Ensemble modeling of Coronal Mass Ejection dynamics and forecasts at 1 AU with a semi-analytic flux-rope model

TL;DR

This paper uses a semi-analytic flux-rope model within a Monte Carlo framework to quantify how uncertainties in CME and solar wind inputs affect forecast diagnostics at 1 AU, highlighting key parameters influencing prediction accuracy.

Contribution

It introduces an ensemble modeling approach that incorporates sheath effects and updated drag forces to better understand CME forecast uncertainties from eruption to 1 AU.

Findings

Time-of-arrival spread is 2.4-7.7 hours, mainly influenced by flux injection history.

Leading-edge speed spread is 28-53 km/s, primarily controlled by background solar wind.

Magnetic field prediction uncertainty varies, with sheath fields more tightly constrained than internal flux-rope fields.

Abstract

This study quantifies how uncertainty in physically meaningful coronal mass ejection (CME) and solar-wind inputs propagates into forecast-relevant diagnostics from eruption to 1 AU. We use a semi-analytic erupting flux rope (EFR) model to simulate CME initiation and Sun-to-1 AU propagation under Lorentz, gravitational, and drag forces, driven by a prescribed time-dependent poloidal-flux injection. Relative to the original EFR formulation, we include sheath and pile-up effects through an effective mass and update the drag term for CME solar-wind coupling. The model is embedded in a Monte Carlo framework with truncated-normal sampling of key eruption and background solar-wind inputs. Across six CME events, the ensembles show event-dependent dispersion in the 1 AU diagnostics. For +/- 20% input sampling, all spreads are 1-sigma ensemble standard deviations. The time-of-arrival spread is 2.4-7.7 h and is mainly controlled by the poloidal-flux injection history, upstream wind speed, and drag coefficient. The leading-edge speed spread is 28-53 km/s and is primarily controlled by background-flow properties. Magnetic-field diagnostics show two regimes: the sheath field is relatively tightly distributed, with a spread of 1-3.5 nT and sensitivity to upstream wind, size, and expansion scaling, whereas the internal flux-rope field has a larger spread of 1-7.6 nT and is governed mainly by eruption-driving and flux-content parameters. The impact-duration spread is 2.4-6.3 h and is controlled mostly by geometric size and expansion scaling, with additional sensitivity to the driving timescale. These results establish a quantitative link between EFR input uncertainties and the resulting spread in CME arrival and impact diagnostics, identifying the physical parameters that most strongly limit forecast precision at 1 AU.