引用本文: | 王云霞,杨国范,林茂森,等.基于landsat卫星影像的水库水体总磷质量浓度反演研究[J].灌溉排水学报,2017,36(4):. |
| WANG Yunxia,YANG Guofan,LIN Maosen,et al.基于landsat卫星影像的水库水体总磷质量浓度反演研究[J].灌溉排水学报,2017,36(4):. |
|
摘要: |
以清河水库为研究区域,利用Landsat8卫星OLI数据及实地观测数据,建立了适于清河水库水体总磷质量浓度的最小二乘支持向量机(LS-SVM)遥感反演模型,并对清河水库总磷量进行了反演分析。结果表明,LS-SVM模型的平均相对误差为6.06%,相比于单波段线性回归反演模型及波段组合线性回归模型,平均误差分别降低了20.77%、12.53%,显著提高了清河水库水体总磷质量浓度反演的模型精度,达到遥感反演预测精度要求;利用LS-SVM反演模型对清河水库总磷量反演显示,水体中总磷质量浓度主要集中在0.04~0.08 mg/L,水库水体总磷质量浓度总体偏高。 |
关键词: 总磷; 遥感; 模型; 清河水库 |
DOI: |
分类号: |
基金项目: |
|
Calculating Total Phosphorus in Reservoirs Using the Satellite Landsat Data |
WANG Yunxia , YANG Guofan, LIN Maosen, YANG Shuting
|
Shenyang Agriculture University, Shenyang 110866, China; College of Water Conservancy and Hydropower Engineering, Hohai University, Nanjing 210098, China
|
Abstract: |
A least square support vector machine (LS-SVM) model was developed in this paper to establish the relationship between the data of OLI from landsat8 and the measured total phosphorus in Qinghe Reservoir. The total phosphorus (TP) in the reservoir was then inversely calculated based on the measured data from the LS-SVM. The results showed that the LS-SVM model increased the accuracy of estimated TP concentration with an average relative error of 6.06%, much lower than the 20.77% and 12.53%, the errors of the single band linear regression model and the band combination linear regression model, respectively. The results of The LS-SVM model also showed that the concentration of TP in Qinghe reservoir ranges from in 0.04~0.08 mg/L, higher than the normal range. |
Key words: the total phosphorus; remote sensing; model; Qinghe reservoir |