Uncertainty Quantification and Global Sensitivity Analysis for CME simulations in the SWMF

The Space Weather Modeling Framework (SWMF) offers efficient and flexible Sun-to-Earth simulations based on coupled first principles and/or empirical models. This encompasses computing the quiet solar wind, generating a coronal mass ejection (CME), propagating the CME through the heliosphere, and calculating the magnetospheric impact via geospace models. However, useful long-term predictions of space weather events are affected by uncertainty and variation in many model parameters and inputs, and require systematic uncertainty quantification (UQ) and data assimilation (DA). In this work, as part of the NextGen SWMF project funded by the NSF SWQU program, we discuss UQ results for CME simulations. Prior work used Global Sensitivity Analysis to identify the most influential parameters for background solar wind simulations under solar maximum and minimum conditions. Then, Bayesian DA was performed to generate a small ensemble of background wind solutions from the posterior distribution. Next, for launching a CME, we include new uncertain flux rope parameters such as size, orientation, field strength etc. We propose new experimental designs on these parameters for each of the best-performing background runs. CME simulations are then conducted for these designs using the Alfvén Wave Solar atmosphere Model (AWSoM) for different events such as the September 2014 and March 2015 eruptions. Global Sensitivity Analysis is performed on predictions of radial velocity (Ur) and number density (Np) at 1 au to identify the most important flux rope parameters. Further, we also quantify the uncertainty in trajectory of the modeled CME by sampling time histories of the predicted quantities Ur, Np and Bz component of magnetic field at different locations around the Earth. These are compared with in-situ observation data at L1 as well as steady-state background quantities sampled at different latitudes prior to the CME eruption.