汤逊湖水体叶绿素浓度遥感估测研究

Research on remote sensing estimation of chlorophyll concentration in water body of Tangxun Lake

内陆水体叶绿素浓度遥感监测的各种经验模型在模型的方法和参数上都有一定的时空限制。为了实现武汉汤逊湖叶绿素浓度的遥感监测,利用同步实测的光谱数据和水质数据进行了该湖泊叶绿素浓度的遥感定量反演研究。首先分析了汤逊湖水体的反射率光谱曲线;其次通过相关分析的方法,发现了汤逊湖水体反射率在单波段法、一阶微分法和比值法建模中的最优波段或组合,并进行了一元线性建模;最后对3种模型的反演精度进行了比较分析。结果表明,对于汤逊湖叶绿素浓度遥感反演,一阶微分法和比值法优于单波段法,一阶微分法的最优波段在446.9 nm附近,比值法的最优波段组合为R861.1/R865.7,二者回归R2都在0.86以上。本项研究可为汤逊湖叶绿素浓度的遥感估测提供了参考。

Water quality plays a key role to the water management,especially for inland water（e.g.lake,reservoir,and river）,due to the increasing stress on the available water resources.Chlorophyll is one of the important inland water quality parameters,while existing models of chlorophyll concentration extraction from remote sensing data are limited to some spatiotemporal study sites.To monitor chlorophyll concentration by remote sensing in the Tangxun Lake of Wuhan,field-observed data and reflectivity spectrum were used to study the chlorophyll concentration.First,the field-observed reflectivity spectrum curve of the water body in Tangxun Lake was analyzed.Second,correlation analysis was taken to get the most suitable bands from the three algorithms of single band method,first differential method,and the method of spectral reflectivity ratio at different bands,and constructed related models for Tangxun Lake.Last,the accuracy of three models was compared and the most suitable model to monitor chlorophyll concentration in Tangxun Lake by remote sensing was obtained.For the extraction of chlorophyll concentration in Tangxun Lake,the first differential method and the method of spectral reflectivity ratio at different bands are better than the single band method.The optimum band of the first differential method is close to 446.9nm and the optimum bands combination of method for ratio of spectral reflectivities is R861.1/R865.7.The R2 of the two models both exceed 0.86.The finding of the research could give a reference to remote sensing estimation of chlorophyll concentration in the water body of Tangxun Lake.