels. Next, Build retrieval model of total suspended solids concentration based on measured data. Then through the scaler of measured spectral data (ASD spectrometer data) and HJ-1 satellite, we will get the relations of the modified parameters. At last, This paper is to build total suspended solids concentration retrieval model of Taihu lake which is based on HJ-1 satellite. Satellite data comes from HJ-1BCCD2 satellite.Through the research we get the following conclusions:(1) Water optical properties of Taihu lake and the suitable band for retrieval of total suspended solids concentration.Taihu lake has typical inland II water’s optical properties. By water absorption and scattering properties of each component analysis and combined with the samples’ measured spectrum, we get the total suspended solids concentration’s appropriate estimated band is in the 600 nm-700nm.(2) Build the total suspended solids concentration estimation model based on ASD spectrometer data.ASD measured by spectral simulation HJ-1 satellite, four channels (HJ1, HJ2, HJ3, HJ4), by spectral simulation ASD band (ASD1, ASD2, ASD3 and ASD4).By analyzing the correlation between total suspended solids concentration and band combinations we get ASD3, ASD4, ASD3* ASD4/ASD1, ASD3* ASD4/(ASD1 +ASD2). Using these four factors as independent variables to build models of linear, logarithmic, polynomial, power and other forms. Through a series of comparative analysis, the final choice is zhe linear model:SS=2004* ASD3-34.95.(3) Correction between ASD spectrometer and HJ-1B CCD2.Total suspended solids concentration estimation mode based on HJ-1 multi-spectral data. By scaling we get the total suspended solids concentration’s parameter relationship between ASD spectrometer and HJ-1B CCD2:ASD3= 0.284* HJ3-0.014. Finally the total suspended solids concentration’s estimation model which based on HJ-1B CCD2:SS= 569.5* HJ3-63.9.Except a few abnormal samples, the model’s average relative error is 0.25 and compared to traditional methods, this model improved in accuracy.(4) Temporal and Spatial Variations of total suspended solids concentration in Taihu lake.Firstly in the time scale, total suspended solids concentration of Taihu lake has higher levels at autumn and winter.Order:winter> autumn> spring, summer. especially in January and December, total suspended solids concentration is highest. At spatial distribution, mostly, southwest region higher than the northeast region, relatively central region is higher. In Xintang Hong Kong, Mei mouth, new port Meiliang Lake area will be relatively high. East Taihu lake has low value. In general, total suspended solids concentration’laws in Taihu lake is not obvious compare to the other water quality parameters.更多还原