Abstract:Uncertainties in flood forecasts that result from error propagation in meteorological–hydrological simulation processes present spatial and temporal dependences that are the principal sources of risk to real-time flood control in a multireservoir system. Risk analysis models that neglect such dependences potentially yield biased results regarding flood control operations. To enhance the accuracy of flood risk analysis and inform decision making regarding real-time flood control for a multireservoir system, this study proposed an integrated flood risk identification, analysis, and diagnosis model framework that incorporates the multidimensional dependences of forecast uncertainties. The framework includes an uncertainty identification model that uses the copula function to characterize the joint probability function of the mixed dependent uncertainties. It also incorporates a real-time flood risk analysis module that couples simulation and optimization to convert the complex multiobjective operation under uncertainty into a tractable model capable of flexibly evaluating individual and joint flood risks. The third component is a risk diagnosis module that calculates risk contributions using conditional probability, and identifies stable and vulnerable reservoirs within the system to alter operational decisions toward risk reduction. Methodologies were verified through application to a mixed four-reservoir system in the Pi River Basin, China. The principal findings were as follows. (1) Positive-dominated dependences of forecast uncertainties were identified because of error propagation through the meteorological and rainfall–runoff forecasts within the spatial and temporal continuity connection, and the copula function accurately described and quantified the multidimensional temporal and spatial dependences of the uncertainties. (2) In neglecting the spatial dependences of forecast uncertainties, the joint flood risk of the reservoir system was underestimated; in ignoring high-order temporal dependences, individual flood risk was underestimated owing to biased estimation in error variance. (3) Systematic reorganization of the operational strategies of diagnosed stable and vulnerable reservoirs could provide guidance for reduction and homogenization of flood risks. The proposed framework could be used to improve the accuracy of flood risk analysis and support reliable real-time flood control operation of a multireservoir system.

Flood forecasting and error simulation using copula entropy method

Copula entropy coupled with artificial neural network for rainfall–runoff simulation

Copula Entropy coupled with Wavelet Neural Network Model for Hydrological Prediction

Efficient joint probability analysis of flood risk

Modeling flood susceptibility using data-driven approaches of naive Bayes tree, alternating decision tree, and random forest methods

Application of a novel artificial neural network model in flood forecasting

Artificial neural network based hybrid modeling approach for flood inundation modeling

Application of Artificial Neural Network in the typhoon flood prediction system - A case study in Shanghai, China

An ANN-based Emulation Modelling Framework for Flood Inundation Modelling: Application, Challenges and Future Directions.

Probabilistic and ensemble simulation approaches for input uncertainty quantification of artificial neural network hydrological models

Hybrid artificial neural network and cooperation search algorithm for nonlinear river flow time series forecasting in humid and semi-humid regions

Exploration of dual-attention mechanism-based deep learning for multi-step-ahead flood probabilistic forecasting

Real-time error correction for flood forecasting based on machine learning ensemble method and its uncertainty assessment

Performance evaluation of artificial intelligence paradigms—artificial neural networks, fuzzy logic, and adaptive neuro-fuzzy inference system for flood prediction

Flood Forecasting Using Artificial Neural Network

Application of neural network and adaptive neuro-fuzzy inference systems for river flow prediction

Integrated real-time flood risk identification, analysis, and diagnosis model framework for a multireservoir system considering temporally and spatially dependent forecast uncertainties

A Flood Forecasting Model Based on Deep Learning Algorithm Via Integrating Stacked Autoencoders with BP Neural Network

Flood frequency analysis using generalized distributions and entropy-based model selection method

Flood prediction based on feature selection and a hybrid deep learning network

Generalizability of deep-learning-based emulation of hydrodynamic flood models