Monthly Streamflow Forecasts Research Articles

Monthly streamflow forecasting based on meteorological data from a nearby station

ABSTRACT Monthly streamflow forecasting is critical for improving water resource management. In this study, several base-classifier data-mining algorithms – conjunctive rule (CR), isotonic regression (ISOR), sequential minimal optimization regression (SMOR) – as well as several hybrid data-mining techniques – disjoint aggregating or dagging (DA)-CR, DA-ISOR, and DA-SMOR – that combine dagging with these algorithms were developed and applied to forecasting streamflow 3 and 12 months into the future (i.e., Qt+3 and Qt+12) based on meteorological data from a nearby station. Thirty years of data (from 1988 to 2018) that included precipitation, minimum relative humidity, maximum relative humidity, evaporation, hours of sunshine, maximum temperature, minimum temperature, wind speed, and streamflow were collected at the Kermanshah synoptic station were input to develop and evaluate several models. Varying combinations of the input data were tested to find the optimal set to employ. The models were validated and compared using several quantitative statistical indices. Though all satisfactorily predicted monthly streamflow, the hybrid models (DA-CR, DA-ISOR, and DA-SMOR) outperformed the base-classifier models (CR, ISOR, and SMOR), proving that dagging improved data-mining models significantly. Of the hybrid models, D-SMOR was the best. The models developed in this study are cost-effective tools for quick and accurate monthly streamflow forecasting.

Investigating the potential of EMA-embedded feature selection method for ESVR and LSTM to enhance the robustness of monthly streamflow forecasting from local meteorological information

Accurate forecast of monthly streamflow is helpful to improve the social capability in risk management. Diverse input features are crucial to the accuracy of machine learning-based monthly streamflow forecast models. The embedded feature selection (EFS) method based on binary-coded metaheuristic algorithm is adept at parallel optimization of feature selection and model hyperparameter. It has been rarely studied coupled with deep learning models in monthly streamflow forecasting. This study further investigated the potential of EFS method using the well-known Long short-term memory network (LSTM) and a state-of-the-art model named Extreme support vector regression (ESVR), with the Evolutionary mating algorithm (EMA) as the search algorithm. The benchmark models were established using EMA for hyperparameter optimization without feature selection. Ten basins with distinct hydroclimate characteristics minimally impacted by human activities from the Catchment Attributes and Meteorology for Large-sample Studies (CAMELS) data set were used to comprehensively examine the proposed models. Nonparametric approaches including the Friedman test and Wilcoxon signed ranks test were utilized to rank the models and detect the significance of the performance variation brought by EFS. The kernel ESVR model coupled with EFS-EMA (EFS-KESVREMA) ranked the first in the 1-month-ahead experiment with NSE = 0.734–0.944. EFS-ESVREMA and EFS-LSTMEMA ranked second and third, with NSE = 0.746–0.94 and NSE = 0.701–0.94 respectively. The results also demonstrated that EFS-EMA helps the benchmark models improve prediction accuracy at a confidence level of 95 % but it has no obvious effect on the standard SVR. As the forecast horizon increased to 3- and 5-month-ahead, EFS-ESVREMA maintained a reliable performance. Whereas EFS-LSTMEMA experienced a significant accuracy decrease in the 5-month-ahead experiment. The proposed models and relevant findings in this study have certain reference value for researchers in monthly streamflow forecasting.

A framework of integrating heterogeneous data sources for monthly streamflow prediction using a state-of-the-art deep learning model

Deep learning has been widely used in hydrological prediction such as monthly streamflow and its performance is usually dependent on the abundance of training data. Even though the interest in using predictors from multiple data sources (e.g., streamflow observations, local meteorological data, and large-scale climate indexes) to train deep learning models for monthly streamflow prediction is growing, these predictors are usually selected from historical periods. Such approaches have limitations that the non-stationary future climate information is not included in the deep learning models. Climate models can provide non-stationary climate information for the future period, which may be useful for monthly streamflow prediction. Therefore, the objectives of this study are to (1) investigate the added value of using predictors derived from global climate models (GCMs) for monthly streamflow prediction based on a state-of-the-art deep learning model, and (2) propose a framework for integrating heterogeneous data sources for monthly streamflow prediction. The framework consists of five integral components: data collection, predictor combination, predictor selection, model construction, and results evaluation. In this study, a hybrid deep learning model combining Convolutional Neural Network and Gated Recurrent Unit was applied for six hydrological stations from mainstream and six stations from tributaries of the Yangtze River. Historical hydroclimate data and GCM hindcasts from 1982 to 2010 are used in monthly streamflow forecasts. Hindcasts are retrospective forecasts for many variables in the past, ideally conducted using the same model used for real-time forecasts. The results show that GCM hindcasts are useful predictors to improve the prediction accuracy for monthly streamflow predictions, especially for the 1- and 3-month lead times. Combining GCM hindcasts with either historical meteorological data or historical streamflow observations and meteorological data as predictors generally provides the best predictive performance. In addition, using large-scale climate indexes as ancillary information is able to improve the predictive performance at a lead time of 6 months. For lead times of 1, 3, and 6 months, the Kling‐Gupta efficiency (KGE) and the mean relative error (MRE)) metrics calculated based on the best-performing predictor combinations are satisfactory for hydrological stations in both mainstream and tributaries, with the median KGE being higher than 0.85 and 0.62, and the median MRE being approximately 20 % and 40 %, respectively, suggesting the monthly streamflow predictions are better for mainstream than for tributaries. Overall results show that (1) the inclusion of predicted information from GCMs can improve the performance for monthly streamflow prediction, and (2) the way of combining various heterogeneous data sources is crucial.

Monthly Streamflow Forecasting Using Decomposition-Based Hybridization with Two-step Verification Method Over the Mangla Watershed, Pakistan

Monthly Streamflow Forecasting Using Decomposition-Based Hybridization with Two-step Verification Method Over the Mangla Watershed, Pakistan

Multi-Step-Ahead Monthly Streamflow Forecasting Using Convolutional Neural Networks

Multi-Step-Ahead Monthly Streamflow Forecasting Using Convolutional Neural Networks

Multi-Variables-Driven Model Based on Random Forest and Gaussian Process Regression for Monthly Streamflow Forecasting

Due to the inherent non-stationary and nonlinear characteristics of original streamflow and the complicated relationship between multi-scale predictors and streamflow, accurate and reliable monthly streamflow forecasting is quite difficult. In this paper, a multi-scale-variables-driven streamflow forecasting (MVDSF) framework was proposed to improve the runoff forecasting accuracy and provide more information for decision-making. This framework was realized by integrating random forest (RF) and Gaussian process regression (GPR) with multi-scale variables (hydrometeorological and climate predictors) as inputs and is referred to as RF-GPR-MV. To validate the effectiveness and superiority of the RF-GPR-MV model, it was implemented for multi-step-ahead monthly streamflow forecasts with horizons of 1 to 12 months for two key hydrological stations in the Jinsha River basin, Southwest China. Other MVDSF models based on the Pearson correlation coefficient (PCC) and GPR with/without multi-scale variables or the PCC and a backpropagation neural network (BP) or general regression neural network (GRNN), with only previous streamflow and precipitation, namely, PCC-GPR-MV, PCC-GPR-QP, PCC-BP-QP, and PCC-GRNN-QP, respectively, were selected as benchmarks. Experimental results indicated that the proposed model was superior to the other benchmark models in terms of the Nash–Sutcliffe efficiency (NSE) for almost all forecasting scenarios, especially for forecasting with longer lead times. Additionally, the results also confirmed that the addition of large-scale climate and circulation factors was beneficial for promoting the streamflow forecasting ability, with an average contribution rate of about 15%. The RF in the MVDSF framework improved the forecasting performance, with an average contribution rate of about 25%. This improvement was more pronounced when the lead time exceeded 3 months. Moreover, the proposed model could also provide prediction intervals (PIs) to characterize forecast uncertainty, as supplementary information to further help decision makers in relevant departments to avoid risks in water resources management.

Metalearning Approach Coupled with CMIP6 Multi-GCM for Future Monthly Streamflow Forecasting

Spatial and temporal variability of streamflow due to climate change affects hydrological processes and irrigation demands at a basin scale. This study investigated the impacts of climate change on the Kurau River in Malaysia using metalearning, an ensemble machine learning technique using support vector regression (SVR) and random forest (RF) coupled with the Coupled Model Intercomparison Project CMIP6 multi-Global Climate Model (GCM). Five global climate models and three shared socioeconomic pathways (SSP1-2.6, SSP2-4.5, and SSP5-8.5) were used. The climate sequences generated by the delta change factor method were applied as input to the metalearning model to predict the streamflow changes in the Kurau River from 2021 to 2080. The model fitted reasonably well, with Kling–Gupta efficiency (KGE), Nash–Sutcliffe efficiency (NSE), percent bias (PBias), and RMS Error (RMSE) of 0.79, 0.83, 2.52, and 4.51, respectively, for the training period (1976–1995) and 0.72, 0.72, 5.85, and 6.90, respectively, for the testing period (1995–2005). Future projections of multi-GCM over the 2021–2080 period under three SSPs predicted an increase in rainfall for all months except April–June during the dry period (off-season), with a higher increase occurring during the wet period (main season). Temperature projections indicated an increase in maximum and minimum temperatures under all SSP scenarios, with a higher increase of approximately 2.0°C under SSP5-8.5 predicted during the 2051–2080 period relative to the baseline period of 1976–2005. The model predicted that the seasonal changes in streamflow of two planting periods range between −7.5% and 7.1% and between 1.2% and 5.9% during the off-season and the main season, respectively. A significant streamflow decrease was predicted in April and May for all SSP scenarios due to high temperatures during the off-season, with SSP5-8.5 being the worst. The impact assessment of climate variabilities on the availability of water resources is vital to identify appropriate adaptation strategies to deal with an expected increase in irrigation demand due to global warming in the future. The predicted future streamflow under the potential climate change impacts is crucial for the Bukit Merah Reservoir to establish suitable operational policies for irrigation release.

Forecasting of Streamflow and Comparison of Artificial Intelligence Methods: A Case Study for Meram Stream in Konya, Turkey

The planning and management of water resources are affected by streamflow. The analysis of the sustainability of water resources has used well-grounded methods such as artificial neural networks, used for streamflow forecasting by researchers in recent years. The main aim of this study is to evaluate the performance of various methods for long-term forecasting from the data of the mean monthly streamflows between 1981 and 2017 from the Kucukmuhsine station on the Meram Stream in the Turkish province of Konya. For that reason, the multilayer perceptron (MLP), long short-term memory (LSTM), and adaptive neuro-fuzzy inference system (ANFIS) artificial intelligence techniques were employed in this study for the long-term forecasting of streamflow for 12 months, 24 months, and 36 months. The mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R2) were used to evaluate the performance of the models developed to make predictions using the data from 1981 to 2017, and the Mann-Whitney test was applied to examine the differences between the actual data from 2018 to 2020 and each model’s forecasted results for those three years. The LSTM model showed superiority based on the values of R2 (calculated as 0.730) and RMSE (lowest value of 0.510), whereas the MLP yielded better prediction accuracy as reflected by the value of MAE (lowest value of 0.519). The ANFIS model did not have the best prediction ability for any of the criteria. In accordance with the Mann-Whitney test results, LSTM and MLP indicated no significant difference between the actual data from 2018 to 2020 and the forecasted values; whereas, there was a significant difference for the ANFIS model at a confidence level of 95%. The results showed that the LSTM model had a better prediction performance, surpassing the MLP and ANFIS models, when comparing mean monthly streamflow forecasts.

SARIMA approach for modelling and forecasting of monthly stream flow of Godavari river in India

SARIMA approach for modelling and forecasting of monthly stream flow of Godavari river in India

A Stepwise and Dynamic C-Vine Copula–Based Approach for Nonstationary Monthly Streamflow Forecasts

AbstractIn recent years, the stationary assumption of hydroclimatic variables as well as their temporal and spatial dependence structure have been challenged due to anthropogenic and climate change...

Monthly Streamflow Forecasting Using Convolutional Neural Network

Monthly streamflow forecasting is vital for managing water resources. Recently, numerous studies have explored and evidenced the potential of artificial intelligence (AI) models in hydrological forecasting. In this study, the feasibility of the convolutional neural network (CNN), a deep learning method, is explored for monthly streamflow forecasting. CNN can automatically extract critical features from numerous inputs with its convolution–pooling mechanism, which is a distinct advantage compared with other AI models. Hydrological and large-scale atmospheric circulation variables, including rainfall, streamflow, and atmospheric circulation factors are used to establish models and forecast streamflow for Huanren Reservoir and Xiangjiaba Hydropower Station, China. The artificial neural network (ANN) and extreme learning machine (ELM) with inputs identified based on cross-correlation and mutual information analyses are established for comparative analyses. The performances of these models are assessed with several statistical metrics and graphical evaluation methods. The results show that CNN outperforms ANN and ELM in all statistical measures. Moreover, CNN shows better stability in forecasting accuracy.

Comparative study of forecasting approaches in monthly streamflow series from Brazilian hydroelectric plants using Extreme Learning Machines and Box & Jenkins models

Abstract Several activities regarding water resources management are dependent on accurate monthly streamflow forecasting, such as flood control, reservoir operation, water supply planning, hydropower generation, energy matrix planning, among others. Most of the literature is focused on propose, compare, and evaluate the forecasting models. However, the decision on forecasting approaches plays a significant role in such models’ performance. In this paper, we are focused on investigating and confront the following forecasting approaches: i) use of a single model for the whole series (annual approach) versus using 12 models, each one responsible for predicting each month (monthly approach); ii) for multistep forecasting, the use of direct and recursive methods. The forecasting models addressed are the linear Autoregressive (AR) and Periodic Autoregressive (PAR) models, from the Box & Jenkins family, and the Extreme Learning Machines (ELM), an artificial neural network architecture. The computational analysis involves 20 time series associated with hydroelectric plants indicated that the monthly approach with the direct multistep method achieved the best overall performances, except for the cases in which the coefficient of variation is higher than two. In this case, the recursive approach stood out. Also, the ELM overcame the linear models in most cases.

Examination and comparison of binary metaheuristic wrapper-based input variable selection for local and global climate information-driven one-step monthly streamflow forecasting

The use of data-driven models to forecast streamflow has received substantial attention from scholars in recent years. However, systematic studies have not been performed to examine binary metaheuristic wrapper-based input variable selection (BMWIVS) in real-world streamflow forecasting. In this study, we explored binary metaheuristic-based shallow machine learning wrappers for one-step monthly streamflow forecasting using local weather information and global climate indices from three catchments with different hydroclimatic conditions. First, the maximal information coefficient (MIC) was employed to investigate the correlations among the forecasting target, streamflow and candidate input variables, which included both local and global climate information. Then, the BMWIVS models obtained by combining eight binary metaheuristic algorithms, five commonly used shallow machine learning algorithms, two combined filter-based input variable selection (FIVS) methods, and two forecasting methods were examined. Finally, the performance of each model was compared with the performance of typical benchmark models, including the univariate seasonal autoregressive integrated moving average model, five machine learning algorithms with no input variable selection, and five machine learning algorithms that use five different FIVS methods. The experimental results emphasized three significant findings. First, an appropriate input variable selection method should be selected in practice because several examined wrappers were inferior to the benchmark models. Second, the BMWIVS model that combined the regularized extreme learning machine method, binary gray wolf optimizer, FIVS results-based initialization method, and forecasted values averaged over multiple runs yielded the best performance in the three cases studied. Third, the correlations in terms of the MIC between the global climate indices and streamflow were lower than those between local weather information and streamflow, and the best wrapper and FIVS would select more local weather information variables than global climate index variables, which suggests that global climate information can be complementary to local weather information for one-step monthly streamflow forecasting. These findings have remarkable practical applications for forecasting monthly streamflow.

Binary Grey Wolf Optimization-Regularized Extreme Learning Machine Wrapper Coupled with the Boruta Algorithm for Monthly Streamflow Forecasting

Input variable selection plays a key role in data-driven streamflow forecasting models. In this study, we propose a two-stage wrapper model to drive one-month-ahead streamflow forecasting in the context of high-dimensional candidate input variables. Initially, the Boruta algorithm, a feature selection method, was applied to select all the relevant input variables for the streamflow series. Then, a novel binary grey wolf optimizer (BGWO)-regularized extreme learning machine (RELM) wrapper was derived. We carried out experiments on two US catchments with 132 candidate input variables, including local meteorological information, global climatic indices, and lags of the streamflow series. Furthermore, the sensitivities of the proposed model in terms of the optimal objective function were compared. The results indicate two important findings. First, the proposed model outperformed commonly used models in terms of four error evaluation criteria. Second, for the proposed model, the root mean square error is a more suitable criterion than the Akaike information criterion (AIC) and the Bayesian information criterion (BIC) for the optimal objective function. These findings are of great reference value for developing ELM models for streamflow forecasting.

Forecasting of Monthly Flow for the White Nile River (South Sudan)

Forecasting of monthly streamflow for the White Nile River at Malakal station is a crucial aspect for different water resources projects in both countries Sudan and South Sudan. For instance, the operation of Jabal al Awliya dam in central Sudan entirely depends upon the measured flow of this station. In this paper, linear stochastic models well-known as seasonal autoregressive integrated moving average [SARIMA] models were used to model and forecast monthly flow of White Nile River in Malakal station, South Sudan. For the analysis, monthly flow data for the years running from 1970 up to 2013 were used. A scrutiny of the original series proves a yearly seasonal pattern. The results of Phillips-Perron (PP) test and Augmented Dickey Fuller (ADF) test on the streamflow series show that this series is not stationary. This non-stationarity was removed using first order seasonal differencing (i.e. twelve-monthly) preceding to the development of the model. The SARIMA (1,0,1)×(0,1,1)12 model was selected as the most suitable for modeling and forecasting monthly flow for White Nile River. It was found that the model was proper to forecast three successive years of monthly flow, which may help the experts to institute priorities for various water resources management in both countries.

Monthly Streamflow Forecasting Using Machine Learning

Streamflow forecasting holds a vital role in planning, design, and management of basin water resources. Accurate streamflow forecast provides a more efficient design of water resources systems technically and economically. In this study, various machine learning algorithms were used to model monthly streamflow data in the Coruh river basin, Turkey. For modeling, Support Vector Machines (SVM), Adaptive Boosting (AdaBoost), K-Nearest Neighbours (KNN) and Random Forest algorithms were considered and compared. Based on the test scores of the considered models with the hyperparameters, Random Forest based model outperforms all other models.

Monthly Streamflow Forecasting Using ELM-IPSO Based on Phase Space Reconstruction

We have developed a hybrid model that integrates chaos theory and an extreme learning machine with optimal parameters selected using an improved particle swarm optimization (ELM-IPSO) for monthly runoff analysis and prediction. Monthly streamflow data covering a period of 55 years from Daiying hydrological station in the Chaohe River basin in northern China were used for the study. The Lyapunov exponent, the correlation dimension method, and the nonlinear prediction method were used to characterize the streamflow data. With the time series of the reconstructed phase space matrix as input variables, an improved particle swarm optimization was used to improve the performance of the extreme learning machine. Finally, the optimal chaotic ensemble learning model for monthly streamflow prediction was obtained. The accuracy of the predictions of the streamflow series (linear correlation coefficient of about 0.89 and efficiency coefficient of about 0.78) indicate the validity of our approach for predicting streamflow dynamics. The developed method had a higher prediction accuracy compared with an auto-regression method, an artificial neural network, an extreme learning machine with genetic algorithm and with PSO algorithm, suggesting that ELM-IPSO is an efficient method for monthly streamflow prediction.

Comparison of eight filter-based feature selection methods for monthly streamflow forecasting – Three case studies on CAMELS data sets

Recently, there has been an increased emphasis on employing data-driven models to forecast streamflow. However, in these data-driven models used for forecasting monthly streamflow, the performances of filter-based feature selection (FFS) methods have not been studied in detail. In this study, we investigated the effectiveness of eight common FFS methods, namely, linear Pearson correlation, partial linear Pearson correlation (PCI), mutual information (MI), conditional MI, partial MI, maximal relevance minimal redundancy Pearson correlation, maximal relevance minimal redundancy MI and gamma test methods, on three regression models, namely multiple linear regression (MLR), ensemble extreme learning machine (enELM) and k-nearest neighbor (KNN) regression, for real-world one-month-ahead streamflow forecasting. The study was conducted on three cases from the Catchment Attributes and Meteorology for Large-sample Studies (CAMELS) data sets. Furthermore, two termination criterion (TC) methods, the Hampel test and resampling, were comparatively analyzed. The results of this study highlight three important findings. First, there was no dominant FFS method that coupled with enELM or KNN. Second, when resampling was applied to select a final model in the candidate combinations of the eight FFS methods and three regression models, PCI was the most favorable FFS method for the final model. Finally, the Hampel test TC was superior to the resampling TC in terms of stability and anti-overfitting. These findings have significant practical reference value for real-world monthly streamflow forecasting.

Streamflow and rainfall forecasting by two long short-term memory-based models

Prediction of streamflow and rainfall is important for water resources planning and management. In this study, we developed two hybrid models, based on long short-term memory network (LSTM), for monthly streamflow and rainfall forecasting. One model, wavelet-LSTM (namely, WLSTM), applied a trous algorithm of wavelet transform to do series decomposition, and the other, convolutional LSTM (namely, CLSTM), coupled convolutional neural network to extract temporal features. Two streamflow datasets and two rainfall datasets are used to evaluate the proposed models. The prediction accuracy of WLSTM and CLSTM was compared with that of multi-layer perceptron (MLP) and LSTM. Results indicated that LSTM was applicable for time series prediction, but WLSTM and CLSTM were superior alternatives.

Long-term monthly streamflow forecasting in humid and semiarid regions

Long-term monthly streamflow forecasting has great importance in the water resource system planning. However, its modelling in extreme cases is difficult, especially in semiarid regions. The main purpose of this paper is to evaluate the accuracy of artificial neural networks (ANNs) and hybrid wavelet-artificial neural networks (WA-ANNs) for multi-step monthly streamflow forecasting in two different hydro-climatic regions in Northern Algeria. Different issues have been addressed, both those related to the model’s structure and those related to wavelet transform. The discrete wavelet transform has been used for the preprocessing of the input variables of the hybrid models, and the multi-step streamflow forecast was carried out by means of a recursive approach. The study demonstrated that WA-ANN models outperform the single ANN models for the two hydro-climatic regions. According to the performance criteria used, the results highlighted the ability of WA-ANN models with lagged streamflows, precipitations and evapotranspirations to forecast up to 19 months for the humid region with good accuracy [Nash–Sutcliffe criterion (Ns) equal 0.63], whereas, for the semiarid region, the introduction of evapotranspirations does not improve the model’s accuracy for long lead time (Ns less than 0.6 for all combinations used). The maximum lead time achieved, for the semiarid region, was about 13 months, with only lagged streamflows as inputs.