基于高光谱分析的长荡湖水质监测技术研究

Research on water quality monitoring technology of Changdang Lake based on hyperspectral analysis

以江苏省长荡湖为例,对15项常规水质参数进行实地采样试验,并同步测定采样点的水体高光谱信息。为研究高光谱信息在该湖泊水质反演中的适用性,对水质参数和对应光谱特征进行定量分析,筛选高光谱对水质指标的敏感波段。分别选用波段差值、波段比值和单波段建立半经验数学模型,并根据实测值验证各模型反演精度。结果表明,利用波段组合反射率建立的模型对TP、浊度、悬浮物SS、水温和透明度具有较好的拟合效果。

Taking Changdang Lake as an example,field sampling experiments were conducted on 15 sets of normal water quality parameters and the hyperspectral information of the water body at the sampling points was simultaneously measured.In order to study the applicability of hyperspectral information in water quality inversion in the lake,the water quality parameters and corresponding spectral characteristics were quantitatively analyzed,and the sensitive bands of hyperspectral to water quality indicators were selected.The semi-empirical mathematical models were established with difference of band reflectance,ratio of band reflectance and single band reflectance respectively,and the accuracy of inversion models were evaluated by the verification group.The results show that the model established by the reflectance of band mixed has a good fitting effect on total phosphorus,turbidity,suspended matter concentration,water temperature and transparency satisfactorily.