Long Term Streamflow Forecasting Using a Hybrid Entropy Model

Dariane, Alireza B.; Farhani, M.; Azimi, Sh.

doi:10.1007/s11269-017-1878-0

Cited by 22 publications

(8 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Auto correlation function (ACF) (Dariane et al 2018) (Eq. 23) is an effective method to know stationarity of time series, which is helpful for mining the potential characteristics of data.…”

Section: Time Series Exploration and Decompositionmentioning

confidence: 99%

Multi-Model Coupling Water Demand Prediction Optimization Method for Megacities Based on Time Series Decomposition

Liu

Sang

Chang

et al. 2021

Preprint

View full text Add to dashboard Cite

The fluctuation of water supply is affected by the living habits and population mobility, so the daily water supply is significantly non-stationarity, which presents a great challenge to the water demand prediction based on data-driven model. To solve this problem, the Hodrick-Prescott (HP) and wavelet transform (WT) time series decomposition methods, and ensemble learning (EL) were introduced, coupling model bidirectional long short term memory (BLSTM), seasonal autoregressive integrated moving average (SARIMA) and Gaussian radial basis function neural network (GRBFNN) were developed, and interval prediction was carried out based on student's t-test (T-test). This research method was applied to the daily water demand prediction in Shenzhen and cross-validation was performed. It is found that the decomposed subseries has obvious law, and WT is superior to HP decomposition method. However, the maximum decomposition level (MDL) of WT should not be set too high, otherwise the trend characteristics of subseries will be weakened. The results show that the potential characteristics and quantitative relationships of historical data can be learned accurately based on WT and coupling model. Although the corona virus disease 2019 (COVID-19) outbreak in 2020 caused a variation in water supply law, this variation is still within the interval prediction. The WT and coupling model satisfactorily predicted water demand and provided the lowest mean square error (0.17%), mean relative error (0.1) and mean absolute error (3.32%) and the highest Nash-Sutcliffe efficiency (97.21%) and correlation coefficient (0.99) in testing set.

show abstract

Section: Time Series Exploration and Decompositionmentioning

confidence: 99%

Multi-Model Coupling Water Demand Prediction Optimization Method for Megacities Based on Time Series Decomposition

Liu

Sang

Chang

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Previous data-driven streamflow forecasting has focused on time series models, such as Box-Jenkins and autoregressive moving average (ARMA) models (Ramaswamy and Saleh 2020;Sun et al 2019;Zhang et al 2011). However, these time series models fail to reasonably forecast the nonlinear runoff series due to the stationarity assumption (Dariane et al 2018;He et al 2019). Machine learning models such as support vector regression (SVR) (Cheng et al 2015;Hadi and Tombul 2018;Lin et al 2009;Maity et al 2010;Vapnik et al 1996), gradient boosting regression trees (GBRT) (He et al 2020;Persson et al 2017;Zhang and Haghani 2015) and artificial neural networks (ANNs) (Chua and Wong 2011;Gauch et al 2020;Kisi et al 2012;Kratzert et al 2018) can address the nonstationary and nonlinear problems of runoff prediction (Friedman 2001;Kumar et al 2019;Vapnik et al 1996).…”

Section: Introductionmentioning

confidence: 99%

Climate-driven Model Based on Long Short-Term Memory and Bayesian Optimization for Multi-day-ahead Daily Streamflow Forecasting

Lian

Luo

Wang³

et al. 2021

Preprint

View full text Add to dashboard Cite

Many previous studies have developed decomposition and ensemble models to improve runoff forecasting performance. However, these decomposition-based models usually introduce large decomposition errors into the modeling process. Since the variation in runoff time series is greatly driven by climate change, many previous studies considering climate change focused on only rainfall-runoff modeling, with few meteorological factors as input. Therefore, a climate-driven streamflow forecasting (CDSF) framework was proposed to improve the runoff forecasting accuracy. This framework is realized using principal component analysis (PCA), long short-term memory (LSTM) and Bayesian optimization (BO) referred to as PCA-LSTM-BO. To validate the effectiveness and superiority of the PCA-LSTM-BO method with which one autoregressive LSTM model and two other CDSF models based on PCA, BO, and either support vector regression (SVR) or, gradient boosting regression trees (GBRT), namely, PCA-SVR-BO and PCA-GBRT-BO, respectively, were compared. A generalization performance index based on the Nash-Sutcliffe efficiency (NSE), called the GI(NSE) value, is proposed to evaluate the generalizability of the model. The results show that (1) the proposed model is significantly better than the other benchmark models in terms of the mean square error (MSE<=185.782), NSE>=0.819, and GI(NSE) <=0.223 for all the forecasting scenarios; (2) the PCA in the CDSF framework can improve the forecasting capacity and generalizability; (3) the CDSF framework is superior to the autoregressive LSTM models for all the forecasting scenarios; and (4) the GI(NSE) value is demonstrated to be effective in selecting the optimal model with a better generalizability.

show abstract

“…Presently, the state of the art of climate services are seasonal climate forecasts over a six-month horizon, coupled with downscaling models to turn climate input into streamflow (Yuan et al 2015). Forecasts for further-reaching time periods are an active field of research (Smith et al 2016, Dariane et al 2018), but are not currently used by industry. However, also downscaled seasonal forecasts are out of reach for most water agencies and utilities: it hence makes sense developing data-driven forecast systems to be integrated into the DSS to improve the quality of decisions, while research progresses, and resources and skills for implementing hydrological forecast are being developed.…”

Section: Introductionmentioning

confidence: 99%

A multi-scenario Decision Support System for real-time operation of over-year multi-reservoir systems 1. Model structure

2019

El Sawah, S. (Ed.) MODSIM2019, 23rd International Congress on Modelling and Simulation.

View full text Add to dashboard Cite

Although in these systems the risk of highly variable inflows is partially internalized in the design process by letting these reservoirs have large storage capacities, operation of these systems is challenged by changing demand levels, changing purposes and objectives, and climate change. Given their dimensions, these systems are often managed in a centralized fashion, with one authority planning allocations to different users. As inadequate management may operate the system either too conservatively or in a fashion unaware of the occurrence of long, multi-year dry spells, the decision-making process should be supported by appropriate decision support tools. When used on a regular basis, in a real-time mode, by updating current information on water resources status and using inflow forecasts, such support tools should be able to identify risky situations suitably in advance, suggest appropriate hedging policies and optimize the use of additional, costly, water resources that can help mitigate the impacts of sustained dry periods. One of the key factors for a successful management is obviously the ability to predict future inflows into the system suitably in advance. Presently, the state of the art of climate services is provided by seasonal climate forecasts over a six-month horizon, coupled with downscaling models to turn climate input into streamflow. Forecasts for further-reaching time periods are an active field of research, but are not currently used by industry. However, also downscaled seasonal forecasts are out of reach for most water agencies and utilities, as they imply the availability of ad-hoc skills, resources and facilities to acquire climate forecasts and turn them into inflow forecasts. It hence makes sense to develop simpler, data-driven forecast systems to be integrated into the decision support tools to improve the quality of decisions, while research progresses, and resources and skills for implementing hydrological forecast are being developed To this end, in this paper we introduce a multi-scenario mathematical programming (algebraic) model for real time management of a multi-reservoir system for water supply where consideration of a multi-year Forecasting Horizon (FH) is necessary given the physical features of the system. It is a linearized MIP (mixed integer programming) model that optimizes water allocations to municipal and irrigation demand centres driven by the objective of minimizing weighted total discounted costs (scarcity costs plus costs of additional water resources) along the multi-year FH, being the weights the occurrence probability of each of the three inflow scenarios. Constraints include 1) mass balances at system's nodes, 2) systems' topology, 3) component's capacity, 4) spills, 5) non-empty conditions on reservoir storage at the end of the FH. It is a multi-scenario optimization tool because future, uncertain inflows are modelled, until the end of the current water year, as three different inflow scenarios: low flows, normal flows and high flows. The optimization model...

show abstract

Long Term Streamflow Forecasting Using a Hybrid Entropy Model

Cited by 22 publications

References 27 publications

Multi-Model Coupling Water Demand Prediction Optimization Method for Megacities Based on Time Series Decomposition

Multi-Model Coupling Water Demand Prediction Optimization Method for Megacities Based on Time Series Decomposition

Climate-driven Model Based on Long Short-Term Memory and Bayesian Optimization for Multi-day-ahead Daily Streamflow Forecasting

A multi-scenario Decision Support System for real-time operation of over-year multi-reservoir systems 1. Model structure

Contact Info

Product

Resources

About