基于光谱技术的太湖蓝藻预警因子研究

On Early Warning Indicators of the Blue-green Algae in Taihu Lake Using Reflectance Spectroscopy

针对近年来太湖富营养化过程的逐渐加剧,湖泊的沉积物中N、P、C含量大幅上升,导致藻类大量繁殖并形成水华现象.2011年4月在太湖进行采样和测量,总共获取44个采样点数据,利用紫外、中红外光谱法对太湖蓝藻预警因子总氮、总磷、叶绿素a进行快速测定,构建太湖的水质参数与光谱的关系模型.实验分析结果可以看出:太湖蓝藻预警因子是预测蓝藻及蓝藻水华产生的重要参数;基于中红外光谱的水体总氮和叶绿素a浓度模拟精度较高(R2>0.62);基于紫外的水体总氮、总磷、叶绿素a浓度的模型模拟值与真实值的相关性达到0.78以上.研究表明,基于紫外、中红外光谱法可以快速测定太湖蓝藻预警因子总氮、总磷、叶绿素a,从而预测太湖蓝藻是可行的.

In recent years,as eutrophication of Taihu Lake mounts up,increasing N,P,C contents in lake sediments lead to algae blooms.Forty-four typical stations were selected to measure ultraviolet and mid-infrared spectroscopy in Taihu Lake in April,2011 and three water quality parameters were analyzed by every station, including Total Phosphorous (TP),Total Nitrogen (TN),Chlorophyll-a (CHL) in the study.The correlations between the water quality parameters and mid-infrared reflectance were analyzed based on synergy interval partial least squares (SiPLS).The results show that,for TN and CHL,the SiPLS models have the smallest RMSE,and TN and CHL can be detected efficiently by in situ mid-infrared spectral analysis (R2>0.62).Similarly,TN,TP and CHL can be monitored efficiently by ultraviolet bands with SiPLS models (R2>0.78).The study indicates that early warning indicators of the blue-green algae can be monitored by ultraviolet and mid-infrared spectroscopy technology,which can provide the water supply and conservancy authorities with referenced spatial information to manage water resources.