Variational Data Assimilation with a Learned Inverse Observation Operator
This work addresses a bottleneck in operational weather forecasting systems by enhancing optimizability, though it is incremental as it builds on existing variational data assimilation methods.
The paper tackles the difficult optimization problem in variational data assimilation for forecasting chaotic systems by learning a mapping from observational data to physical states, which improves forecast quality in experiments on the Lorenz96 model and a two-dimensional turbulent fluid flow.
Variational data assimilation optimizes for an initial state of a dynamical system such that its evolution fits observational data. The physical model can subsequently be evolved into the future to make predictions. This principle is a cornerstone of large scale forecasting applications such as numerical weather prediction. As such, it is implemented in current operational systems of weather forecasting agencies across the globe. However, finding a good initial state poses a difficult optimization problem in part due to the non-invertible relationship between physical states and their corresponding observations. We learn a mapping from observational data to physical states and show how it can be used to improve optimizability. We employ this mapping in two ways: to better initialize the non-convex optimization problem, and to reformulate the objective function in better behaved physics space instead of observation space. Our experimental results for the Lorenz96 model and a two-dimensional turbulent fluid flow demonstrate that this procedure significantly improves forecast quality for chaotic systems.