Analysis Of The Ba River Basin Water Pollution And Exploration Of UV-Vis Pollution Traceability

Urban rivers provide a suitable living environment for humans and other organisms.With the rapid development of human society,the discharges of urban sewage have increased dramatically,resulting in a plethora of river pollution problems.During the 14 th Five-Year Plan period,Xi’an City in Shaanxi Province set up 6 national examination monitoring sections,of which 2 are located in the downstream of Bahe River Basin.The absence of monitoring sections in the upstream of Bahe River Basin makes it difficult to trace the source of pollution if the water quality in the control section exceeds the standard.At present,traditional manual investigation and traceability are inefficient,costly,and easy to lose timeliness.There is an urgent need to construct a low-cost,simple and efficient identification model or system of pollution sources for monitoring urban rivers.The ultraviolet-visible absorption light spectroscopy(UV-Vis)analysis method has the advantages of low cost,high efficiency,greenness and simplicity,and is one of the important technical means for monitoring the dynamic changes of the regional water environment and early warning of river pollution traceability.Therefore,this study selected the whole section of Bahe River Basin as the research object,based on the analysis and diagnosis of pollution factors and pollution sources in the study area,combined with the MIKE 11 hydrodynamic and water quality models,the simulation results were visualized and demonstrated by Arc GIS spatial analysis module,thus estimating the scope of pollution sources in the whole section of Ba River.Through the field sampling in the pollution source range,the main pollution sources affecting the water environment in the river basin are identified by V-Vis and MATLAB and the corresponding control countermeasures are also proposed.Besides,the "fingerprint database" of pure pollutants was designed according to typical pollution emergencies,the influence of different training groups on performance of it was explored,and the best model for predicting pollutant concentration was established.This research can improve the identification efficiency of pollution sources,reduce the manual identification process and costs,provide reference for the management of the water environment in the Bahe River Basin and the future expansion of the "fingerprint database" of different water bodies.The main research contents and conclusions are as follows:(1)The water environment problems in the Bahe River Basin were diagnosed through the evaluation method of three pollution factors and the improved output coefficient method.The results of the pollution factor evaluation show that the water quality during the flat water period was better than the abundant water period and the dry period in Bahe River Basin.At the spatial scale,the high values of water quality indicators in the river are mainly concentrated in the Xianning Bridge of Chanhe River ～ the confluence of the Chanhe and Bahe river ～ Sanlang Village,and the low values are concentrated in Lantian County ～ the confluence of the Chanhe river and Bahe rivers.The results of the pollution source evaluation indicate that the main source of pollution in the Bahe River Basin is non-point source pollution.Among them,COD emission is the largest,followed by TN,and TP emission is the smallest.From the perspective of pollutant discharge methods,chickens in livestock and poultry breeding have the largest pollution emissions,followed by rural populations and the smallest discharge from grassland.The calculation results of the districts and counties in the river basin show that the areas with low non-point source pollutant load intensity are mainly distributed in the northwest side of the downstream of the Bahe River and the southwest side of the middle of the Bahe River,and the areas with high non-point source pollutant load intensity are mainly distributed in the upstream of the Bahe River,midstream and downstream of the Chanhe River,and the confluence of the Chanhe river and Bahe river.(2)The migration and conversion process of major pollutants was simulated and the scope of pollution sources was estimated in the Bahe River through the coupling model of hydrodynamic and water quality.According to the operation results of hydrodynamic and convective diffusion modules,it can be seen that the Nash-Suttcliffe coefficient of both is greater than 0.5,and the model reflects better the change process of hydrodynamics and water quality of the river.Arc GIS is used to visualize the simulation results,and the results suggest that the areas with high concentrations of COD and TP are mainly distributed in the northeast of Baqiao District during the flood period.The areas with high TP concentration during the dry period are mainly concentrated in Chanhe River and the section from the confluence of the Chanhe river and Bahe river to Sanlang Village,and the areas with high COD concentration are mainly distributed in the confluence of the Chanhe river and Bahe river ～ Sanlang Village.The areas with high COD and TP concentration during the flat water period are mainly at the confluence of the Chanhe river and Bahe river.The areas with high concentration pollutants during the three periods are roughly located at the confluence of the Chanhe river and Bahe river ～ Sanlang Village,and the water quality of Lantian County in upstream of the Bahe River was good.(3)The main pollution sources in the Bahe River Basin were identified through UV-Vis pollution traceability technology.8 classification models were established by different algorithms,and the "PCA+KNN" classification model was selected as the best model combining the efficiency and performance of the classification models.Base on the operation results of the optimal model,it can be seen that the pollution sources from the confluence of the Chanhe River and Bahe River to Sanlang Village were more likely from the Xi’an World Expo Park and the Fifth Sewage Treatment Plant,besides the Pearson correlation was used to prove that the UV-Vis traceability model was reliable.Based on the traceability results,many pollution control countermeasures were proposed,for example,the above the Yu Kou Chanhe River should focus on water connotation and ecological protection,and the above the Yu Kou Chanhe River should focus on the embankment treatment and ecological restoration,the sediment dredging is carried out in the rubber dam in the confluence of the Chanhe river and Bahe river ～ Sanlang Village;the precise regulation of the rubber dam in the confluence of the Chanhe river and Bahe river is implemented to increase the flow velocity and speed up the diffusion and dilution effect of the river.(4)The UV-Vis "fingerprint database" of pure pollutants was designed based on the type of river pollution emergency.The optimal physical conditions for establishing the database were also explored by the mathematical,and the results showed that the number of training groups had no significant influence on the model performance,and the accuracy,sensitivity and accuracy of the models with different training groups were relatively high,all of which exhibited good generalization and robustness.In the same model,the number of training sample groups(40～80)has a linear relationship with the running time,the running time rose with the increase of the number of sample groups;when the number of training sample groups exceeds80,the relationship between the number of model sample groups and the running time was not obvious.When the number of training samples reached 30,“PCA+KNN” was the best model.The constructed regression model by the spectral data of pure chemical solution can well predict the concentration of target substances in the mixed water sample.The model is less susceptible to interference from other substances in the water sample,and suitable for the quantitative analysis of pollution prediction in the later stage.