Perspectives of using physics-based models for forecasting of total electron content of the ionosphere

Abstract:

We will use new results from an evaluation of physics-based ionosphere-thermosphere models (TIE-GCM, CTIPe and GITM) prediction capability to reveal key challenges of using physics-based models for ionosphere-thermosphere forecasting with multi-day lead times. Our findings examine the True Skill Statistic (TSS) metric for the prediction of total electron content (TEC), noting that this is but one, yet important, metric to assess predictive skill. The meanings of contingency table elements for the prediction performance are analyzed in the context of ionosphere modeling. We will evaluate the prediction metric for different magnitudes of ionospheric TEC decreases and increases using data-derived benchmarks. TEC predictions made with physics-based modeling of storm dynamics and with assuming non-storm dynamics are compared. We will explore uncertainties of current physics-based modeling for TEC prediction and its sensitivity to forecastable drivers. Our approach provides quantitative information on the degree to which physics-based modeling is useful in an ambitious prediction scenario.