湖泊水质下降和富营养化问题日益严重,传统监测方法已经日益不能满足水质监测的需要。该文通过野外高光谱仪器测定查干湖水体反射光谱,分析研究水体反射光谱特征与水质参数叶绿素a含量、透明度和浊度之间的关系,运用多种半经验算法建立反演模型。结果表明:单波段光谱反射率与叶绿素a和浊度的相关系数较小,但与透明度相关系数较高;光谱反射率通过对数、比值转换,可以有效的提高叶绿素a和浊度估测模型的精度,却对透明度作用不大。总体来说,叶绿素a和浊度估测模型效果较好,透明度高光谱估测模型相对较差,需进一步研究。
The lakes are one of the important natural resources, however, because of the strong human influence, water quality is seriously polluted and need monitoring. Because the routine monitoring method is time-consuming and strenuous, and remote sensing technology, especially hyperspectral remote sensing, can solve these problems effectively. This research tried to study the relations between water body reflectance and three main water quality variables, such as chlorophyll-a contents, transparency and turbidity, and set up three hyperspectral estimate model. The result showed: The coefficient that single band with chlorophyll-a and turbidity was relatively little, but the transparency coefficient was high; Spectral reflectance transformed through logarithm and ratio could effectively improved the precision of chlorophyll-a and turbidity model to estimate, but no help to transparency.