Real-time multi-step-ahead water level forecasting by recurrent neural networks for urban flood control
Introduction
Urban flood control is a crucial and challenging task, particularly in developed cities. Urban floods are flashy in nature mainly due to severe thunderstorms and occur both on urbanized surfaces and in small urban creeks, which deliver mass water to cities. On account of more impervious areas resulting from the rapid urbanization in metropolitan areas, less water infiltration has resulted in an increase in the flow rate and the amount of surface runoff over the last decades. Taiwan is located in the northwestern Pacific Ocean where subtropical air currents frequently introduce typhoons and convective rains. The urban flood hydrographs in Taiwan typically have large peak flows and fast-rising limbs in a matter of minutes, which could cause serious disasters. For example, Typhoon Nari brought massive rainfalls at an astounding level of 500 mm/day on September 17th in 2001, which resulted in 27 deaths, inundations at some stations of the Taipei Metro System, and countless economic losses. The heavy rainfall event on June 12th in 2012 brought astonishing rainfalls with a cumulative amount of 54.1 mm/hr, which directly resulted in quick and wide surface flooding such that the transportation system collapsed in most of the southern Taipei City. It appears floods cannot be prevented, but planning emergency measures through flood management might mitigate disastrous consequences.
In response to the flood threat to residents and property, the Taipei City Government has long-term endeavored in developing flood control-related infrastructures, such as increasing levee heights and enhancing sewerage systems, and urban inundations have been significantly mitigated and controlled in recent years. As a result, the main threat to the city turns out to be the floodwater inside the levee system. A surface inundation will inevitably take place if surface runoff exceeds the capacity of a storm drainage system. To tackle this problem, pumping stations play an important role in flood mitigation at metropolitan areas and are principal hydraulic facilities built to manage internal stormwater flows at places under the condition that gravity drainage cannot be achieved. The operation of a pumping station highly depends on the water level information of its floodwater storage pond (FSP). Within the catchment of a pumping station, surface runoff will drain to its FSP for storage and subsequent disposal through gravity drainage. When the water level of the FSP reaches the start level of duty pumps, the pumps will be activated according to operation rules for discharging the stored floodwater into the nearby river. For floodwater control management during heavy rainfall or typhoon events, it is imperative to construct an efficient and accurate model to forecast many step-ahead FSP water levels by utilizing the information of the current FSP water level and the rainfall measured at the neighboring rainfall gauging stations of the pumping station. The proposed model is expected to provide sufficient response time for warming up the pumps in advance for enhancing secure pumping operations and urban flood control management.
Artificial neural networks (ANNs) possess the ability to approximate nonlinear functions, and therefore become useful tools for handling water resources problems such as rainfall forecasting (Hung et al., 2009, Nasseri et al., 2008), stream flow forecasting (Akhtar et al., 2009, Besaw et al., 2010, Chen et al., 2013, Nayak et al., 2013, Sudheer et al., 2008, Toth, 2009, Sahoo et al., 2009), water level forecasting (Alvisi et al., 2006; Makarynskyy et al., 2004, Ali Ghorbani et al., 2010), and applications in urban drainage systems (Bruen and Yang, 2006, Loke et al., 1997, Chang et al., 2008, Chiang et al., 2010). Signal delays play an important role in neurobiological information processing. This concept has led to the development of dynamic neural networks. Recurrent neural networks (RNNs) that facilitate time delay units through feedback connections are computationally more powerful and biologically more plausible than other adaptive approaches such as feedforward networks, and thus have attracted much attention for years (Assaad et al., 2005, Chang et al., 2012, Coulibaly and Baldwin, 2005, Coulibaly and Evora, 2007, Ma et al., 2008, Muluye, 2011, Serpen and Xu, 2003). RNNs can be trained to learn sequential or time-varying patterns and are considered very effective in modeling the dynamics of complex hydrological processes with accurate forecasts; consequently their capability in modeling multi-step-ahead forecasts in highly variable time series is investigated. As known, multi-step-ahead forecasting is much more complex to deal with than one-step-ahead forecasting (Sorjamaa et al., 2007), and we believe ANNs, especially recurrent ones, can play an important role in tackling these complex tasks.
The greatest success in flood forecasting is commonly achieved on large rivers. Nevertheless, flash urban floods associated with heavy thunderstorms in cities are often very uncertain and are more difficult to predict due to complex dynamic phenomena involved. Many studies demonstrated the predictability of streamflow through soft computation methods (Nayak et al., 2004, Maity and Kumar, 2008) while only few papers investigated the prediction performance of inundation and/or sewerage systems in urban areas (Chiang et al., 2010). In this study, we intend to investigate the reliability and accuracy of short-term (10–60-min) forecast models for the floodwater storage pond (FSP) of a sewer-pumping system in Taipei City, Taiwan. The multi-step-ahead FSP water level forecast models for flood pumping control during heavy rainfall and/or typhoon events are tailored made through a static ANN (the back-propagation neural network-BPNN) and two dynamic ANNs (the Elman NN; the nonlinear autoregressive with eXogenous input-NARX network). Consequently, the comparison results of these three ANN models are evaluated for the effectiveness of recurrent connections. The forecasting system is designed to anticipate the occurrence of flooding and to take measures necessary to reduce flood-induced losses. The study will give a boost to the efforts for urban flood disaster management and strengthen the Taipei City Government with more proactive disaster preparedness.
Section snippets
Methodology
In this study, various ANNs are used to make water level forecasts for representing the behavior of the rainfall-sewer flow processes in storm events. Flood levels can be forecasted on the basis of (a) rainfall data; (b) previous water levels; and (c) a combination of both data sets. We adopt three ANNs coupled with statistical techniques to construct real time multi-step-ahead FSP forecast models. The implementation procedure is shown in Fig. 1. The time span of rainfall affecting the rise of
Study area and dataset
Taiwan, an island located in the subtropical zone of the North Pacific Ocean, is covered with mountainous terrains and steep landforms. Taipei City, situated in the Taipei Basin of northern Taiwan, is surrounded by the Danshui River whose narrow estuary makes it difficult to discharge water effectively from the city. Consequently, the high levees along the Danshui River have been built to prevent outer flood into the city with a return period of two-hundred-year flood protection standard.
Results and discussion
This section presents the selection result of effective rainfall factors and the forecast performance of the static (BPNN) and dynamic (Elman NN and NARX) neural networks in two scenarios (w/ and w/o current FSP water level information). The results and discussion are addressed in details, which are shown as follows.
Conclusions
In this study, three ANN models (one static, two dynamic) are developed to make forecasts on the evolution of water level at floodwater storage pond (FSP) as a function of current FSP water level and rainfall information based on the inputs extracted by an advanced factor selection method (GT) for allowing sufficient time advance to warm up the pumping system and enhancing secure pumping operations to prevent the city from flooding. The temporal resolution of water level and rainfall data is 10
Acknowledgements
The authors gratefully acknowledge the funding support from the Hydraulic Engineering Office, Public Works Department, Taipei City Government, Taiwan, ROC, (Grant No.: H-102-03-102124) for this research. The authors sincerely appreciate the Editor, Associate Editor and Reviewers for their valuable comments and constructive suggestions.
References (43)
- et al.
Sea water level forecasting using genetic programming and comparing the performance with artificial neural networks
Comput. Geosci.
(2010) - et al.
Advances in ungauged streamflow prediction using artificial neural networks
J. Hydrol.
(2010) - et al.
Counterpropagation fuzzy-neural network for city flood control system
J. Hydrol.
(2008) - et al.
Regional estimation of groundwater arsenic concentrations through systematical dynamic-neural modeling
J. Hydrol.
(2013) - et al.
Reinforced recurrent neural networks for multi-step-ahead flood forecasts
J. Hydrol.
(2013) - et al.
Nonstationary hydrological timeseries forecasting using nonlinear dynamic methods
J. Hydrol.
(2005) - et al.
Comparison of neural network methods for infilling missing daily weather records
J. Hydrol.
(2007) Finding structure in time
Cognit. Sci.
(1990)- et al.
How embedded memory in recurrent neural network architectures helps learning long-term temporal dependencies
Neural Networks
(1998) - et al.
Artificial neural networks as a tool in urban storm drainage
Water Sci. Technol.
(1997)
Predicting sea level variations with artificial neural networks at Hillarys Boat Harbour, Western Australia
Estuarine Coastal Shelf Sci.
Long-term time series prediction with the NARX network: an empirical evaluation
Neurocomputing
Evaporation estimation using artificial neural networks and adaptive neuro-fuzzy inference system techniques
Adv. Water Resour.
River flow forecasting through conceptual models: 1. A discussion of principles
J. Hydrol.
Optimized scenario for rainfall forecasting using genetic algorithm coupled with artificial neural network
Expert Syst. Appl.
A neuro-fuzzy computing technique for modeling hydrological time series
J. Hydrol.
Rainfall runoff modelling using conceptual, data driven and wavelet based computing approach
J. Hydrol.
Assessment of input variables determination on the SVM model performance using PCA, Gamma test, and forward selection techniques for monthly stream flow prediction
J. Hydrol.
Methodology for long-term prediction of time series
Neurocomputing
A note on the gamma test
Neuralcomput. Appl.
River flow forecasting with artificial neural networks using satellite observed precipitation pre-processed with flow length and travel time information: case study of the Ganges river basin
Hydrol. Earth Syst. Sci.
Cited by (200)
Analysing the performance of the NARX model for forecasting the water level in the Chikugo River estuary, Japan
2024, Environmental ResearchComparison of strategies for multistep-ahead lake water level forecasting using deep learning models
2024, Journal of Cleaner ProductionData driven real-time prediction of urban floods with spatial and temporal distribution
2024, Journal of Hydrology XEvent-based decision support algorithm for real-time flood forecasting in urban drainage systems using machine learning modelling
2023, Environmental Modelling and SoftwareHigh temporal resolution urban flood prediction using attention-based LSTM models
2023, Journal of Hydrology