A deep learning technique-based data-driven model for accurate and rapid flood prediction

Abstract

Abstract. An accurate and rapid urban flood prediction model is essential to support decision-making on flood management, especially under increasing extreme precipitation conditions driven by climate change and urbanization. This study developed a deep learning technique-based data-driven flood prediction model based on an integration of LSTM network and Bayesian optimization. A case study in north China was applied to test the model performance and the results clearly showed that the model can accurately predict flood maps for various hyetograph inputs, meanwhile with substantial improvements in computation time. The model predicted flood maps 19,585 times faster than the physical-based hydrodynamic model and achieved a mean relative error of 9.5 %. For retrieving the spatial patterns of water depths, the degree of similarity of the flood maps was very high. In a best case, the difference between the ground truth and model prediction was only 0.76 % and the spatial distributions of inundated paths and areas were almost identical. The proposed model showed a robust generalizability and high computational efficiency, and can potentially replace and/or complement the conventional hydrodynamic model for urban flood assessment and management, particularly in applications of real time control, optimization and emergency design and plan.

Using geomorphologic indicators in preparation for flood zoning and flood risk maps in the Kashafroud basin, Iran

“…On the other hand, with the development of computer science, many software and numerical models have been presented for flood modeling. Among software models, HEC‐RAS (Ghimire et al, 2022; Namara et al, 2022; Sejati et al, 2022), HEC‐HMS (Altaf & Romshoo, 2022; Nadeem et al, 2022; Shah & Lone, 2022), MIKEs (Mike11, Mike21, Mike Flood, Mike Urban) (Dikici et al, 2022; Goumas et al, 2022; Li et al, 2022; Zhang et al, 2022; Zhou, Teng, et al, 2022; Zhou, Zheng, et al, 2022), Bentley Open Flow (de Andrade et al, 2022; Ellwood et al, 2022; Khakhar et al, 2022), Flo‐2D (Gerundo et al, 2022; Komolafe, 2022; Liu et al, 2022; Wang, Luo, et al, 2022), and CCHE (Lee et al, 2020; Kakati et al, 2022; Poorzaman et al, 2022) models can be mentioned. Also, the most applied numerical models include artificial neural network (Ahmad et al, 2022; Dahri et al, 2022; Wang, Yang, et al, 2022), random forest (Abedi et al, 2022; El‐Magd & Ahmed, 2022; Ha & Kang, 2022; Zhu & Zhang, 2022), and analytic hierarchy process (Bouamrane et al, 2022; Roy et al, 2023; Souissi et al, 2022; Vilasan & Kapse, 2022).…”

A hybrid deep neural network with statistical learning algorithms for flood susceptibility modeling

“…Various studies have confirmed that such hybridization can be applied to detect flood susceptibility regions where the technologies applying artificial intelligence in combination with spatial analysis seek to classify and process the geospatial data in a way that goes beyond simple input/output protocols. One of the main factors in determining the flood susceptibility areas is a calculation of the weight of flood predictors, which can be conducted based on the use of past flood locations or expert judgment (Zhou et al 2022). The efficiency of statistical learning approaches coupled with DNN has been evaluated by many different researches focusing on weighting the predictors (Shahab et al 2020).…”