Comparison of Machine Learning Models Performance on Simulating Reservoir Outflow: A Case Study of Two Reservoirs in Illinois, U.S.A.

Abstract

Reservoir outflow is an important variable for understanding hydrological processes and water resource management. Natural streamflow variation, in addition to the streamflow regulation provided by dams and reservoirs, can make streamflow difficult to understand and predict. This makes them a challenge to accurately simulate hydrologic processes at a daily scale. In this study, three Machine Learning (ML) algorithms, Random Forest (RF), Support Vector Machine (SVM), and Artificial Neural Network (ANN), were examined and compared to model reservoir outflow. Past, current, and future hydrologic and meteorological data were used as model inputs, and the outflow of next day was used as prediction. Simulation results demonstrated that all three models can reasonably simulate reservoir outflow. For Carlyle Lake, the coefficient of determination and Nash–Sutcliffe efficiency were each close to one for the three models. The coefficient of determination, relative mean bias, and root mean square error indicated that the SVM performed better than the RF and ANN, but the SVM output displayed a larger relative mean bias than that from RF and ANN. For Lake Shelbyville, the ANN model performed better than RF and SVM when considering the coefficient of determination, Nash–Sutcliffe efficiency, relative mean bias, and root mean square error. The study results demonstrate that the three ML algorithms (RF, SVM, and ANN) are all promising tools for simulating reservoir outflow. Both the accuracy and efficacy of the three ML algorithms are considered to support practitioners in planning reservoir management.

Toward Improved Simulations of Disruptive Reservoirs in Global Hydrological Modeling

“…The predictors utilized in this study to construct the RF model, their usage proxy and their ranks across the reservoirs based on permutation accuracy importance (Strobl et al, 2007), are shown in Table 2. These type of predictors have been widely used in the literature for example, Qie et al (2022); Tounsi et al (2022). Optimized RF models were obtained for each reservoir of the experiment.…”

“…Reservoir operation signal can be defined as the difference between the regulated and natural streamflow time series, where the latter would have been measured without the reservoir (Brunner & Naveau, 2023). There are numerous works that reconstruct the reservoir operation signal from observed streamflow time series measured downstream of a reservoir, such as the use of wavelet transform (Shiau & Huang, 2014; White et al., 2005), artificial neural networks (Ehsani et al., 2016; Qie et al., 2022), fuzzy rules (Coerver et al., 2018), harmonic regression models (Turner et al., 2021), Random forest (RF) and Support Vector Machines (Qie et al., 2022), generalized additive models (Brunner & Naveau, 2023), among others. With this in mind, it remains to be seen how useful the S2023 would be in conjunction with ML techniques in HMs.…”

“…where, V 1 is storage at first day of operational year, V f is the reservoir capacity, No analysis or exclusion threshold recommendation. regression models (Turner et al, 2021), Random forest (RF) and Support Vector Machines (Qie et al, 2022), generalized additive models (Brunner & Naveau, 2023), among others. With this in mind, it remains to be seen how useful the S2023 would be in conjunction with ML techniques in HMs.…”

Near-term forecasting of water reservoir storage capacities using long short-term memory

“…Similarly, Shamim et al (2016) showed that localized linear ML models are capable of predicting reservoir levels in Pakistan. Qie et al (2022) analyzed reservoir outflow for two sites in Illinois, USA and showed promising results using multiple different statistical techniques. ML models are also frequently used in water quality research, including recent studies in Vietnam (Nguyen et al, 2021), Hong Kong (Deng et al, 2021), and Ghana (Ewusi et al, 2021).…”

Modelling and forecasting daily streamflow with reservoir operationin the upper Chao Phraya River basin, Thailand

“…In the end, all the trees are combined into the so-called forest as it averages all predictions to produce a final output value. Considering its strengths, RF has been increasingly applied in reservoir modelling studies (e.g., Yang et al, 2016;Li et al, 2020;Qie et al, 2022).…”