Study On Water Quality Risk Assessment And Its Response To Land Use Types In Gansu Section Of The Yellow River

Water quality plays an important role in ecological security,physical health and economic development.The identification and assessment of watershed water quality risk factors are the basis for determining water quality and establishing water quality policies.At present,there has been little research on water quality assessments,water quality risk identification and technological tools in large-scale watersheds.Gansu Province is located in the upper reaches of the Yellow River Basin,which is the primary source of drinking water for millions of people in Gansu Province.It is of great significance to carry out comprehensive water quality assessments and water quality risk identification in the Yellow River Basin for the health of residents along the river basin,ecological security in Northwest China and the socioeconomic development of Gansu Province.In this paper,based on the regional characteristics of the Gansu monitoring section and the relationship between water quality indices,the Canadian Water Quality Index method（CWQI）and BP neural network method were used for comprehensive water quality evaluation to identify water quality pollution factors.In addition,a model developed by the US Environmental Protection Agency（USEPA）was used to identify the water quality health risks in the Gansu section of the Yellow River.Based on water pollution and water quality health risk factors,a relationship model between land use types and water quality risk factors was established by using a redundancy analysis（RDA）and the partial least squares regression（PLSR）method to obtain large-scale water quality risk projections,which provide the scientific basis for water quality planning and comprehensive management in the Yellow River Basin in Gansu Province and lay the foundation for large-scale water quality projections.The main research contents and achievements of this dissertation are as follows:（1）Pollution risk analysis of water quality comprehensive evaluation based on the single factor,CWQI and BP neural network methods.The single factor method,the CWQI method and BP neural network method were used to carry out comprehensive evaluation and pollution risk analysis of the 2019 water quality of the Gansu section of the Yellow River.The results showed that the overall water quality in the Gansu section was good,reaching an average to good level.Total nitrogen（TN）and faecal coliform were the main pollution indicators affecting the water quality of the section.When they were not included in the evaluation,the water quality basically reached the water quality requirements of functional zoning.The CWQI evaluation results show that the water quality in Gansu section of the Yellow River is at average to good level.The proportion of the sections with average water quality is 72.2%,the proportion of the sections with good water quality is16.7%,and the proportion of the sections with poor quality is 11.1%.The trained BP neural network water quality evaluation results show that the water quality in Gansu section of the Yellow River is mainly good,accounting for 77.8%;the section with excellent water quality accounts for 11.1%;the section with general water quality accounts for 11.1%;there is no section with poor or extremely poor water quality.The results of the principal component analysis showed that the main sources of water pollution risk in the Gansu section of the Yellow River were organic matter,TN and microbial index.（2）Assessment and analysis of water quality health risks in the Gansu section of the Yellow River based on the adaptive adjustment of exposure parameters in the USEPA model.Based on the exposure parameters of the USEPA model,such as age,body weight,drinking water volume,exposure frequency,skin contact area and contact time,an adaptive adjustment was made according to the actual situation in Gansu Province,and a water quality health risk assessment under drinking water and skin contact routes was carried out.The results showed that the concentrations of metals were all lower than the limit of ClassⅠwater in the Surface Water Environmental Quality Standard（GB3838-2002）.The heavy metal Nemerow index and HPI index were all less than 1 and 50,respectively,indicating that the heavy metal pollution level in water was mild.The total health risk of each section in the Gansu section of the Yellow River was 1.61×10^-4～1.91×10^-3(a^-1),which exceeded the maximum acceptable level of health risks recommended by the USEPA.The heavy metal factors that cause health risks are As and Cr,which mainly come from industrial sources,such as mining,smelting and coal burning;domestic sources,such as traffic,waste settlement and household waste;and agricultural non-point sources,such as chemical fertilisers and pesticides.（3）Study on the response relationship between land use type and water quality in theGansu section of the Yellow River based on RDA and the PLSR method.The characteristics of land use in the Gansu section of the Yellow River were analysed based on remote sensing technology.The optimal interpretation scale of land use to water quality was determined by a variance decomposition method;the specific relationship between land use and water quality was determined by redundancy analysis method;the relationship modelling was carried out by the PLSR method;and a set method system for predicting future water quality changes based on land use type was established.The accuracy of the model was tested.The results showed that the proportion of forest to grassland areas was positively correlated with organic pollution,such as CODcr,COD_Mn and NH₃-N.Spatial scales,construction land and the Patch density（PD）index were positively correlated with TN,NH3-N and faecal coliform in water.Aromatic diversity index（SHDI）was negatively correlated with CODcr and COD_Mn.Increasing the diversity of land use types and landscape patterns was conducive to reducing organic pollutants in water due to the riparian buffer zone,the fitting degree of regression model is greater than 0.6.Water quality indices such as DO,COD_Mn and faecal coliform have a good multiple linear relationships with land use type.（4）Water environment management countermeasures in Gansu section of the Yellow RiverThrough comprehensive water quality evaluation and water quality health risk evaluation by various methods,it is concluded that the water quality evaluation result of Gansu section of the Yellow River water source conservation area is lower than that of other sections,and the overall health risk value of the upstream area is lower than that of the middle and downstream area.Water environment management departments should continue to pay attention to organic pollution,nitrogen pollution,excessive microorganisms and other problems in the basin,strengthen the investigation and management of point source pollution,strengthen the strict monitoring of heavy metals Cr and As and,especially,strengthen the management of water quality risks in the water source area.In urban construction,the planning department should try to control the development scale of construction land within 1 km of the riparian zone,implement an ecological red line and reasonably and scientifically control the development and utilisation of land.Agricultural departments should develop intensive agriculture policies and attempt to arrange farmland and other land use types that have positive correlations with water quality indices in an area 1～5 km away from rivers.Municipal management should combine regional characteristics,enhance the richness of landscape patterns and reduce the dispersion of unused land in the riparian zone to reduce the occurrence of water and soil erosion.The water quality of the Gansu section of the Yellow River was comprehensively evaluated by the single factor,CWQI and BP neural network methods.In addition,the exposure parameters in the USEPA health risk ratings model were adjusted adaptively to evaluate the water quality of the Gansu section of the Yellow River.The relationship model between land use type and water quality based on PLSR,RDA,variance partitioning analysis and other technologies can be applied to some water quality indices.This can then predict water quality changes in the future,as well as evaluate water quality risks,providing support for the establishment of water quality protection and management systems.It also lays a foundation for large-scale watershed quality projections.