A Novel Physics‐Aware Machine Learning‐Based Dynamic Error Correction Model for Improving Streamflow Forecast Accuracy

Roy, Abhinanda; Kasiviswanathan, K. S.; Patidar, Sandhya; Adeloye, Adebayo J.; Soundharajan, B.; Ojha, Chandra Shekhar Prasad

doi:10.1029/2022wr033318

Cited by 8 publications

(6 citation statements)

References 141 publications

(448 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Roy et al [95] recently proposed a dynamic error-correction approach for streamflow forecasting using an RF model with the HBV model. The framework demonstrated consistent accuracy, especially for long lead time and low flow predictions.…”

Section: Residual Error Modelingmentioning

confidence: 99%

Enhancing Streamflow Prediction Physically Consistently Using Process-Based Modeling and Domain Knowledge: A Review

Yifru,

Lim,

Lee

2024

Sustainability

View full text Add to dashboard Cite

Streamflow prediction (SFP) constitutes a fundamental basis for reliable drought and flood forecasting, optimal reservoir management, and equitable water allocation. Despite significant advancements in the field, accurately predicting extreme events continues to be a persistent challenge due to complex surface and subsurface watershed processes. Therefore, in addition to the fundamental framework, numerous techniques have been used to enhance prediction accuracy and physical consistency. This work provides a well-organized review of more than two decades of efforts to enhance SFP in a physically consistent way using process modeling and flow domain knowledge. This review covers hydrograph analysis, baseflow separation, and process-based modeling (PBM) approaches. This paper provides an in-depth analysis of each technique and a discussion of their applications. Additionally, the existing techniques are categorized, revealing research gaps and promising avenues for future research. Overall, this review paper offers valuable insights into the current state of enhanced SFP within a physically consistent, domain knowledge-informed data-driven modeling framework.

show abstract

Section: Residual Error Modelingmentioning

confidence: 99%

Enhancing Streamflow Prediction Physically Consistently Using Process-Based Modeling and Domain Knowledge: A Review

Yifru,

Lim,

Lee

2024

Sustainability

View full text Add to dashboard Cite

show abstract

“…However, in the low-carbon energy supply chain, due to its particularity, the traditional game theory model and optimization method are facing great challenges. As an algorithm technology that can automatically learn and improve, machine learning can analyze and predict a large amount of data in a low-carbon energy supply chain, thus providing data-based decision support [5]. Game theory can work out the optimal decision-making scheme by analyzing the strategy and interest relationship among supply chain participants.…”

Section: Introductionmentioning

confidence: 99%

Optimal scheduling method of low-carbon energy supply chain based on machine learning and game theory

Yan,

Zhang,

Wang

et al. 2024

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

Aiming at the quality and stability of a low-carbon new energy supply chain, this paper puts forward an optimal scheduling method for a low-carbon energy supply chain based on machine learning and game theory. By establishing an index and preprocessing keywords, the similarity between the index and the data in the platform database is calculated, and the data-sharing result of the low-carbon energy supply chain is obtained. This paper analyzes the game benefits of each subject in the optimization of the low-carbon energy supply chain, obtains the game matrix, and realizes the optimal scheduling of the low-carbon energy supply chain. The experimental results show that the scheduling convergence error of the low-carbon energy supply chain is low, the accuracy is high, and the load supply equipment can run stably.

show abstract

“…Nevertheless, pure ML/DL algorithms are hardly interpretable and do not use the effective physical models, solvers as well as adjoint (differentiation) techniques developped over the past century. Hybrid approaches, that leverage ML/DL in sequential combination with process-based numerical models via their inputs/outputs, have been explored recently and enable to improve the accuracy of hydrologic predictions (e.g., Konapala et al (2020), with DA in Roy et al (2023) or with UQ in Tran et al (2023)).…”

Section: Introductionmentioning

confidence: 99%

Multiscale Learnable Physical Modeling and Data Assimilation Framework: Application to High-Resolution Regionalized Hydrological Simulation of Flash Floods

Huynh,

Garambois,

Renard

et al. 2024

Preprint

View full text Add to dashboard Cite

To advance the discovery of scale-relevant hydrological laws while better exploiting massive multi-source data, combining machine learning and process-based modeling is compelling, as recently demonstrated in lumped hydrological modeling. This article introduces MLPM-PR, a new and powerful framework standing for Multiscale spatially distributed Learnable Physical Modeling and learnable Parameter Regionalization with data assimilation. MLPM-PR crucially builds on a differentiable model that couples (i) two neural networks for processes learning and parameters regionalization, (ii) a grid-based differentiable conceptual hydrological model, and (iii) a simple kinematic wave routing. The approach is tested on a challenging flash flood-prone multi-catchment modeling setup at high spatio-temporal resolution (1km, 1h). Discharge prediction performances highlight the accuracy and robustness of MLPM-PR, both spatially and temporally compared to classical approaches. The physical interpretability of spatially distributed parameters and internal states shows the nuanced behavior of the hybrid model and its adaptability to diverse hydrological responses.

show abstract

A Novel Physics‐Aware Machine Learning‐Based Dynamic Error Correction Model for Improving Streamflow Forecast Accuracy

Abstract: Accurate and reliable streamflow forecasts provide valuable information ensuring effective management of water resource structures designed for flood protection, allocation of irrigation water, hydropower generation, and many more (

Cited by 8 publications

References 141 publications

Enhancing Streamflow Prediction Physically Consistently Using Process-Based Modeling and Domain Knowledge: A Review

Enhancing Streamflow Prediction Physically Consistently Using Process-Based Modeling and Domain Knowledge: A Review

Optimal scheduling method of low-carbon energy supply chain based on machine learning and game theory

Multiscale Learnable Physical Modeling and Data Assimilation Framework: Application to High-Resolution Regionalized Hydrological Simulation of Flash Floods

Contact Info

Product

Resources

About