Urban Pluvial Flood Modeling Integrating Two-Dimensional CNN and SHAP-Driven Interpretability

Abstract

Urban pluvial floods have become increasingly frequent under the combined effects of climate change and urbanization, leading to increased disaster losses. This study developed a data-driven urban pluvial flood prediction model using convolutional neural networks (CNNs) to enhance computational efficiency while maintaining simulation accuracy. A high-resolution cellular-based flood model generated the training dataset through systematic patch-based sampling combined with fixed step size selection strategies. The established framework enabled flood simulations through integrated analysis of topographic features and rainfall processes. Shapley additive explanations (SHAP) and Group masking analysis (GMA) were implemented to interpret the decision-making mechanisms of CNN model. The model was validated in a relatively independent drainage area, demonstrating strong agreement with conventional cellular model outputs across six design storm scenarios and two historical rainfall events. Computational experiments showed that the CNN model reduced simulation time from minutes to seconds compared to process-based approaches, while maintaining low absolute errors in water depth predictions. Both SHAP and GMA interpretation revealed that topographic features, particularly building, digital elevation model (DEM), and aspect, exert dominant influence on model predictions. This data-driven framework established an efficient computational paradigm for urban flood modeling, with SHAP and GMA analysis guiding input variable selection while explaining model behavior. The methodology demonstrated potential for real-time monitoring integration, supporting rapid flood risk assessment and resilience enhancement.

Keywords

Convolutional neural network / Data-driven model / Numerical simulation / SHAP and GMA analysis / Urban pluvial flood

References

[1] Bentivoglio R, Isufi E, Jonkman SN, Taormina R. Deep learning methods for flood mapping: A review of existing applications and future research directions. Hydrology and Earth System Sciences, 2022, 26(16): 4345-4378

[2] Bhattarai, Y., S. Bista, R. Talchabhadel, S. Duwal, and S. Sharma. 2024. Rapid prediction of urban flooding at street-scale using physics-informed machine learning-based surrogate modeling. Total Environment Advances 12: Article 200116.

[3] Chen, W.J., W.Q. Wang, G.R. Huang, Z.L. Wang, C.G. Lai, and Z.Y. Yang. 2021. The capacity of grey infrastructure in urban flood management: A comprehensive analysis of grey infrastructure and the green-grey approach. International Journal of Disaster Risk Reduction 54: Article 102045.

[4] Ferreira MD, Correa DC, Nonato LG, de Mello RF. Designing architectures of convolutional neural networks to solve practical problems. Expert Systems with Applications, 2018, 94: 205-217

[5] Gao, W.Z., Y.X. Liao, Y.H. Chen, C.G. Lai, S.J. He, and Z.L. Wang. 2024. Enhancing transparency in data-driven urban pluvial flood prediction using an explainable CNN model. Journal of Hydrology 645(Part A): Article 132228.

[6] Glenis V, Kutija V, Kilsby CG. A fully hydrodynamic urban flood modelling system representing buildings, green space and interventions. Environmental Modelling & Software, 2018, 109: 272-292

[7] Guo KH, Guan MF, Yu DP. Urban surface water flood modelling – A comprehensive review of current models and future challenges. Hydrology and Earth System Sciences, 2021, 25(5): 2843-2860

[8] Guo ZF, Leitao JP, Simoes NE, Moosavi V. Data-driven flood emulation: Speeding up urban flood predictions by deep convolutional neural networks. Journal of Flood Risk Management, 2021, 14(1): 2843-2860

[9] Haces-Garcia, F., N. Ross, C.L. Glennie, H.S. Rifai, V. Hoskere, and N. Ekhtari. 2025. Rapid 2D hydrodynamic flood modeling using deep learning surrogates. Journal of Hydrology 651: Article 132561.

[10] He, J., L. Zhang, T. Xiao, H. Wang, and H. Luo. 2023. Deep learning enables super-resolution hydrodynamic flooding process modeling under spatiotemporally varying rainstorms. Water Research 239: Article 120057.

[11] Jamali, B., E. Haghighat, A. Ignjatovic, J.P. Leitao, and A. Deletic. 2021. Machine learning for accelerating 2D flood models: Potential and challenges. Hydrological Processes 35(4): Article e14064.

[12] Kabir, S., S. Patidar, X.L. Xia, Q.H. Liang, J. Neal, and G. Pender. 2020. A deep convolutional neural network model for rapid prediction of fluvial flood inundation. Journal of Hydrology 590: Article 125481.

[13] Ke, E.T., J.C. Zhao, and Y.L. Zhao. 2025. Investigating the influence of nonlinear spatial heterogeneity in urban flooding factors using geographic explainable artificial intelligence. Journal of Hydrology 648: Article 132398.

[14] Leandro J, Chen AS, Djordjevic S, Savic DA. Comparison of 1D/1D and 1D/2D coupled (sewer/surface) hydraulic models for urban flood simulation. Journal of Hydraulic Engineering, 2009, 135(6): 495-504

[15] LeCun Y, Boser B, Denker JS, Henderson D, Howard RE, Hubbard W, Jackel LD. Backpropagation applied to handwritten zip code recognition. Neural Computation, 1989, 1(4): 541-551

[16] Liang, Y.S., W.H. Liao, Z. Zhang, H.C. Li, and H. Wang. 2024. Using a multiphysics coupling-oriented flood modelling approach to assess urban flooding under various regulation scenarios combined with rainstorms and tidal effects. Journal of Hydrology 645(Part A): Article 132189.

[17] Liang QH, Xia XL, Hou JM. Efficient urban flood simulation using a GPU-accelerated SPH model. Environmental Earth Sciences, 2015, 74(11): 7285-7294

[18] Liao YX, Wang ZL, Lai CG, Xu CY. A framework on fast mapping of urban flood based on a multi-objective random forest model. International Journal of Disaster Risk Science, 2023, 14(2): 253-268

[19] Liu, W., X. Zhang, Q. Feng, and B.A. Engel. 2025. City-scale integrated flood risk prediction under future climate change and urbanization based on the shared socioeconomic pathways (SSP) scenarios. Journal of Hydrology 655: Article 132971.

[20] Löwe, R., J. Böhm, D.G. Jensen, J. Leandro, and S.H. Rasmussen. 2021 U-FLOOD – Topographic deep learning for predicting urban pluvial flood water depth. Journal of Hydrology 603(Part A): Article 126898.

[21] Luo, P.P., M.T. Luo, F.Y. Li, X.G. Qi, A.D. Huo, Z.H. Wang, B. He, K. Takara, et al. 2022. Urban flood numerical simulation: Research, methods and future perspectives. Environmental Modelling & Software 156: Article 105478.

[22] Madayala, A.B., A. Jain, and B. Lohani. 2022. Development of a physics-guided neural network model for effective urban flood management. Journal of Hydrologic Engineering 27(9). https://doi.org/10.1061/(ASCE)HE.1943-5584.0002196.

[23] Mudashiru, R.B., N. Sabtu, I. Abustan, and W. Balogun. 2021. Flood hazard mapping methods: A review. Journal of Hydrology 603(Part A): Article 126846.

[24] Piadeh, F., K. Behzadian, and A.M. Alani. 2022. A critical review of real-time modelling of flood forecasting in urban drainage systems. Journal of Hydrology 607: Article 127476.

[25] Ren, H.C., B. Pang, G. Zhao, H.J. Yu, P.N. Tian, and C.R. Xie. 2025. Incorporating dynamic drainage supervision into deep learning for accurate real-time flood simulation in urban areas. Water Research 270: Article 122816.

[26] Ro B, Garfin G. Participatory risk governance for Seoul, South Korea’s flood risk management. International Journal of Disaster Risk Science, 2024, 15(3): 317-331

[27] Song, W.K., M.F. Guan, K.H. Guo, and D.P. Yu. 2025. Rapid flood inundation mapping by integrating deep learning-based image super-resolution with coarse-grid hydrodynamic modeling. Engineering Applications of Computational Fluid Mechanics 19(1): Article: 2481115.

[28] Tellman B, Sullivan JA, Kuhn C, Kettner AJ, Doyle CS, Brakenridge GR, Erickson TA, Slayback DA. Satellite imaging reveals increased proportion of population exposed to floods. Nature, 2021, 596(7870): 80-86

[29] Wang WQ, Chen WJ, Huang GR. Urban stormwater modeling with local inertial approximation form of shallow water equations: A comparative study. International Journal of Disaster Risk Science, 2021, 12(5): 745-763

[30] Wang, J.H., J.M. Hou, S.W. Li, J.X. Sun, J. Jing, and J.Y. Wang. 2024. Advancements in enhancing flood evolution and urban inundation predictions: A study of local time stepping algorithm and GPU-accelerated hydrodynamic model. Journal of Hydrology 641: Article 131744.

[31] Xing Y, Liang QH, Wang G, Ming XD, Xia XL. City-scale hydrodynamic modelling of urban flash floods: The issues of scale and resolution. Natural Hazards, 2019, 96(1): 473-496

[32] Xu K, Han ZT, Xu HS, Bin LL. Rapid prediction model for urban floods based on a light gradient boosting machine approach and hydrological-hydraulic model. International Journal of Disaster Risk Science, 2023, 14(1): 79-97

[33] Yan, R., J.Q. Liao, J. Yang, W. Sun, M.Y. Nong, and F.P. Li. 2021. Multi-hour and multi-site air quality index forecasting in Beijing using CNN, LSTM, CNN-LSTM, and spatiotemporal clustering. Expert Systems with Applications 169: Article 114513.

[34] Yang JR, Liu K, Wang M, Zhao G, Wu W, Yue QR. A convolutional neural network-weighted cellular automaton model for the fast prediction of urban pluvial flooding processes. International Journal of Disaster Risk Science, 2024, 15(5): 754-768

[35] Zeng ZY, Lai CG, Wang ZL, Wu XS, Huang GR, Hu QF. Rapid simulation of urban rainstorm flood based on WCA2D and SWMM model. Advances in Water Science, 2020, 31(1): 29-38

[36] Zhang, Z.Y., W.C. Tian, C.K.X. Lu, Z.L. Liao, and Z.G. Yuan. 2024. Graph neural network-based surrogate modelling for real-time hydraulic prediction of urban drainage networks. Water Research 263: Article 122142.

[37] Zhao, W.G., Z.X. Zhang, N. Khodadadi, and L.Y. Wang. 2025. A deep learning model coupled with metaheuristic optimization for urban rainfall prediction. Journal of Hydrology 651: Article 132596.

[38] Zheng, J.X., X. Chen, K. Kawaike, K. Yamanoi, T. Koshiba, and G.R. Huang. 2024. Response of urban flood resilience to climate change: An exploration with a novel performance-based metric considering the socioeconomic impacts of damage costs. Journal of Hydrology 645(Part B): Article 132260.

RIGHTS & PERMISSIONS

The Author(s)