Physics-Guided Machine Learning for Scientific Discovery: An Application in Simulating Lake Temperature Profiles

Abstract

Physics-based models of dynamical systems are often used to study engineering and environmental systems. Despite their extensive use, these models have several well-known limitations due to simplified representations of the physical processes being modeled or challenges in selecting appropriate parameters. While-state-of-the-art machine learning models can sometimes outperform physics-based models given ample amount of training data, they can produce results that are physically inconsistent. This paper proposes a physics-guided recurrent neural network model (PGRNN) that combines RNNs and physics-based models to leverage their complementary strengths and improves the modeling of physical processes. Specifically, we show that a PGRNN can improve prediction accuracy over that of physics-based models, while generating outputs consistent with physical laws. An important aspect of our PGRNN approach lies in its ability to incorporate the knowledge encoded in physics-based models. This allows training the PGRNN model using very few true observed data while also ensuring high prediction accuracy. Although we present and evaluate this methodology in the context of modeling the dynamics of temperature in lakes, it is applicable more widely to a range of scientific and engineering disciplines where physics-based (also known as mechanistic) models are used, e.g., climate science, materials science, computational chemistry, and biomedicine.

arXiv commentarXiv admin note: text overlap with arXiv:1810.13075

References in corpus (10)

• Adam: A Method for Stochastic Optimization
• Discovering governing equations from data: Sparse identification of nonlinear dynamical systems
• Deep Hidden Physics Models: Deep Learning of Nonlinear Partial Differential Equations
• Multistep Neural Networks for Data-driven Discovery of Nonlinear Dynamical Systems
• Achieving Conservation of Energy in Neural Network Emulators for Climate Modeling
• Hidden Fluid Mechanics: A Navier-Stokes Informed Deep Learning Framework for Assimilating Flow Visualization Data
• Hybrid modeling and prediction of dynamical systems
• Learning Parameters and Constitutive Relationships with Physics Informed Deep Neural Networks
• DR-RNN: A deep residual recurrent neural network for model reduction
• Physics-Based Generative Adversarial Models for Image Restoration and Beyond

Cited by in corpus (9)

• Deep Learning for One-dimensional Consolidation
• Long short-term memory embedded nudging schemes for nonlinear data assimilation of geophysical flows
• Information fusion and machine learning for sensitivity analysis using physics knowledge and experimental data
• Physics-constrained Deep Learning for Robust Inverse ECG Modeling
• Physics-informed deep learning for flow and deformation in poroelastic media
• Modeling Stochastic Microscopic Traffic Behaviors: a Physics Regularized Gaussian Process Approach
• A Data Scientist's Guide to Streamflow Prediction
• Monotonic Neural Network: combining Deep Learning with Domain Knowledge for Chiller Plants Energy Optimization
• Semi-supervised physics guided deep learning framework for predicting the I-V characteristics of GAN HEMT