Variable coupling between the bottomside and topside thickness of the ionosphere

Data of foF2, hmF2 and h'F deduced from ionograms, and GPS-derived TEC recorded at 24 observatories for the solar minimum (April 1995) and solar maximum (April 2002) are analyzed to explore variations of the bottomside and topside thickness of the ionosphere. Monthly median parameters represent quiet space weather conditions, and daily-hourly values are used during the storms. Reconstruction of the electron density distribution is performed with the International Reference Ionosphere extended up to the GPS satellite height at 20000 km, IRI*, matched to the observations. The Kriging mapping technique has been applied for the spatial interpolation of the multi-stations ionosonde and GPS-TEC data at the area of latitudes of 35-70 degrees N, and longitudes of -10 to 40 degrees E. An analytical model of superposition of two 3rd degree polynomials is developed to represent IRI bottomside thickness B0 via two variables-the shape parameter B1 and the ratio of Ne/NmF2 for any given profile anchor point below the F2 peak. The real height calculated from the minimum virtual height h'F allows a determination of the semi-thickness of the bottomside ionosphere (r(bot)). TEC input in IRI* is used for selecting the topside profile semi-thickness (r(top)) minimizing the difference between the model TEC and GPS-derived TEC. We have found that the topside thickness increment for the stormy conditions is opposite to the bottomside thickness. The topside semi-thickness to bottomside semi-thickness ratio, r(top)/r(bot), is growing towards larger foF2. This ratio was 2.5-0.5 times for quiet median conditions and 8-0.5 times for ionospheric storm. The extreme low ratios less than I during daytimes refer to the cases when the topside semi-thickness is equal to or less than the bottomside semi-thickness. A numerical model is derived to give the close approximation to the relation between the topside and bottomside ionosphere thickness giving smooth variations as a function of local time, solar and geomagnetic activity, and the foF2 critical frequency. (c) 2006 Elsevier Ltd. All rights reserved.