Remote Sensing Retrieval Model For Chlorophyll-a Concentration Of Water In Backwater Area,Three Gorges Reservoir

Chlorophyll-a concentration is the most important index to reflect the eutrophication of water body.At present,the remote sensing retrieval methods of chlorophyll a concentration in two inland water bodies are mainly concentrated in ordinary lakes,such as Taihu,Poyang Lake and so on.There are few studies on the stream lakes in the backwater area of the Three Gorges Reservoir area.In this study,the Three Gorges Reservoir Area Gaoyang lake,backwater area of Han Feng Lake as the study area,using the GF-1 WFV remote sensing data and the measured data,the establishment of remote sensing model,large area dynamic inversion of the tributaries of the Three Gorges reservoir backwater area of chlorophyll a concentration.The research work and main conclusions are as follows:(1)The use of GF-1 WFV remote sensing data and measured the concentration of chlorophyll a,establish the band ratio model of regression analysis,inverse backwater area,chlorophyll a concentration,band ratio model formula for X for remote sensing image band combination of B4/B3 values,the root mean square error of RMSE is 5.6739,the coefficient of determination is 0.68672-R.Band ratio model and the inversion effect is not ideal,the reason for the backwater area of Three Gorges Reservoir area is two water,the optical properties of the complex,application of a class of linear inversion water model fitting is one-sided,it is difficult to obtain the best fitting effect.(2)Compared with the linear regression method of band ratio,the BP neural network model has the function of simulating complex nonlinear problems.This study established the model structure of BP neural network model of 4-4-1,by correlation analysis,B4-B3,B4-B2,B4/ selection(B2+B3)and B4/B3 four band combinations do 4 input neurons of the BP neural network model,set for single hidden layer,hidden layer node number is 4,the measured concentration of chlorophyll a value as the output neuron model training.After R-BP neural network model after training was 0.8389,RMSE was 3.8745.The GF-1 WFV images extracted from the waters were used for the trained BP neural network model,and the chlorophyll a concentration distribution in the backwater area of the Three Gorges Reservoir Area in May 2016-8 was retrieved,and the inversion results were good.(3)The comparison of two kinds of chlorophyll a inversion accuracy of the model,the coefficient of determination R-BP neural network model and band ratio models were 0.8389 and 0.6867;the root mean square error of RMSE were 3.8745 and 5.6739;the average relative error of E were 20.6% and 55.9%.The comparison results show that the GF-1 WFV model is applied to retrieve the chlorophyll a concentration in the backwater area of the Three Gorges Reservoir Area by using the BP neural network model.The accuracy of the model is higher than that of the band ratio model,and the inversion effect is better than the band ratio model.