摘要
利用遥感方法对大伙房水库营养状态进行评价。首先,确定了适用于大伙房水库遥感反演透明度和叶绿素a的最佳波段组合;其次,建立2指标的波段组合和最小二乘支持向量机(LS-SVM)模型并评估模型精度;再次,确定LS-SVM模型为水库2项指标预测模型,计算2指标综合营养状态指数并进行营养状态分级;最后,成功将LS-SVM模型应用于2017年7月2指标反演,并利用2指标综合营养状态指数法评价水库水体营养状态。结论:LS-SVM模型适用于2指标的反演预测,2项指标综合营养状态指数法适用于大伙房水库水体的营养状态评价。
This study had assessed the trophic state of Dahuofang reservoir by remote sensing. Firstly, the optimal band combination used for transparency and chlorophyll-a calculation had been analyzed. Secondly, the band combination model and least squares support vector machine model( LS-SVM) for estimating the concentration of chlorophyll-a and the value of transparency had been built up, meanwhile, the accuracy had been assessed. Thirdly, LS-SVM had been used for estimating those value, and the trophic state index model TLI( ∑) had been built up with chlorophyll-a and the transparency. Lastly, the LS-SVM model for chlorophyll-a and the transparency and the two factors' TLI( ∑) for trophic state assessment had been successfully applied to satellite image in July,2017. To sum up, LS-SVM was a suitable model for estimating the concentration chlorophyll-a and the value of transparency, two factors' TLI( ∑) could also be used to assess the trophic state in Dahuofang reservoir.
作者
谢志钢
XIE Zhigang(Hydrology Bureau of Liaoning Province,Shenyang 110003,China)
出处
《人民珠江》
2018年第7期115-119,共5页
Pearl River
关键词
遥感评价
营养状态指数
最小二乘支持向量机
remote sense assessment
trophic state index
least squares support vector machine