Differentiable Programming for Earth System Modeling

Abstract

Earth System Models (ESMs) are the primary tools for investigating future Earth system states at time scales from decades to centuries, especially in response to anthropogenic greenhouse gas release. State-of-the-art ESMs can reproduce the observational global mean temperature anomalies of the last 150 years. Nevertheless, ESMs need further improvements, most importantly regarding (i) the large spread in their estimates of climate sensitivity, i.e., the temperature response to increases in atmospheric greenhouse gases, (ii) the modeled spatial patterns of key variables such as temperature and precipitation, (iii) their representation of extreme weather events, and (iv) their representation of multistable Earth system components and their ability to predict associated abrupt transitions. Here, we argue that making ESMs automatically differentiable has huge potential to advance ESMs, especially with respect to these key shortcomings. First, automatic differentiability would allow objective calibration of ESMs, i.e., the selection of optimal values with respect to a cost function for a large number of free parameters, which are currently tuned mostly manually. Second, recent advances in Machine Learning (ML) and in the amount, accuracy, and resolution of observational data promise to be helpful with at least some of the above aspects because ML may be used to incorporate additional information from observations into ESMs. Automatic differentiability is an essential ingredient in the construction of such hybrid models, combining process-based ESMs with ML components. We document recent work showcasing the potential of automatic differentiation for a new generation of substantially improved, data-informed ESMs.

arXiv comment17 pages, 2 figures

References in corpus (21)

• Deep learning to represent sub-grid processes in climate models
• Machine learning accelerated computational fluid dynamics
• Firedrake: automating the finite element method by composing abstractions
• Enforcing Analytic Constraints in Neural-Networks Emulating Physical Systems
• Earth System Modeling 2.0: A Blueprint for Models That Learn From Observations and Targeted High-Resolution Simulations
• From calibration to parameter learning: Harnessing the scaling effects of big data in geoscientific modeling
• Will Artificial Intelligence supersede Earth System and Climate Models?
• Deep Learning for Physical Processes: Incorporating Prior Scientific Knowledge
• Automated derivation of the adjoint of high-level transient finite element programs
• Least Squares Shadowing sensitivity analysis of chaotic limit cycle oscillations
• Stiff Neural Ordinary Differential Equations
• Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers
• A Differentiable Programming System to Bridge Machine Learning and Scientific Computing
• JAX-FLUIDS: A fully-differentiable high-order computational fluid dynamics solver for compressible two-phase flows
• Achieving Conservation of Energy in Neural Network Emulators for Climate Modeling
• Use of neural networks for stable, accurate and physically consistent parameterization of subgrid atmospheric processes with good performance at reduced precision
• Sensitivity analysis on chaotic dynamical system by Non-Intrusive Least Square Shadowing (NILSS)
• Predicting Climate Change using Response Theory: Global Averages and Spatial Patterns
• JAX-COSMO: An End-to-End Differentiable and GPU Accelerated Cosmology Library
• Response and Sensitivity Using Markov Chains
• Toward a chaotic adjoint for LES

Cited by in corpus (4)

• Neural General Circulation Models for Weather and Climate
• Explainable Offline-Online Training of Neural Networks for Parameterizations: A 1D Gravity Wave-QBO Testbed in the Small-data Regime
• Reconstructing Historical Climate Fields With Deep Learning
• Constructing Extreme Heatwave Storylines with Differentiable Climate Models