针对洪涝灾害直接损失动态刻画不足及单一模型难以兼顾不同时间尺度间接影响的问题，构建基于损失率函数的洪涝灾害损失动态评估方法。通过多因子比拟、Beta分布和多函数拟合构建区域损失率函数，将淹没水深与历时纳入直接损失计算，并采用IO和CGE模型评估间接损失。以巢湖流域2020年特大洪水为例，结果表明：①直接经济损失随洪水演进呈“缓慢增加—快速增加—峰值稳定”特征，反映水深、历时及受淹范围共同作用下的损失累积；②模型评估直接经济损失190.61亿元，与官方统计值201.33亿元相比，相对误差为5.32%，总体吻合较好；③IO和CGE模型评估的间接经济损失分别为318.61亿元和246.09亿元，为直接损失的1.67倍和1.29倍，凸显了产业关联对经济损失的放大作用，两类模型分别反映短期产业链传导与中长期经济调整，具有互补性。研究结果可为洪涝灾害损失评估及防灾减灾决策提供科学依据。

[Objectives] Existing flood-loss assessment methods often have difficulty characterizing the temporal accumulation of direct economic losses throughout flood evolution, while a single economic model cannot adequately represent indirect economic impacts at different time scales. To address these limitations, this study develops a dynamic flood-loss assessment method based on loss-rate functions to describe the time-varying evolution of direct economic losses and evaluate indirect economic losses from complementary economic perspectives.[Methods] The catastrophic flood in the Chaohu Lake Basin in 2020 was selected as the case study. Regional flood loss-rate functions were established through multi-factor analogical transfer, Beta-distribution probability analysis, and multiple-function fitting. Transfer factors representing differences in economic development, flood-control capacity, and land-surface characteristics were used to transfer historical loss-rate samples from reference regions to the study area. Candidate loss rates at different inundation-depth levels were screened according to their dispersion, and representative values were determined from the maximum-probability values of the fitted Beta distributions. Multiple functional forms were then compared to derive loss-rate functions for different exposed elements. For direct-loss assessment, spatialized socio-economic data were overlaid with simulated inundation depth and duration at successive time steps. Both inundation depth and duration were incorporated into asset-loss and output-loss rates, and annual output values were converted according to inundation duration. Direct economic losses were accumulated over time to obtain their temporal evolution. For indirect-loss assessment, an input-output (IO) model and a computable general equilibrium (CGE) model were applied separately. The IO model characterized short-term propagation of production shocks through relatively fixed inter-industry linkages, whereas the CGE model represented medium- to long-term adjustments involving price responses, supply-demand changes, sectoral substitution, and resource reallocation.[Results] The results show that: (1) Direct economic losses exhibited a three-stage pattern of “slow increase-rapid increase-peak stabilization” during flood evolution. In the early stage, the inundated area was limited, with shallow water depths and short inundation durations in most affected units, resulting in slow loss accumulation. As river and lake water levels rose, inundation expanded toward low-lying areas, while inundation depth and duration increased simultaneously, causing more exposed elements to enter higher loss-rate intervals and cumulative losses to rise rapidly. When the flood remained at a high but relatively stable stage, newly inundated areas decreased and changes in inundation depth and duration weakened, so cumulative losses gradually approached a stable peak. (2) The model-estimated direct economic loss was CNY 19.061 billion, compared with the officially reported value of CNY 20.133 billion, corresponding to a relative error of 5.32%. The close agreement indicates that the proposed method can reasonably reproduce the overall magnitude of direct economic losses caused by the 2020 flood in the Chaohu Lake Basin. (3) The indirect economic losses estimated by the IO and CGE models were CNY 31.861 billion and CNY 24.609 billion, respectively, equivalent to 1.67 and 1.29 times the direct economic loss. Both exceeded the direct loss, demonstrating a pronounced amplification effect of inter-industry linkages. The IO model produced the higher estimate because production shocks were transmitted through relatively fixed input-output relationships, whereas the CGE model incorporated adaptive responses such as price adjustment, sectoral substitution, and resource reallocation. The two models therefore reflect different economic mechanisms and time scales and provide complementary information for indirect flood-loss assessment.[Conclusions] Flood-related economic losses are characterized by temporal accumulation during inundation and amplification through inter-industry linkages. By integrating dynamic inundation characteristics with regionalized loss-rate functions, the proposed method can describe the temporal evolution of direct economic losses. The complementary use of the IO and CGE models enables indirect economic impacts to be assessed from the perspectives of short-term industrial-chain transmission and medium- to long-term economic adjustment. The framework has potential for application to other regions, but the loss-rate functions and related socio-economic parameters are region dependent; therefore, regional recalibration and validation using local flood-loss data are required before transfer to other study areas. The results can provide quantitative support for flood-loss assessment, stage-specific emergency response, and disaster prevention and mitigation decision-making.