Research on the Quantitative Sensing Images Inversion Model of Cyanobacteria Bloom in the Xiangjiang River Based on Back Propagation Neural Network
In recent years,the frequent occurrence of cyanobacterial blooms in some lakes and rivers has led to an increasing amount of related research.Based on the cyanobacterial bloom monitoring data of the Xiangjiang River on September 26,2021 and Sentinel-3 satellite images,taking the total algae density as the research object,the b17/b14 ratio band combination is determined as the feature band,and a quantitative inversion model for cyanobacterial blooms in the Xiangjiang River is constructed using the correlation coefficient method and Back Propagation(BP)neural network method.The regression value(R)of the model is 0.95465,the root mean square error is 244×104 cells/L,and the Nash efficiency coefficient is 0.90,indicating that the model is relatively reliable.The validation results show that the model is suitable for the inversion of cyanobacterial blooms with total algae density between 1200×104 cells/L and 5000×104 cells/L,with a relative error of less than 60%for 80%of the validation points.Compared with the results obtained by Maximum Chlorophyll Index(MCI),the consistency rate of cyanobacterial bloom in all sections of the Xiangjiang River is above 80%,and the consistency rate in Changsha and Yueyang sections was above 90%.It indicates that the model can be used for the quantitative inversion of cyanobacterial blooms in the Xiangjiang River.
cyanobacterial bloomtotal algal densitySentinel-3neutral networkquantitative remote sensingthe Xiangjiang River
万方数据
维普
