在城市化进程加速与气候变化的背景下，实现防洪韧性的精准评估对城市的可持续发展具有重要意义。现有的防洪韧性评估研究多采用经典统计方法，对小样本数据较为敏感，且未考虑空间相关性对城市防洪韧性的影响。该文采用基于Markov链Monte Carlo(MCMC)方法的Bayes时空交互模型(BSTIM)，将邻接城市的空间信息纳入时空演化过程，深入揭示防洪韧性在时间与空间上的耦合过程，最后将其应用于长三角地区的防洪韧性评估研究。结果表明：BSTIM对城市防洪韧性的时空解译度达到93%以上，对时空数据的拟合较好，实现了对防洪韧性时空特征的精准评估。长三角地区的防洪韧性在研究期间呈现上升趋势，空间布局整体上形成“低-高-中”的空间格局，局部变化趋势呈现“东高西低，北部居中”的特点。城市防洪韧性空间格局的冷热聚集差异性显著，局部变化冷热聚集形成“西冷东热”的格局；半数地区呈现“热-热”与“冷-热”的聚集特点，表明长三角地区城市抗洪的未来趋势向好。该研究成果可为城市防洪韧性发展重点的有效识别提供理论支撑与决策参考。

Objective: Against the backdrop of accelerated urbanization and climate change, accurately assessing flood resilience holds significant implications for the sustainable development of cities. Previous studies on flood resilience assessment have employed classical statistical methods, which are relatively sensitive to small-sample data and fail to consider the effect of spatial correlation on urban flood resilience. To mitigate disaster risks and enhance cities' capacity to withstand urban floods, this study focuses on examining the intrinsic spatial structures of flood resilience while accounting for spatial correlation and space-time heterogeneity. Methods: This study employed the Bayesian space-time interaction model (BSTIM) based on the Markov chain Monte Carlo method to achieve an accurate assessment of flood resilience. The model incorporated the spatial information of adjacent cities into the space-time evolution of flood resilience, enabling the revelation of the coupled space-time dynamics. In addition, a comparative study of the model was conducted to verify the effectiveness of the selected parameter. Finally, BSTIM was applied to assess flood resilience in 41 cities within the Yangtze River Delta region, and targeted policy recommendations were proposed. Results: The results showed that the space-time interpretation degree of BSTIM for urban flood resilience exceeded 93%, demonstrating excellent performance in fitting space-time data and enabling an accurate assessment of the space-time characteristics. (1) Flood resilience in the Yangtze River Delta region showed an upward trend during the study period, with the highest annual growth rate of 17.89% in 2017 and the lowest of 14.71% in 2009. (2) Significant differences existed in the spatial distribution of flood resilience across the region, forming an overall "low-high-medium" spatial pattern. (3) Local change trends varied considerably, presenting a characteristic pattern of "high in the east, low in the west, and medium in the north." (4) The spatial pattern of urban flood resilience exhibited significant differences in cold-hot spot clustering, with local changes forming a "cold in the west, hot in the east" pattern. Over half of the areas in the Yangtze River Delta region displayed "hot-hot" and "cold-hot" agglomeration characteristics, with "hot-hot" being the most prevalent. Conclusion: BSTIM provided an accurate assessment of space-time interaction effects in the space-time evolution of flood resilience. In addition, the model identified spatial distribution patterns, local change trends, cold-hot agglomeration distribution, and future development trends in flood resilience. Based on the results, the study captured the development dynamics of flood resilience in various cities of the Yangtze River Delta region, highlighting key priorities for flood prevention and mitigation. This research provides important reference value and theoretical support for effectively identifying critical areas for urban flood resilience development, thereby enhancing the flood resistance and disaster relief capabilities of urban agglomerations.