A southern hemisphere error in the IRI

I would like to call attention to the fact that the IRI computes erroneously the F2-layer semithickness parameter B₀ at southern hemisphere locations. The values of B₀, based on northern latitude observations, have a seasonal variation which must be preserved at southern latitudes.The error was found in the course of a study to develop a new ionospheric model for radio-propagation predictions. We observed at southern latitudes major discrepancies between the IRI and the Bradley-Dudeney (1973) model in relation to the F2-layer semithickness. This is estimated in the latter model as the difference between the height of maximum electron concentration (hmF2) and the ionospheric characteristic h′F,F2, the minimum observed virtual height of reflection from the F2-layer, corrected taking into account underlying ionization. The profiles for both models were drawn using the same values of foF2 and hmF2. Then, our analysis served also to test the IRI model with h′F,F2 data obtained from CCIR maps but not used as primary inputs by the IRI.