A ConvLSTM-Based Model for Urban Flood Prediction Under Dynamic Rainfall Patterns and Exploration on Its Extrapolation Capability

Abstract

Urban flooding induced by heavy rainfall is increasingly frequent, necessitating accurate and timely flood forecasting to mitigate risks. Although data-driven models have demonstrated significant potential for real-time flood prediction due to their computational efficiency, current implementations frequently neglect the critical influence of rainfall spatial heterogeneity, resulting in inaccuracies in flood prediction. Therefore, this study designed diverse rainfall scenarios featuring moving rainstorm centers and proposed a fast simulation method for urban flooding under complex rainfall conditions, utilizing the convolutional long short-term memory (ConvLSTM) model. The efficacy of the proposed method was validated across three study areas. The results indicate that the ConvLSTM model has superior performance in predicting flood inundation depth and extent, achieving an average R² of 0.964, outperforming two other deep learning models. Notably, this model achieved predictions within seconds based on input rainfall data, offering high computational efficiency that is hundreds of times faster than hydrological–hydrodynamic coupled models. Furthermore, we explored the model’s extrapolation capability when rainfall intensities exceed the maximum value of the training set. This research contributes insights to the advancement and refinement of rapid urban flood forecasting methodologies.

Keywords

ConvLSTM / Moving rainstorm / Rapid prediction / Urban flooding

References

[1] Ackom EK, Adjei KA, Odai SN. Spatio-temporal rainfall trend and homogeneity analysis in flood prone area: Case study of Odaw River basin—Ghana. SN Applied Sciences, 2020, 2: Article 2141.

[2] Barredo JI. Major flood disasters in Europe: 1950–2005. Natural Hazards, 2007, 42: 125-148.

[3] Barzegar R, Aalami MT, Adamowski J. Coupling a hybrid CNN-LSTM deep learning model with a boundary corrected maximal overlap discrete wavelet transform for multiscale lake water level forecasting. Journal of Hydrology, 2021, 598: Article 126196.

[4] Bates PD, De Roo APJ. A simple raster-based model for flood inundation simulation. Journal of Hydrology, 2000, 236: 54-77.

[5] Bates PD, Dawson RJ, Hall JW, Horritt MS, Nicholls RJ, Wicks J, Hassan Mohamed Ahmed Ali Mohamed. Simplified two-dimensional numerical modelling of coastal flooding and example applications. Coastal Engineering, 2005, 52: 793-810.

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

[7] Chang T, Yu H, Wang C, Chen AS. Overland-gully-sewer (2d–1d-1d) urban inundation modeling based on cellular automata framework. Journal of Hydrology, 2021, 603: Article 127001.

[8] Chen J, Li Y, Zhang C. The effect of design rainfall patterns on urban flooding based on the Chicago method. International Journal of Environmental Research and Public Health, 2023, 20: Article 4245.

[9] Chen G, Hou J, Liu Y, Li X, Qiao X, Li D. Study on the sensitivity of urban inundation and watershed flood simulation to rainfall data spatial resolution. Urban Climate, 2024, 57: Article 102125.

[10] Costabile P, Macchione F, Natale L, Petaccia G. Flood mapping using LIDAR DEM. Limitations of the 1-D modeling highlighted by the 2-D approach. Natural Hazards, 2015, 77: 181-204.

[11] Dang TQ, Tran BH, Le QN, Tanim AH, Bui VH, Mai ST, Thanh PN, Anh DT. Integrating intelligent hydro-informatics into an effective early warning system for risk-informed urban flood management. Environmental Modelling & Software, 2025, 183: Article 106246.

[12] Deng Z, Wang Z, Wu X, Lai C, Liu W. Effect difference of climate change and urbanization on extreme precipitation over the Guangdong-Hong Kong-Macao Greater Bay Area. Atmospheric Research, 2023, 282: Article 106514.

[13] Dottori F, Todini E. Testing a simple 2D hydraulic model in an urban flood experiment. Hydrological Processes, 2013, 27: 1301-1320.

[14] El Baida M, Boushaba F, Chourak M, Hosni M. Real-time urban flood depth mapping: Convolutional neural networks for pluvial and fluvial flood emulation. Water Resources Management, 2024, 38: 4763-4782.

[15] Fowler HJ, Lenderink G, Prein AF, Westra S, Allan RP, Ban N, Barbero R, Berg P, et al. . Anthropogenic intensification of short-duration rainfall extremes. Nature Reviews Earth & Environment, 2021, 2: 107-122.

[16] Fu G, Jin Y, Sun S, Yuan Z, Butler D. The role of deep learning in urban water management: A critical review. Water Research, 2022, 223: Article 118973.

[17] Gao W, Liao Y, Chen Y, Lai C, He S, Wang Z. Enhancing transparency in data-driven urban pluvial flood prediction using an explainable CNN model. Journal of Hydrology, 2024, 645: Article 132228.

[18] Guidolin M, Chen AS, Ghimire B, Keedwell EC, Djordjević S, Savić DA. A weighted cellular automata 2D inundation model for rapid flood analysis. Environmental Modelling & Software, 2016, 84: 378-394.

[19] Guo K, Guan M, Yu D. Urban surface water flood modelling—A comprehensive review of current models and future challenges. Hydrology and Earth System Sciences, 2021, 25: 2843-2860.

[20] Guo Z, Leitão JP, Simões NE, Moosavi V. Data-driven flood emulation: Speeding up urban flood predictions by deep convolutional neural networks. Journal of Flood Risk Management, 2021, 14: Article e12684.

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

[22] Hoshino T, Yamada TJ. Spatiotemporal classification of heavy rainfall patterns to characterize hydrographs in a high-resolution ensemble climate dataset. Journal of Hydrology, 2023, 617: Article 128910.

[23] Houngue NR, Almoradie AD, Thiam S, Komi K, Adounkpè JG, Begedou K, Evers M. Climate and land-use change impacts on flood hazards in the Mono River catchment of Benin and Togo. Sustainability, 2023, 15(7): Article 5862.

[24] Huang H, Wang Z, Liao Y, Gao W, Lai C, Wu X, Zeng Z. Improving the explainability of CNN-LSTM-based flood prediction with integrating SHAP technique. Ecological Informatics, 2024, 84: Article 102904.

[25] Jayapadma JMMU, Wickramaarachchi TN, Silva GHAC, Ishidaira H, Magome J. Coupled hydrodynamic modelling approach to assess land use change induced flood characteristics. Environmental Monitoring and Assessment, 2022, 194: Article 354.

[26] Ji S, Xu W, Yang M, Yu K. 3D convolutional neural networks for human action recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 35: 221-231.

[27] Ji S, Zhang C, Xu A, Shi Y, Duan Y. 3D convolutional neural networks for crop classification with multi-temporal remote sensing images. Remote Sensing, 2018, 10: Article 75.

[28] Lai C, Luo Y, Li X, Yu H, Zeng Z, Li S, Gao W, Wang Z. Assessment on vulnerability of road networks considering the dynamic impact of urban waterlogging and the mitigation effect of LID measures. Journal of Hydrology, 2024, 643: Article 132005.

[29] LeCun Y, Bengio Y, Hinton G. Deep learning. Nature, 2015, 521(7553): 436-444.

[30] Lehmann J, Coumou D, Frieler K. Increased record-breaking precipitation events under global warming. Climatic Change, 2015, 132: 501-515.

[31] Li S, Wang Z, Lai C, Lin G. Quantitative assessment of the relative impacts of climate change and human activity on flood susceptibility based on a cloud model. Journal of Hydrology, 2020, 588: Article 125051.

[32] Li S, Wang Z, Wu X, Zeng Z, Shen P, Lai C. A novel spatial optimization approach for the cost-effectiveness improvement of LID practices based on SWMM-FTC. Journal of Environmental Management, 2022, 307: Article 114574.

[33] Liao Y, Wang Z, Chen X, Lai C. Fast simulation and prediction of urban pluvial floods using a deep convolutional neural network model. Journal of Hydrology, 2023, 624: Article 129945.

[34] Liao Y, Wang Z, Lai C, Xu C. 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.

[35] Lin R, Zheng F, Ma Y, Duan H, Chu S, Deng Z. Impact of spatial variation and uncertainty of rainfall intensity on urban flooding assessment. Water Resources Management, 2022, 36: 5655-5673.

[36] Luo Z, Liu J, Zhang S, Shao W, Zhou J, Zhang L, Jia R. Spatiotemporal evolution of urban rain islands in China under the conditions of urbanization and climate change. Remote Sensing, 2022, 14: Article 4159.

[37] Ma X, Man Q, Yang X, Dong P, Yang Z, Wu J, Liu C. Urban feature extraction within a complex urban area with an improved 3D-CNN using airborne hyperspectral data. Remote Sensing, 2023, 15: Article 992.

[38] Maas, A.L., A.Y. Hannun, and A.Y. Ng. 2013. Rectifier nonlinearities improve neural network acoustic models, 2013. In Proceedings of the 30th International Conference on Machine Learning, 17–19 June 2013, Atlanta, Georgia, USA.

[39] Meng D, Liao Y, Deng Z, Chen Y, Lai C, Chen X, Wang Z. Spatially moving non-uniform rainstorms may exacerbate urban flooding disasters. Journal of Hydrology, 2025, 660: Article 133374.

[40] Mignot E, Li X, Dewals B. Experimental modelling of urban flooding: A review. Journal of Hydrology, 2019, 568: 334-342.

[41] Moishin M, Deo RC, Prasad R. Designing deep-based learning flood forecast model with ConvLSTM hybrid algorithm. IEEE Access, 2021, 9: 50982-50993.

[42] Morales-Hernández M, Sharif MB, Kalyanapu A, Ghafoor SK, Dullo TT, Gangrade S, Kao SC, Norman MR, et al. . TRITON: A multi-GPU open source 2D hydrodynamic flood model. Environmental Modelling & Software, 2021, 141: Article 105034.

[43] Muller CL, Chapman L, Johnston S, Kidd C, Illingworth S, Foody G, Overeem A, Leigh RR. Crowdsourcing for climate and atmospheric sciences: Current status and future potential. International Journal of Climatology, 2015, 35: 3185-3203.

[44] Neal J, Dunne T, Sampson C, Smith A, Bates P. Optimisation of the two-dimensional hydraulic model LISFOOD-FP for CPU architecture. Environmental Modelling & Software, 2018, 107: 148-157.

[45] Pan X, Hou J, Gao X, Chen G, Li D, Imran M, Li X, Yang N, et al. . LSTM model-based rapid prediction method of urban inundation with rainfall time series. Water Resources Management, 2025, 39: 661-688.

[46] Park K, Lee M. The development and application of the urban flood risk assessment model for reflecting upon urban planning elements. Water, 2019, 11(5): Article 920.

[47] Patro S, Chatterjee C, Mohanty S, Singh R, Raghuwanshi NS. Flood inundation modeling using MIKE FLOOD and remote sensing data. Journal of the Indian Society of Remote Sensing, 2009, 37: 107-118.

[48] Prestininzi P. Suitability of the diffusive model for dam break simulation: Application to a CADAM experiment. Journal of Hydrology, 2008, 361: 172-185.

[49] Qi W, Ma C, Xu H, Chen Z, Zhao K, Han H. A review on applications of urban flood models in flood mitigation strategies. Natural Hazards, 2021, 108: 31-62.

[50] Qiu Y, Schertzer D, Tisserand B, Tchiguirinskaia I. Spatio-temporal rainfall variability and its impacts on the hydrological response of nature-based solutions. Urban Water Journal, 2024, 21: 1147-1163.

[51] Rasool U, Yin X, Xu Z, Padulano R, Rasool MA, Siddique MA, Hassan MA, Senapathi V. Rainfall-driven machine learning models for accurate flood inundation mapping in Karachi, Pakistan. Urban Climate, 2023, 49: Article 101573.

[52] Sadeghi F, Rubinato M, Goerke M, Hart J. Assessing the performance of LISFLOOD-FP and SWMM for a small watershed with scarce data availability. Water, 2022, 14: Article 748.

[53] Shao Y, Chen J, Zhang T, Yu T, Chu S. Advancing rapid urban flood prediction: A spatiotemporal deep learning approach with uneven rainfall and attention mechanism. Journal of Hydroinformatics, 2024, 26: 1409-1424.

[54] Shen J, Tong Z, Zhu J, Liu X, Yan F. A new rapid simplified model for urban rainstorm inundation with low data requirements. Water, 2016, 8(11): Article 512.

[55] Shi X, Chen Z, Wang H, Yeung D, Wong W, Woo W. Convolutional LSTM network: A machine learning approach for precipitation nowcasting. Advances in Neural Information Processing Systems, 2015, 28: 802-810

[56] Singh J, Karmakar S, PaiMazumder D, Ghosh S, Niyogi D. Urbanization alters rainfall extremes over the contiguous United States. Environmental Research Letters, 2020, 15: Article 74033.

[57] Tarasova L, Basso S, Zink M, Merz R. Exploring controls on rainfall-runoff events: 1. Time series-based event separation and temporal dynamics of event runoff response in Germany. Water Resources Research, 2018, 54: 7711-7732.

[58] Thakur A, Gupta M, Sinha DK, Mishra KK, Venkatesan VK, Guluwadi S. Transformative breast cancer diagnosis using CNNs with optimized ReduceLROnPlateau and early stopping enhancements. International Journal of Computational Intelligence Systems, 2024, 17: Article 14.

[59] Thrysøe C, Balstrøm T, Borup M, Löwe R, Jamali B, Arnbjerg-Nielsen K. FloodStroem: A fast dynamic GIS-based urban flood and damage model. Journal of Hydrology, 2021, 600: Article 126521.

[60] Treinish LA, Praino AP, Cipriani JP, Mello UT, Mantripragada K, Real LV, Sesini PA, Saxena V, et al. . Enabling high-resolution forecasting of severe weather and flooding events in Rio de Janeiro. IBM Journal of Research and Development, 2013, 57: 1-7.

[61] Wang Z, Lai C, Chen X, Yang B, Zhao S, Bai X. Flood hazard risk assessment model based on random forest. Journal of Hydrology, 2015, 527: 1130-1141.

[62] Wang Y, Fang Z, Hong H, Peng L. Flood susceptibility mapping using convolutional neural network frameworks. Journal of Hydrology, 2020, 582: Article 124482.

[63] Wang Z, Chen Y, Zeng Z, Chen X, Li X, Jiang X, Lai C. A tight coupling model for urban flood simulation based on SWMM and TELEMAC-2D and the uncertainty analysis. Sustainable Cities and Society, 2024, 114: Article 105794.

[64] Wang X, Hou J, Gao X, Wang T, Zhou Q, Li Y, Sun X. Urban inundation response law analysis to characteristics of designed rainstorms based on coupled hydrodynamic and rainfall-tracking model. Journal of Hydrology, 2024, 632: Article 130870.

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

[66] Wang Z, Lyu H, Fu G, Zhang C. Time-guided convolutional neural networks for spatiotemporal urban flood modelling. Journal of Hydrology, 2024, 645: Article 132250.

[67] Wei H, Wu H, Zhang L, Liu J. Urban flooding simulation and flood risk assessment based on the InfoWorks ICM model: A case study of the urban inland rivers in Zhengzhou, China. Water Science and Technology, 2024, 90: 1338-1358.

[68] Wu X, Wang Z, Guo S, Liao W, Zeng Z, Chen X. Scenario-based projections of future urban inundation within a coupled hydrodynamic model framework: A case study in Dongguan City, China. Journal of Hydrology, 2017, 547: 428-442.

[69] Wu G, Chen W, Jung H. Gated attention recurrent neural network: A deeping learning approach for radar-based precipitation nowcasting. Water, 2022, 14: Article 2570.

[70] Wu M, Wei X, Ge W, Chen G, Zheng D, Zhao Y, Chen M, Xin Y. Analyzing the spatial scale effects of urban elements on urban flooding based on multiscale geographically weighted regression. Journal of Hydrology, 2024, 645: Article 132178.

[71] Xiang X, Guo S, Li C, Sun B, Liang Z. Deep learning model for flood probabilistic forecasting considering spatiotemporal rainfall distribution and hydrologic uncertainty. Journal of Hydrology, 2025, 655: Article 132879.

[72] Xiong L, Ding W, Huang X, Huang W. CLSTAN: ConvLSTM-based spatiotemporal attention network for traffic flow forecasting. Mathematical Problems in Engineering, 2022, 2022: Article 1604727.

[73] Xu Y, He C, Guo Z, Chen Y, Sun Y, Dong Y. Simulation of water level and flow of catastrophic flood based on the CNN-LSTM coupling network. Water, 2023, 15: Article 2329.

[74] Yan L, Rong H, Yang W, Lin J, Zheng C. A novel integrated urban flood risk assessment approach based on one-two dimensional coupled hydrodynamic model and improved projection pursuit method. Journal of Environmental Management, 2024, 366: Article 121910.

[75] Yang S, Jhong B, Jhong Y, Tsai T, Chen C. Long short-term memory integrating moving average method for flood inundation depth forecasting based on observed data in urban area. Natural Hazards, 2023, 116: 2339-2361

[76] Yu R, Zhai P, Chen Y. Facing climate change-related extreme events in megacities of China in the context of 1.5°C global warming. Current Opinion in Environmental Sustainability, 2018, 30: 75-81.

[77] Zeng Z, Lai C, Wang Z, Chen Y, Chen X. Future sea level rise exacerbates compound floods induced by rainstorm and storm tide during super typhoon events: A case study from Zhuhai. China. Science of the Total Environment, 2024, 911: Article 168799.

[78] Zhang W, Zhang X, Liu Y, Tang W, Xu J, Fu Z. Assessment of flood inundation by coupled 1D/2D hydrodynamic modeling: A case study in mountainous watersheds along the coast of southeast China. Water, 2020, 12(3): Article 822.

[79] Zhang M, Xu M, Wang Z, Lai C. Assessment of the vulnerability of road networks to urban waterlogging based on a coupled hydrodynamic model. Journal of Hydrology, 2021, 603: Article 127105.

[80] Zhang W, Liu Y, Tang W, Chen S, Xie W. Rapid spatio-temporal prediction of coastal urban floods based on deep learning approaches. Urban Climate, 2023, 52: Article 101716.

[81] Zhang X, Kang A, Lei X, Wang H. Urban drainage efficiency evaluation and flood simulation using integrated SWMM and terrain structural analysis. Science of the Total Environment, 2024, 957: Article 177442.

[82] Zhang R, Li Y, Chen T, Zhou L. Flood risk identification in high-density urban areas of Macau based on disaster scenario simulation. International Journal of Disaster Risk Reduction, 2024, 107: Article 104485.

[83] Zhou Q, Leng G, Su J, Ren Y. Comparison of urbanization and climate change impacts on urban flood volumes: Importance of urban planning and drainage adaptation. Science of the Total Environment, 2019, 658: 24-33.

[84] Zhou Z, Smith JA, Baeck ML, Wright DB, Smith BK, Liu S. The impact of the spatiotemporal structure of rainfall on flood frequency over a small urban watershed: An approach coupling stochastic storm transposition and hydrologic modeling. Hydrology and Earth System Sciences, 2021, 25: 4701-4717.

[85] Zhou Y, Wu Z, Jiang M, Xu H, Yan D, Wang H, He C, Zhang X. Real-time prediction and ponding process early warning method at urban flood points based on different deep learning methods. Journal of Flood Risk Management, 2024, 17: Article e12964.

RIGHTS & PERMISSIONS

The Author(s)