Quantitative Analysis Of Water Quality Indexes Based On Laser Induced Breakdown Spectroscopy And Hyperspectral Technology

With the development of hyperspectral remote sensing technology,water quality monitoring has changed from qualitative description to quantitative analysis,but there are still some problems,such as low spatial resolution and real-time,huge amount of data and redundancy of information.Laser induced breakdown spectroscopy(LIBS)can accurately analyze the simultaneity of multi-elements in multi-phase samples(solid,liquid and gas),and has the ability of complete element analysis and remote analysis.Therefore,based on the theories of geography,ecology and environmental sciences,using the accumulated water quality data of rivers and lakes,remote sensing data,ground measured spectral data and laser-induced breakdown spectra as data sources,this study attempts to retrieve the parameters of inland waters from the traditional ground measured hyperspectral data.In order to discuss the estimation accuracy,maneuverability and application convenience of the two techniques of surface hyperspectral and LIBS in the quantitative analysis of water parameters,the quantitative analysis of water quality based on laser-induced breakdown spectroscopy was compared with that based on LIBS.In addition,on the basis of comparative study on the quantitative analysis of water body parameters by these two techniques,the optimal model and algorithm for quantitative analysis of water body parameters are attempted to be constructed through the combination of algorithmic methods,in order to construct the optimal model for monitoring water quality indicators in arid areas,and to improve the accuracy and efficiency of water quality indicators detection.The conclusions are as follows:(1)According to the inversion model results of total phosphorus and total nitrogen concentration based on the characteristic peak intensity and baseline area of total phosphorus and total nitrogen in LIBS spectral curve,the estimation model based on the baseline area of characteristic peak has higher overall accuracy and stronger reliability,and can accurately estimate the total phosphorus and total nitrogen concentration in water body.(2)Compared with the hyperspectral model of total phosphorus and total nitrogen water quality parameters constructed by spectral transformation and spectral index alone,the model based on spectral index and spectral transformation co-inversion has higher accuracy.The determination coefficient R~2 of the model is about 0.6,which can better estimate the total phosphorus and total nitrogen concentration.(3)Compared with the single estimating model of total phosphorus and total nitrogen based on LIBS characteristic peak baseline area,hyperspectral remote sensing image and measured hyperspectral data,the combined forecasting model based on LIBS characteristic peak baseline area and hyperspectral remote sensing image,or the combined forecasting model based on LIBS characteristic peak baseline area and measured hyperspectral data,can effectively improve total phosphorus.The prediction accuracy of the nitrogen concentration estimation model reflects the feasibility of the combined prediction model method.It not only makes full use of the advantages of each model,but also minimizes the prediction error and makes the prediction model more stable and adaptable.