Development of a Regional F‐Region Critical Frequency Model for Southern Africa

A new localized linear regression model, called the SANSA model, is presented. The model provides a forecast of the F‐region critical frequency (foF2) over Grahamstown, South Africa (33.31°S, 26.52°E). The input parameters of this model are the maximum daily solar elevation angle, the F10.7 cm radio flux index and the planetary K index. The output parameter is the forecasted foF2. The historical data‐set consists of ionosonde foF2 observations made in Grahamstown for the period 2010–2019. The historical and predicted data sets used to develop the model are obtained from databases maintained by the South African National Space Agency, the National Aeronautics and Space Administration and the National Oceanic and Atmospheric Administration. The model improves on the results of the Ionospheric Communications Enhanced Profile Analysis and Circuit Prediction Program in terms of the root mean square error skill score metric. The work presented shows that a linear regression model based on the solar elevation angle as well as the F10.7 and planetary K indices can closely approximate observed foF2 magnitudes during quiet and storm times. Plain Language Summary The F‐region critical frequency (foF2) is the highest frequency that a signal can be transmitted, from the ground, and still be reflected off the lower ionosphere back to the ground. Signals with frequencies above the foF2 will penetrate the ionosphere and propagate into space, thus all information coded within that signal will be lost to space. The work presented here aims to demonstrate that a foF2 forecast model can be developed by combining the F10.7 cm radio flux index, local maximum daily solar elevation angle and the planetary K‐index. We based our model on analyzing the historical foF2 variation measured between 2010 and 2019 above Grahamstown, South Africa. The historical and predicted data sets used to develop the model are obtained from databases maintained by the South African National Space Agency, the National Aeronautics and Space Administration and the National Oceanic and Atmospheric Administration. The proposed model improves on the results of the Ionospheric Communications Enhanced Profile Analysis and Circuit Prediction Program in terms of the root mean square error skill score metric. Key Points We present a methodology for deriving a regional foF2 model for Grahamstown, South Africa The model used hourly foF2 ionosonde data measured over a decade, the F10.7 index, maximum daily solar el