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.