Enhancing robustness of monthly streamflow forecasting model using gated recurrent unit based on improved grey wolf optimizer

Zhao, Xuehua; Lv, Hanfang; Shujin, Lv; Sang, Yuting; Wei, Yizhao; Zhu, Xiao

doi:10.1016/j.jhydrol.2021.126607

Cited by 59 publications

(25 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The primary stages of the grey wolf hunting process contain hunting, encircling and attacking the prey [58]. In the GWO algorithm, the hierarchy is structured as the fittest solution, the second-best solution, the third-best solution, and the rest of the candidate solutions [59]. Encircling the prey in the hunting process is represented by the following equations (Equations ( 4) and ( 5)):…”

Section: Grey Wolf Optimizationmentioning

confidence: 99%

Short-Term Streamflow Forecasting Using Hybrid Deep Learning Model Based on Grey Wolf Algorithm for Hydrological Time Series

Kilinç¹,

Yurtsever

2022

Sustainability

View full text Add to dashboard Cite

The effects of developing technology and rapid population growth on the environment have been expanding gradually. Particularly, the growth in water consumption has revealed the necessity of water management. In this sense, accurate flow estimation is important to water management. Therefore, in this study, a grey wolf algorithm (GWO)-based gated recurrent unit (GRU) hybrid model is proposed for streamflow forecasting. In the study, daily flow data of Üçtepe and Tuzla flow observation stations located in various water collection areas of the Seyhan basin were utilized. In the test and training analysis of the models, the first 75% of the data were used for training, and the remaining 25% for testing. The accuracy and success of the hybrid model were compared via the comparison model and linear regression, one of the most basic models of artificial neural networks. The estimation results of the models were analyzed using different statistical indexes. Better results were obtained for the GWO-GRU hybrid model compared to the benchmark models in all statistical metrics except SD at the Üçtepe station and the whole Tuzla station. At Üçtepe, the FMS, despite the RMSE and MAE of the hybrid model being 82.93 and 85.93 m3/s, was 124.57 m3/s, and it was 184.06 m3/s in the single GRU model. We achieved around 34% and 53% improvements, respectively. Additionally, the R2 values for Tuzla FMS were 0.9827 and 0.9558 from GWO-GRU and linear regression, respectively. It was observed that the hybrid GWO-GRU model could be used successfully in forecasting studies.

show abstract

Section: Grey Wolf Optimizationmentioning

confidence: 99%

Short-Term Streamflow Forecasting Using Hybrid Deep Learning Model Based on Grey Wolf Algorithm for Hydrological Time Series

Kilinç¹,

Yurtsever

2022

Sustainability

View full text Add to dashboard Cite

show abstract

“…(2021) used this advanced model for highly accurate forecasting of the urban water demand in Qom, Iran. In addition, the LSSVM model was used for water resource management by enhancing the accuracy of the prediction of mid‐to long‐term streamflow (Zhao et al., 2021). In methane transport modeling, Taherdangkoo et al.…”

Section: Introductionmentioning

confidence: 99%

“…Regarding the prospects for the application of the LSSVM model, Rezaali et al (2021) used this advanced model for highly accurate forecasting of the urban water demand in Qom, Iran. In addition, the LSSVM model was used for water resource management by enhancing the accuracy of the prediction of mid-to long-term streamflow (Zhao et al, 2021). In methane transport modeling, Taherdangkoo et al (2021) employed the LSSVM model to estimate methane solubility in aquatic environments for a variety of temperatures and pressures.…”

mentioning

confidence: 99%

Reconstructing Daily Discharge in a Megadelta Using Machine Learning Techniques

Thanh

Binh

Kantoush

et al. 2022

Water Resources Research

View full text Add to dashboard Cite

River discharge is a crucial indicator to understand terrestrial water cycles and supplies necessary information about water resource management (Adnan et al., 2020). Direct measurement of river discharge, such as employing the acoustic Doppler current profiler, is complicated, costly, time-consuming, and labor-intensive because it requires a number of current sensors and repeated surveys performed by boats and is thus unsafe under unfavorable flow and weather conditions (Gisen & Savenije, 2015;Matte et al., 2018). Other noncontact methods, including large-scale particle image velocimetry (LSPIV) (Akbarpour et al., 2020) and remote sensing (Kebede et al., 2020), have recently begun to be used for discharge measurements. Nevertheless, the use of these methods in contiguous monitoring of river discharge is not feasible; for example, LSPIV cannot measure discharge in large rivers because of limited camera coverage, while satellite images are not always available due to cloud cover, particularly during rainy seasons. As a result, at hydrological stations situated on rivers worldwide, flow discharge is not directly measured; rather, it is indirectly estimated either from the widely used stage-discharge rating curve (RC) method or from cubature, rating-fall, tide-correction, and coaxial graphical-correction methods (Matte et al., 2018), in which the stage (water level) is recorded at specific intervals (e.g., daily, hourly, or sub-daily) depending on the goal of the measurements. Due to technical, financial, maintenance and political instability issues, long-term flow discharge datasets may have gaps, resulting in the loss of information or the misinterpretation of historical flow regime changes and hydrological processes (Tencaliec et al., 2015). Therefore, it is important to reconstruct missing discharge values to reliably provide helpful information for water resource management at the basin scale.Several methods, including statistical methods, numerical models, and machine learning (ML) algorithms, have been employed to predict river flows. Recently, ML techniques, such as support vector regression (SVR) (Adnan et al., 2020;Luo et al., 2019), random forest (RF), Gaussian process regression (GPR) (Sun et al., 2014), M5

show abstract

“…However, under changing environments, the runoff process and associated hydrological system have been altered by human activities and climate change (Song et al, 2018;Sun et al, 2018Sun et al, , 2022Jiang et al, 2019;Lu et al, 2020;Yan et al, 2020;Hu et al, 2022), and the runoff series becomes nonlinear and nonstationary, which makes it challenging to capture the variation characteristics of runoff (Sun et al, 2014;Lin et al, 2020;Yan et al, 2021b;Samantaray et al, 2022a;Samantaray et al, 2022b;Samantaray et al, 2022c;Zhou et al, 2022). Therefore, there is an urgent need to develop a runoff prediction model with robustness and high forecasting accuracy under a changing environment (Sit et al, 2020;Niu et al, 2021;Zhao et al, 2021). In recent years, there have been many studies trying to transform the complex runoff series into stationary subsequences using wavelets or mode decomposition methods, and then predict the sub-sequences to improve the accuracy of prediction (Meng et al, 2019;Feng et al, 2020b;Niu et al, 2020).…”

Section: Introductionmentioning

confidence: 99%

Climate-informed monthly runoff prediction model using machine learning and feature importance analysis

Yan

Lei

Jiang

et al. 2022

Front. Environ. Sci.

View full text Add to dashboard Cite

Accurate runoff prediction can provide a reliable decision-making basis for flood and drought disaster prevention and scientific allocation of water resources. Selecting appropriate predictors is an effective way to improve the accuracy of runoff prediction. However, the runoff process is influenced by numerous local and global hydrometeorological factors, and there is still no universal approach about the selection of suitable predictors from these factors. To address this problem, we proposed a runoff prediction model by combining machine learning (ML) and feature importance analysis (FIA-ML). Specifically, take the monthly runoff prediction of Yingluoxia, China as an example, the FIA-ML model uses mutual information (MI) and feature importance ranking method based on random forest (RF) to screen suitable predictors, from 130 global climate factors and several local hydrometeorological information, as the input of ML models, namely the hybrid kernel support vector machine (HKSVM), extreme learning machine (ELM), generalized regression neural network (GRNN), and multiple linear regression (MLR). An improved particle swarm optimization (IPSO) is used to estimate model parameters of ML. The results indicated that the performance of the FIA-ML is better than widely-used long short-term memory neural network (LSTM) and seasonal autoregressive integrated moving average (SARIMA). Particularly, the Nash-Sutcliffe Efficiency coefficients of the FIA-ML models with HKSVM and ELM were both greater than 0.9. More importantly, the FIA-ML models can explicitly explain which physical factors have significant impacts on runoff, thus strengthening the physical meaning of the runoff prediction model.

show abstract

Enhancing robustness of monthly streamflow forecasting model using gated recurrent unit based on improved grey wolf optimizer

Cited by 59 publications

References 60 publications

Short-Term Streamflow Forecasting Using Hybrid Deep Learning Model Based on Grey Wolf Algorithm for Hydrological Time Series

Short-Term Streamflow Forecasting Using Hybrid Deep Learning Model Based on Grey Wolf Algorithm for Hydrological Time Series

Reconstructing Daily Discharge in a Megadelta Using Machine Learning Techniques

Climate-informed monthly runoff prediction model using machine learning and feature importance analysis

Contact Info

Product

Resources

About