Data-based optimization of a simple shortwave fadeout absorption model

Electron density enhancement caused by electromagnetic radiation emitted during a solar X-ray flare has the potential to increase high frequency (HF; 3–30 MHz) absorption in the dayside D-region ionosphere, impacting shortwave radio signals by reducing the signal strength, a phenomenon commonly referred to as shortwave fadeout. Data-based optimization of a simple absorption model is performed incorporating solar X-ray flux data and 30 MHz riometer data from stations distributed across Canada. In a single event study the data-based optimization model is shown to overestimate absorption by 1% for the duration of an X2.1 solar X-ray flare. This corrects an underestimation by the NOAA D-region Absorption Prediction (D-RAP) model. In a statistical study, based on 87 events, data-based optimization performed on an event-by-event basis showed excellent overall agreement between measured and modelled data: the Pearson correlation coefficient was R = 0.88, and the slope of the best-fit line to the data was m = 0.91. A generalized model was developed using data from all 87 events collectively. Although good agreement was found between the measured and modelled data sets, correlation and slope were slightly reduced to R = 0.75 and m = 0.80. Model accuracy is characterized by prediction efficiency (PE) which peaked at PE = 0.78 for the event-by-event evaluation and PE = 0.48 for the collective data set. The results of this study highlight the advantages of data-based optimization in modelling absorption due to shortwave fadeout.