Flash floods occur frequently and distribute widely in mountainous areas because of complex geographic and geomorphic conditions and various climate types. Effective flash flood forecasting with useful lead times remains a challenge due to its high burstiness and short response time. Recently, machine learning has led to substantial changes across many areas of study. In hydrology, the advent of novel machine learning methods has started to encourage novel applications or substantially improve old ones. This study aims to establish a discharge forecasting model based on Long Short-Term Memory (LSTM) networks for flash flood forecasting in mountainous catchments. The proposed LSTM flood forecasting (LSTM-FF) model is composed of T multivariate single-step LSTM networks and takes spatial and temporal dynamics information of observed and forecast rainfall and early discharge as inputs. The case study in Anhe revealed that the proposed models can effectively predict flash floods, especially the qualified rates (the ratio of the number of qualified events to the total number of flood events) of large flood events are above 94.7% at 1–5 h lead time and range from 84.2% to 89.5% at 6–10 h lead-time. For the large flood simulation, the small flood events can help the LSTM-FF model to explore a better rainfall-runoff relationship. The impact analysis of weights in the LSTM network structures shows that the discharge input plays a more obvious role in the 1-h LSTM network and the effect decreases with the lead-time. Meanwhile, in the adjacent lead-time, the LSTM networks explored a similar relationship between input and output. The study provides a new approach for flash flood forecasting and the highly accurate forecast contributes to prepare for and mitigate disasters.
As a revolutionary tool leading to substantial changes across many areas, Machine Learning (ML) techniques have obtained growing attention in the field of hydrology due to their potentials to forecast time series. Moreover, a subfield of ML, Deep Learning (DL) is more concerned with datasets, algorithms and layered structures. Despite numerous applications of novel ML/DL techniques in discharge simulation, the uncertainty involved in ML/DL modeling has not drawn much attention, although it is an important issue. In this study, a framework is proposed to quantify uncertainty contributions of the sample set, ML approach, ML architecture and their interactions to multi-step time-series forecasting based on the analysis of variance (ANOVA) theory. Then a discharge simulation, using Recurrent Neural Networks (RNNs), is taken as an example. Long Short-Term Memory (LSTM) network, a state-of-the-art DL approach, was selected due to its outstanding performance in time-series forecasting, and compared with simple RNN. Besides, novel discharge forecasting architecture is designed by combining the expertise of hydrology and stacked DL structure, and compared with conventional design. Taking hourly discharge simulations of Anhe (China) catchment as a case study, we constructed five sample sets, chose two RNN approaches and designed two ML architectures. The results indicate that none of the investigated uncertainty sources are negligible and the influence of uncertainty sources varies with lead-times and discharges. LSTM demonstrates its superiority in discharge simulations, and the ML architecture is as important as the ML approach. In addition, some of the uncertainty is attributable to interactions rather than individual modeling components. The proposed framework can both reveal uncertainty quantification in ML/DL modeling and provide references for ML approach evaluation and architecture design in discharge simulations. It indicates uncertainty quantification is an indispensable task for a successful application of ML/DL.
15 December 2016 in Science China Technological Sciences
It is important to place water quality sensors along open channels in long-distance water transfer projects optimally for rapid source identification and efficient management of sudden water contamination. A new framework which considers multiple objectives, including earliest detection time, lowest missing detection rate and lowest sensor cost, and combines the randomness of injected contaminant type and contaminant incident consisting of contaminant intrusion location, time and mass, was established to obtain optimal placement of water quality sensor with better robustness in this paper. The middle route of the South-to-North Water Diversion Project in China was chosen as a case study, and it was found that both missing detection rate and detection time decrease with sensor cost gradually; furthermore, given the higher detecting precision, the detection accuracy and efficiency would be improved, a smaller number of water quality sensors would be needed, and the ten key placement positions where sensor with different detecting precision placed could be identified. Under the constraints of the allowable maximum missing detection rate, 1.00%, and detection time, 120.00 min, the detecting precision of 0.20 mg/L and 38 sensors placed could be selected as the optimal sensor placement scheme. Finally, with the consideration of contaminant uncertainty, the sensor placement scheme with better robustness could be constructed. The proposed framework would be helpful in solving the problem of water quality sensor placement with high practicality and efficiency in long-distance water transfer projects.
In drought years, most residents fail to improve water use efficiency due to residential water supply normally being prioritized in many regions, which makes other low-priority industrial water users suffer more from water shortage. This paper proposes a Pricing Strategy for Residential Water (PSRW), a water tariff that changes on annual time scale, based on the scarcity value of water resources, aiming to promote residential water conservation and reallocate water resources across the residential and industrial sectors during droughts. An optimization model to maximize the total benefit of residents and industrial sectors is introduced based on marginal benefit and price elasticity. The water shortage of industrial sectors is used to reflect the scarcity of water resources, and the lowest water supply standard for households and the maximum proportion of household water fee expenditure (HWFE) to household disposable income (HDI) are used to ensure the residents’ acceptability to price raising. It shows an “S-type” relationship between the optimal price raising coefficient and industrial water shortage, and two turning points are found in the curve, which are the starting and stopping points of price raising. The appearance of starting point depends on the non-negative net benefit, and the stopping point is affected by the factors that represent the residents’ acceptability to price raising. The application to Tianjin, a city in northern China with the rapid growth of population and economy but scarce water resources, shows PSRW is a potential means to improve water efficiency and optimize water resource allocation in water scarcity situations.