Inversion Of Water Quality Parameters Of Fuyang River In Handan City Based On Multi-source Remote Sensing Data

Water is an important resource needed to sustain human survival and development.The quality of water directly affects people’s water security.Therefore,the monitoring,management and protection of water resources are extremely important.Traditional water quality monitoring is time-consuming and laborious,and it is difficult to show the overall situation of water quality.In recent years,the development of remote sensing technology has become a new regionalization of water quality monitoring methods.Because of the use of remote sensing technology water quality monitoring has a large monitoring range,fast,low cost,can be real-time dynamic and other advantages,it has become a current research hot spot.Taking the Fuyang River basin of Handan city as the study area,this paper uses water quality sampling and testing data,UAV remote sensing data and quasisynchronous satellite remote sensing data from five river sections,South Lake and North Lake of Fuyang River to invert turbidity,suspended solids concentration and ammonia nitrogen,and studies the "UAV+satellite" remote sensing inversion technique for regional water quality monitoring.This study provides a new monitoring technology and data support for the monitoring and protection of water environment in Handan section of Fuyang River.The main research contents and conclusions of the paper include the following three aspects.(1)Based on the pre-processed UAV remote sensing data,the spectral parameter models such as difference index(DI),ratio index(RI)and normalized index(NDI)were constructed to analyze the correlation between the spectral parameters and the actual measured water quality data such as turbidity,suspended solids concentration and ammonia nitrogen.The experimental results showed that the parameters with the highest correlation corresponded to the normalized index of 550 nm and 710 nm,the difference index of 550 nm and 710 nm,and the normalized index of 550 nm and 660 nm,respectively.Based on the above wave combinations,one-dimensional linear models,polynomial models and exponential models were established,and multiple linear regression models for each water quality parameter were established using partial least squares regression methods.The comparison analysis showed that the best inversion model for turbidity and ammonia nitrogen was the multiple linear regression model.The best inversion model for suspended solids concentration was a cubic polynomial model.(2)Quasi-synchronous Sentinel-2 remote sensing data were downloaded for preprocessing,constructing spectral parameter models such as difference index,ratio index and normalized index,and analyzing the correlation between spectral parameters and laboratory assay water quality parameter data of 45 sampling points in South Lake and North Lake.The experimental results showed that the spectral parameters with the highest correlation were 490 nm and 705 nm,and the best inversion models for turbidity,suspended solids concentration and ammonia nitrogen were finally obtained by using the above-mentioned waveband and waveband combinations to construct a one-dimensional linear regression model and a multivariate linear regression model.(3)Using the best inversion models constructed by the two data sources to invert the water quality parameters,it was found that the water quality parameters in the Handan section of the Fuyang River generally showed a distribution of low in the south and high in the north.The three water quality parameters in the South Lake,Zhangzhuang Bridge section,South Ring Road section,Renmin section and the turbidity and suspended matter concentrations in the Orchard Road section were relatively low.The concentration of ammonia nitrogen in the Orchard Road section was the second highest.The Suli section and the North Lake had the most serious pollution.The inversion results and sampling experimental results were basically consistent.