loading page

Cross-attractor transforms: Improving forecasts by learning optimal maps between dynamical systems and imperfect models
  • Niraj Agarwal,
  • Daniel E. Amrhein,
  • Ian Grooms
Niraj Agarwal
University of Colorado Boulder

Corresponding Author:[email protected]

Author Profile
Daniel E. Amrhein
NCAR
Author Profile
Ian Grooms
University of Colorado Boulder
Author Profile

Abstract

Biased, incomplete models are often used for forecasting states of complex dynamical systems by mapping an estimate of a “true’ initial state into model phase space, making a forecast, and then mapping back to the “true’ space. While advances have been made to reduce errors associated with model initialization and model forecasts, we lack a general framework for discovering optimal mappings between the reference dynamical system and the model phase space. Here, we propose using a data-driven approach to infer these maps. Our approach consistently reduces errors in the Lorenz-96 system with an imperfect model constructed to produce significant model errors compared to a reference configuration. Optimal pre- and post-processing transforms leverage “shocks’ and “drifts’ in the imperfect model to make more skillful forecasts of the reference system. The implemented machine learning architecture using neural networks constructed with a custom analog-adjoint layer makes the approach generalizable to numerous applications.
05 Jun 2024Submitted to ESS Open Archive
06 Jun 2024Published in ESS Open Archive