Correlated Time-Series in Multi-Day-Ahead Streamflow Forecasting Using Convolutional Networks

Barino, Felipe Oliveira; Silva, Vinícius N. H.; Lopez-Barbero, A.P.; Honório, Leonardo de Mello; Santos, Alexandre Bessa dos

doi:10.1109/access.2020.3040942

Cited by 12 publications

(5 citation statements)

References 58 publications

(34 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One of those models considered only the river flow in previous days, while the other considered that same variable combined with the turbidity. Both models obtained NSE and R 2 values higher than 0.92, while mean absolute percentage error (MAPE) and normalized RMSE were lower than Usually, in machine learning methods, better results are verified when antecedent streamflow is considered as a forcing variable (Barino et al 2020;Khosravi et al 2022). However, when the model is used in the simulation of future scenarios or periods when no observed data are available, the antecedent streamflow values to feed the model are those already calculated by the model in the previous iterations.…”

Section: D-cnn Modelmentioning

confidence: 90%

“…The results of the 1D-CNN model are in accordance with the results of several authors. Barino et al (2020) used two 1D-CNN models to predict multi-day ahead river flow in Madeira River, a tributary of the Amazon River, Brazil. One of those models considered only the river flow in previous days, while the other considered that same variable combined with the turbidity.…”

Section: D-cnn Modelmentioning

confidence: 99%

See 1 more Smart Citation

Assessing the reliability of a physical-based model and a convolutional neural network in an ungauged watershed for daily streamflow calculation: a case study in southern Portugal

Oliveira,

Ramos,

Simionesei

et al. 2024

Environ Earth Sci

View full text Add to dashboard Cite

The main goal of this study was to estimate inflows to the Maranhão reservoir, southern Portugal, using two distinct modeling approaches: a one-dimensional convolutional neural network (1D-CNN) model and a physically based model. The 1D-CNN was previously trained, validated, and tested in a sub-basin of the study area where observed streamflow values were available. The trained model was here subject to an improvement and applied to the entire watershed by replacing the forcing variables (accumulated and delayed precipitation) to make them correspond to the values of the entire watershed. The same way, the physically based MOHID-Land model was calibrated and validated for the same sub-basin, and the calibrated parameters were then applied to the entire watershed. Inflow values estimated by both models were validated considering a mass balance at the reservoir. The 1D-CNN model demonstrated a better performance in simulating daily values, peak flows, and the wet period. The MOHID-Land model showed a better performance in estimating streamflow values during dry periods and for a monthly analysis. Hence, results show the adequateness of both modeling solutions for integrating a decision support system aimed at supporting decision-makers in the management of water availability in an area subjected to increasing scarcity.

show abstract

Section: D-cnn Modelmentioning

confidence: 90%

Section: D-cnn Modelmentioning

confidence: 99%

Assessing the reliability of a physical-based model and a convolutional neural network in an ungauged watershed for daily streamflow calculation: a case study in southern Portugal

Oliveira,

Ramos,

Simionesei

et al. 2024

Environ Earth Sci

View full text Add to dashboard Cite

show abstract

“…Despite this, in recent times, these networks have been used in the scope of time series forecasting, namely, through 1D-CNNs. The prediction of PM2.5 levels in the air and the river’s flow are some examples of the application of 1D-CNNs. , RNNs constitute a class of networks in which the evolution of the state depends on the current input and the current state. This property makes it possible to perform context-dependent processing, allowing long-term dependencies to be learned.…”

Section: Wastewater Treatment Modeling Using Machine Learningmentioning

confidence: 99%

A Review of Computational Modeling in Wastewater Treatment Processes

Duarte,

Martins,

Oliveira

et al. 2023

ACS EST Water

View full text Add to dashboard Cite

Wastewater treatment companies are facing several challenges related to the optimization of energy efficiency, meeting more restricted water quality standards, and resource recovery potential. Over the past decades, computational models have gained recognition as effective tools for addressing some of these challenges, contributing to the economic and operational efficiencies of wastewater treatment plants (WWTPs). To predict the performance of WWTPs, numerous deterministic, stochastic, and time series-based models have been developed. Mechanistic models, incorporating physical and empirical knowledge, are dominant as predictive models. However, these models represent a simplification of reality, resulting in model structure uncertainty and a constant need for calibration. With the increasing amount of available data, data-driven models are becoming more attractive. The implementation of predictive models can revolutionize the way companies manage WWTPs by permitting the development of digital twins for process simulation in (near) real-time. In data-driven models, the structure is not explicitly specified but is instead determined by searching for relationships in the available data. Thus, the main objective of the present review is to discuss the implementation of machine learning models for the prediction of WWTP effluent characteristics and wastewater inflows as well as anomaly detection studies and energy consumption optimization in WWTPs. Furthermore, an overview considering the merging of both mechanistic and machine learning models resulting in hybrid models is presented as a promising approach. A critical assessment of the main gaps and future directions on the implementation of mathematical modeling in wastewater treatment processes is also presented, focusing on topics such as the explainability of data-driven models and the use of Transfer Learning processes.

show abstract

“…Developed by LeCun and Bengio (1995) to identify handwritten digits, CNN uses convolutional filtering to achieve high correlation with neighboring data. This means that this type of network works based on weight sharing concept, with the filters' coefficients being shared for all input positions and their number and values being essential to capture data patterns (Wang et al, 2019, Barino et al, 2020, Chong et al, 2020. CNNs are thus recognized as more suitable solutions to identify local patterns, with a certain identified pattern being able to be recognized in another independent occurrence (Tao et al, 2019).…”

Section: Neural Network Model For Reservoir Outflow Estimationmentioning

confidence: 99%

Direct integration of reservoirs’ operations in a hydrological model for streamflow estimation: coupling a CLSTM model with MOHID-Land

Oliveira

Ramos

Pinto

et al. 2023

Preprint

View full text Add to dashboard Cite

Abstract. Knowledge about streamflow regimes and values is essential for different activities and situations, in which justified decisions must be made. However, streamflow behavior is commonly assumed as non-linear, being controlled by various mechanisms that act on different temporal and spatial scales, making its estimate challenging. An example is the construction and operation of infrastructures such as dams and reservoirs in rivers. The challenges faced by modelers to correctly describe the impact of dams on hydrological systems are considerable. In this study, an already implemented, calibrated, and validated solution of MOHID-Land model for natural regime flow in Ulla River basin was considered as baseline. The referred watershed comprehends three reservoirs. Outflow values were estimated considering a basic operation rule for two of them (run-of-the-river dams) and considering a data-driven model of Convolutional Long Short-Term Memory (CLSTM) type for the other (high-capacity dam). The outflow values obtained with the CLSTM model were imposed in the hydrological model, while the hydrological model fed the CLSTM model with the level and the inflow of the reservoir. This coupled system was daily evaluated in two hydrometric stations located downstream of the reservoirs, resulting in an improved performance compared with the baseline application. The analysis of the modelled values with and without reservoirs further demonstrated that considering dams’ operations in the hydrological model resulted in an increase of the streamflow during the dry season and a decrease during the wet season but with no differences in the average streamflow. The coupled system is thus a promising solution for improving streamflow estimates in modified rivers.

show abstract

Correlated Time-Series in Multi-Day-Ahead Streamflow Forecasting Using Convolutional Networks

Cited by 12 publications

References 58 publications

Assessing the reliability of a physical-based model and a convolutional neural network in an ungauged watershed for daily streamflow calculation: a case study in southern Portugal

Assessing the reliability of a physical-based model and a convolutional neural network in an ungauged watershed for daily streamflow calculation: a case study in southern Portugal

A Review of Computational Modeling in Wastewater Treatment Processes

Direct integration of reservoirs’ operations in a hydrological model for streamflow estimation: coupling a CLSTM model with MOHID-Land

Contact Info

Product

Resources

About