水污染排放清单是流域水质模拟与排放管控的重要工具。面对污染源类型多样、高频排放监测数据稀缺、污染输入-水质响应关系复杂等挑战，亟需构建具有清晰层级结构、可拓展可计算、支撑流域系统治污的精细化多源污染排放清单，并建立规范化的污染排放清单研制方法与相应的评估模型。 本文以污染输入-水质响应关系为基础，构建了流域精细化水污染排放清单（Water Pollution Emission Inventory with High Spatio-Temporal Resolution，简称WEIHR清单）技术与评估方法。首先，构建了WEIHR清单分类分级体系，采用数据同化技术融合多源数据，运用时空降尺度、数据挖掘、参数拟合等稀疏数据处理方法，建立排放源-活动水平-排放负荷-单一账户核算-多账户清单的完整排放映射关系，构建WEIHR清单并分析了污染排放时空特征。其次，基于环境系统分析方法建立排放清单可靠性评估框架，运用水质模型、全局灵敏度分析、蒙特卡洛模拟、贝叶斯分析等技术对WEIHR清单准确性、参数灵敏性和不确定性及其对水质的叠加性影响进行了系统评估。最后，将WEIHR清单应用于污染输入-水质响应关系分析，定量评估了不同排放管控策略对流域水环境质量的改善效果，识别出流域污染排放精细化管控的可行策略集。 本文以白洋淀上游府河流域为应用案例研究对象。研究成果表明，WEIHR清单整体可靠，清单模拟水质与观测水质逐日平均相对误差小于30%，准确性较传统清单提高27.5%。清单能够较好刻画单一污染账户及全口径账户整体排放特征，单污染源清单污染排放时间变异系数为0.23-6.32，空间变异系数为0.20-3.65，时空变化识别精度较传统清单提高0.2-3.0。清单径流曲线数CN值、水质模型污染物降解系数等参数全局灵敏性较大，参数灵敏度与模拟机理吻合，清单结构可靠。清单不确定性经水质模型传递后，联合不确定性增加2-3%，全年联合不确定性约为13%，模拟水质不确定性分布对观测水质的涵盖度达到0.85以上（95%置信水平）。排放控制策略研究表明，生活源、城市地表径流源与种植源管控策略削峰强度识别效果较传统清单增加0.3-8.9%。其中，生活源策略可削减排放总量20%-56.7%，对平枯水期的污染减排发挥重要作用，可减少研究期约90%超标天数；城市地表径流源策略虽对污染排放总量削减仅贡献0.6%-5%，但对污染物排放峰值强度削减贡献了5-10%，对丰水期污染减排贡献较大。

Water pollution emission inventory is an important tool for watershed water quality modeling and emission control. Faced with the challenges of diverse types of pollution sources, scarcity of high-frequency emission monitoring data, and complex relationship between pollution input and water quality response, there is an urgent need to build a refined multi-source pollution emission inventory with a clear hierarchical structure, scalable and calculable, and capable of supporting watershed system pollution control, and to establish a standardized pollution emission inventory development method and corresponding evaluation model.Based on the relationship between pollution input and water quality response, this paper constructs the technology and evaluation method of the Water Pollution Emission Inventory with High Spatio-Temporal Resolution (WEIHR). First, this study constructs the WEIHR classification and grading system, and uses data assimilation technology to integrate multi-source data. At the same time, methods such as spatio-temporal downscaling, data mining, and parameter fitting are used for sparse data processing, and a complete emission mapping relationship of emission source-activity level-emission load-single account accounting- multi-account inventory is established. The WEIHR is constructed and used to analyze the spatio-temporal characteristics of pollution emissions. Secondly, based on the environmental systematic analysis method, this study establishes a reliability assessment framework for the WEIHR, using water quality model, global sensitivity analysis, Monte Carlo simulation, Bayesian analysis and other techniques to systematically assess the accuracy, parameter sensitivity and uncertainty of the WEIHR and its additive effects on water quality. Finally, the WEIHR is applied to the analysis of the relationship between pollution input and water quality response, quantitatively evaluates the improvement effect of different source emission control strategies on the water environment quality of the basin, and identifies a feasible set of strategies for refined management and control of pollution emission in the basin.This paper takes the Fuhe Basin in the upper reaches of Baiyangdian as the case study. The research results show that the WEIHR inventory is overall reliable, the mean relative error between the simulated water quality and the observed water quality is less than 30%, and the accuracy is 27.5% higher than that of the traditional inventory. The inventory can better describe the overall emission characteristics of single pollution accounts and full-caliber accounts. The time variation coefficient of pollution emission from the single pollution source is 0.23-6.32, and the spatial variation coefficient is 0.20-3.65. Compared with the traditional inventory, the identification accuracy of temporal and spatial changes is improved by 0.2-3.0. The parameters such as the CN value of the inventory runoff curve and the pollutant degradation coefficient of the water quality model are highly sensitive globally. The parameter sensitivity is consistent with the simulation mechanism, and the inventory structure is reliable. After the uncertainty of the inventory is transmitted through the water quality model, the joint uncertainty increases by 2-3%, and the annual joint uncertainty is about 13%. The coverage of the simulated water quality uncertainty distribution to the observed water quality is above 0.85 (95% confidence level). Research on emission control strategies shows that the identification effect of peak shaving intensity of living source, urban surface runoff source and planting source control strategy is 0.3-8.9% higher than that of traditional inventory. Among them, the living source strategy can reduce the total emission by 20%-56.7%, which plays an important role in the reduction of pollution during the normal and dry period, and can reduce the number of days exceeding the standard by about 90% during the study period. Although the urban surface runoff source strategy only contributes 0.6%-5% to the reduction of total pollution emissions, it contributes 5-10% to the reduction of the peak intensity of pollutant emissions, which is a great contribution to the reduction of pollution emissions during the flood season.