| Under the national strategy of continuously promoting ecological protection and high-quality development in the Yellow River Basin,comprehensive management is the core element for achieving sustainable development of the Yellow River.As a first-level tributary of the Yellow River in the middle reaches,Wuding River faces significant ecological health pressure due to climate change and the characteristics of fossil energy industries within its basin.Conducting a health assessment of Wuding River,analyzing and diagnosing ecological problems,and proposing corresponding protection and management measures are crucial for maintaining ecological security and restoring river health.This paper takes Wuding River in Shaanxi Province as the research object,analyzing and considering the integrity of the ecological health of the basin from two aspects:natural ecological environment and social service functions of the river.Based on the current ecological health status of Wuding River,the paper reveals the method and characteristics of constructing a health evaluation system for small and medium-sized rivers in such semi-arid regions.Utilizing a BP neural network model optimized by genetic algorithms to establish a weight learning mechanism for river evaluation,the paper explores the feasibility and applicability of machine learning algorithms in river health assessment.The health status of Wuding River is assessed,analyzing the spatial heterogeneity of different indicator results and the sources of unhealthy pressure,and proposing adaptive management recommendations.The main research conclusions of this paper are as follows:(1)In view of the characteristics of small and medium-sized rivers in the semi-arid regions of the north and the background of the fossil energy industry in the Wuding River basin,the river health evaluation indexes were expanded and adjusted,and a health evaluation system for the Wuding River was constructed,covering 18 health indicators in five aspects of ecological integrity,including hydrological and water resources,physical structure,chemical water quality,aquatic organisms and social service functions.(2)MATLAB was used to build a GA-BP neural network model to establish a weight learning mechanism,combining the traditional subjective and objective weighting methods to obtain the weight matrix of the Wuding River health evaluation indicators.The results show that the weighting of each criterion layer in the comprehensive proportioning system is as follows:water quality and chemistry>social service function>hydrological water resources>aquatic biology>physical structure,among which the rationality of the outfall layout in terms of chemical water quality has the most influence on the river health assignment,the social service function layer has the largest weighting in terms of water supply compliance,and the remaining hydrological water resources,aquatic biology and physical structure layers have the greatest influence on the health of the Wuding River.The remaining indicators that have a significant impact on the health of the river are the degree of ecological water satisfaction,the fish retention index and the longitudinal connectivity of the river.(3)Through the diagnosis of the health status of the Wuding River and its spatial heterogeneity,the final comprehensive health rating of the Wuding River in Shaanxi Province is two categories of health(62.12),of which the chemical water quality level as a whole is one category of health(72.88),with obvious differences in the chemical water quality status of different river sections and a reasonable layout of the upstream and midstream outfalls;the social service function of the Wuding River is in two categories of health(71.29),and the flood control rate is better than the public satisfaction and water supply assurance rate indicators.In terms of hydrological and water resources,the Wuding River as a whole is in a Class Ⅲ sub-healthy state(47.68),with flow variability in the downstream areas significantly better than in the middle and upstream areas,while the ecological flow assurance rate is poor in the whole area except for the two upstream evaluation sections,and the degree of soil erosion control is in a Class Ⅴ sick state.At the level of aquatic life,the river is in an unhealthy state(34.03),and the macrobenthic integrity and fish retention indices show a better distribution pattern in the upper reaches than in the middle and lower reaches;at the level of physical structure,the Wuding River is in a subhealthy state(57.21),with the stability of the riparian zone,the degree of vegetation cover and the longitudinal connectivity of the river in a poor state.(4)Based on the evaluation results,this study proposes corresponding protection and management measures:the level of development in the basin should be adjusted to match the environmental carrying capacity of Wuding River.Furthermore,the water resources management system should be strictly enforced,and a water use warning mechanism should be established to ensure stable water flow in the river channel,while accelerating the rate of soil and water loss control.The ecological environment health zoning and control of Wuding River should be implemented to improve water pollution and enhance ecological resilience.Attention should be paid to the hydrological and cultural construction of the basin,and the river health evaluation should be applied to the river chief system management system.In summary,this study constructed a health evaluation system for Wuding River,determined the weights using machine learning algorithms,revealed the health status and problems of Wuding River in Shaanxi Province,and proposed corresponding protection and management measures.The research results can serve as a reference for the construction of health evaluation systems for small and medium-sized rivers in similar semi-arid regions,and provide a scientific basis for the construction of the Yellow River’s digital twin basin. |
PDF Full Text Request