Speaker
Description
Building on the methods established in Edward-Inatimi et al. (2024), we calibrate an ambient solar-wind ensemble driven by the Wang-Sheeley-Arge (WSA) model. Ensemble methods are powerful tools which allow forecast uncertainty to be better characterised. Using a coupled coronal-heliospheric modelling approach we generate ensembles using the WSA model, used operationally, with the Heliospheric Upwind eXtrapolation with time dependence (HUXt) solar-wind model. We demonstrate how the levels of perturbation applied to the inner-boundary have a large impact on different aspects of forecast performance and that improvements in forecast reliability, the ensemble's capability to generate realistic probabilities, often results in a trade-off in forecast resolution, the ability of the ensemble to meaningfully distinguish between events beyond climatology. Using well-known ensemble and forecasting metrics, we refine a calibration procedure which optimises the scale of perturbation applied to the inner-boundary conditions which results in an optimal balance of reliability and resolution. We provide optimal values of the perturbation parameters for use with WSA-HUXt.
| Do you plan to attend in-person or online? | In-person |
|---|
