Speaker
Description
Accurate three-dimensional (3D) characterization of coronal mass ejections (CMEs) is essential for modelling their propagation through interplanetary space and forecasting their arrival time at Earth. However, forecasting accuracy, assessed through platforms such as the Community Coordinated Modeling Center (CCMC) CME Scoreboard, has shown minimal improvement over the past decade, with persistent mean absolute errors around 13 hours. Several studies underscore fundamental issues in model inputs, notably uncertainties in CME parameter characterisation from coronagraph data as well as lack of knowledge of the solar wind conditions through which the CME propagates.
Current operational forecasting typically employs simplified morphological models such as the cone or Graduated Cylindrical Shell models, often relying on subjective manual fitting methods that underestimate true uncertainties, introduce user biases, and complicate statistically robust ensemble creation. Recent analyses reveal substantial variability in derived parameters depending on user input or viewpoint availability.
To address these challenges, we introduce a 3D CME cone model fitting framework using Bayesian inference, significantly reducing subjectivity by rigorously quantifying both observational and model uncertainties in the parameter space. Unlike traditional methods yielding single best-fit solutions, Bayesian inference provides a comprehensive posterior distribution for CME parameters, rigorously quantifying uncertainties and parameter correlations.
Using our framework, we investigate how parameter uncertainties and correlations change when expanding from a single viewpoint to two viewpoints. Using our posterior distribution, we model ensembles of CMEs using the HUXt model and compare the forecasted time of arrival distributions to more ad-hoc methods that do not account for parameter correlations. Additionally, the posterior distributions could offer informed priors crucial for data assimilation methods incorporating heliospheric imager (HI)-like observations, particularly valuable once missions such as Vigil become operational.
By making our framework's open-source code available, we aim to promote the broader adoption of uncertainty-aware CME characterization, addressing critical gaps identified in contemporary forecasting methodologies.
| Do you plan to attend in-person or online? | In-person |
|---|
