We compare ionospheric parameters including total electron content (TEC), peak density, and height of the F2 layer (NmF2 and hmF2) between FORMOSAT-3/COSMIC (F3/C) GPS radio occultation (RO) technique retrieved and International Reference Ionosphere (IRI-2001) model predicted during different seasons in a low solar activity (LSA) year 2007. The comparison of topside TEC (tTEC, obtained by integrating electron density from hmF2 to 800 km) between IRI and F3/C shows that the IRI overestimates tTEC during both equinox seasons at around +/- 15 degrees magnetic latitudes during daytime, especially in March equinox while underestimation of tTEC predicted by IRI is a dominant feature for both solstices. Further, a common finding is that the IRI overestimates tTEC from evening to prenoon hours irrespective of season at around +/- 20 degrees magnetic latitudes, which is most likely due to the underestimation of hmF2 (by around 30-40 km), a key parameter on which the build of electron density profile depends in the model and inaccurate representation of the real profile by the topside electron density profiler model in IRI. The global distributions and seasonal variations of NmF2 show clear semiannual and annual asymmetry features during daytime and such features also reflected in IRI predicted peak densities with few exceptions. This high degree of agreement between tTEC comparison and the characteristic features of NmF2 global distributions indicates that the majority of contribution for tTEC has come from the F2 region. Further, it is also presented the global distributions of topside vertical scale heights (VSH) computed using electron density profiles retrieved from F3/C and predicted by IRI model during different seasons in year 2007. An important finding is that the topside VSH of F3/C profile data is meticulously following the geomagnetic equator during daytime irrespective of season and tends to increase toward higher latitudes. An appreciable latitudinal difference is found in the season averaged scale heights that derived with F3/C data during daytime while completely opposite results are found for IRI predicted scale heights. The discrepancies of topside VSH between the F3/C derived and predicted by IRI indicating that the shape of the topside electron density profile in the IRI model should desperately be revised accordingly such that it more closely resembles the real situation.
