Abstract
The inversion of electron density from total electron content (TEC) measurements of GPS radio occultation is investigated by means of simulated data from the International Reference Ionosphere IRI-2001 and observations by the GPS/MET satellite experiment. In both cases a meridional slice of electron density is derived for northern summer solstice June/July 1995 from the midnight to the noon sector of the Earth's ionosphere. By means of the simulated occultation data a new 2-D recovery method is tested considering electron density variations along the ray path through a non-spherical ionosphere. This method is as fast as the Abel inversion. The relative retrieval error is less than a few percent around and beyond the F2-layer peak. The resolution (in latitude) of the 2-D recovery method is significantly better than those of the Abel inversion which assumes spherical symmetry of the ionosphere along the ray path. After this simulation test, the 2-D recovery method and the Abel inversion are applied to the noon-midnight GPS/MET data in June/July 1995, near to solar minimum. The advantages of the 2-D recovery method are in case of the GPS/MET observations questionable. This could be due to data gaps and TEC errors disturbing the 2-D recovery method more than the robust Abel inversion. Finally meridional slices are derived for other local times by the Abel inversion. Because of missing data and uncertainty of possible retrieval errors the discussion of the diurnal and global variations of the F region is confined to the most significant features. Clear departures between GPS/MET and IRI are found for the polar winter ionosphere and the nighttime topside ionosphere at low latitudes. Generally the GPS/MET observations and the IRI predictions agree well for most local times and latitude regions.
