By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
 The radio occultation technique is used to study plasma irregularities in the F2 layer ionosphere and below. The data, acquired by the GPS receiver onboard the geophysical research satellite CHAMP for more than three years, allow to study ionospheric irregularities under different geophysical conditions on a global scale. The dual-frequency GPS signals provide a possibility to trace irregularities with a spatial size ranging from about 20 to 250 km. Whereas the irregularities characterized by longer wavelengths dominate at low latitudes, the shorter scale perturbations are very intense at high latitudes. It is found that the high latitude irregularities occur mostly at night time when the ionization level is low.
 Ionospheric GPS radio occultation (IRO) measurements onboard LEO satellites can effectively be used for monitoring the ionospheric ionization on global scale [Hajj and Romans, 1998; Schreiner et al., 1999; Jakowski et al., 2002]. By measuring the phase differences between L1 and L2 GPS carrier frequencies the total electron content (TEC), i.e. the integral of the electron density along the ray path between GPS transmitter and LEO receiver, may be estimated with high accuracy. Because the noise level of the carrier phases is of order 0.1 radian, the relative TEC measurements achieve an accuracy of about 0.05 TEC units (TECU, 1 TECU = 1016 electrons/m2).
 In this study we analyze the vast database of IRO data sets, produced by the radio occultation experiment onboard the German geo-research satellite CHAMP [Reigber et al., 2000]. After three years we collected more than 120,000 occultation profiles [Jakowski et al., 2005].
 As shown by Hocke et al. , Pavelyev et al.  high-rate sampled (50 Hz) IRO measurements can be used effectively to detect spatial ionospheric structures caused by sporadic E layers. In this paper we will show that 1 Hz sampled radio occultation data can also contribute to the study of irregularities in the most dense F region.
 The occultation geometry is illustrated in Figure 1. While the LEO satellite with the GPS receiver onboard moves away from the GPS satellite, the connecting ray moves closer and closer to the Earth surface as indicated by the sequence of positions a–d of the tangent point in Figure 1. For further use, we define the “tangent point displacement” (TPD) as the path length which the tangent point has traversed (e.g., a–c in Figure 1).
 For reasons of simplicity we ignore ray path bending effects. Such simplification is possible if the height of the tangent point remains greater than 50 km.
2. Data Analysis
 Typically, a single occultation event produces a TEC profile, similar to curve a in Figure 2, where TEC is plotted against TPD. TEC first increases due to increasing of ionospheric electron density as well as width of the crossed spherical ionosphere layers. Subsequently the ray incidence angle becomes less oblique and the crossed width of dense layers lessens.
 But often, as curve b in Figure 2 shows, there are features in the TEC profile that obviously indicate plasma irregularities in the ionospheric layers. Since TPD values cover a range of 500–2500 km, it is practically impossible to resolve large-scale ionospheric structures with size above 500 km. However, as we show, the radio occultation technique provides a novel and very useful way to study medium- and small-scale irregularities where little research has been done so far.
 Of course, monitoring occultations with a single LEO satellite has some constraints such as the impossibility to locate exactly the disturbance region along the ray path. Also it is difficult to estimate the wavelength due to the unknown inclination of the wave fronts with respect to the ray. On the other hand, the 1 Hz sampled IRO measurements can achieve a vertical resolution of up to 2.5 km.
 To filter out ionospheric disturbances from TEC profiles, we need to get rid of the characteristic bell-shaped dependency of TEC (see Figure 2). To do this, we approximate the TEC profile with a smooth function and subsequently subtract it from the original profile. As several test calculations have shown, a nine point cubic polynomial approximation proves to be most reliable. Then we apply the Fourier transform to the difference between the experimental data and the approximation curve. Typical Fourier spectra of spatial TEC variations are shown on Figure 3.
 The long-wave part of this spectrum (wavelength >500 km) is formed more by traces of the bell-shaped dependency and by imperfections of the polynomial approximation than by the ionospheric structure. But these factors only slightly influence the middle part of the spectrum (100–250 km). The short-wave range (20–100 km) is almost completely free of them. The observed TEC variations are determined mainly by the plasma irregularities in the ionosphere between 150 and 400 km height where the most of the TEC stems from. Here we note that the true size of the irregularities usually differs from the size of their projections on the tangent point trajectory. Thus, because our estimation of the wavenumbers is semi-quantitative, we select two subranges with different wavelengths: 100–250 km for medium-scale and 25–50 km for small-scale disturbances (see Figure 3). For each of them we compute average TEC spectral density values.
3. Observations and Discussion
 The latitudinal distribution of the irregularities and their seasonal dependency is shown in Figure 4 resulting from the analysis of 46,000 occultations. In Figure 4 the spectral density is plotted against the averaged latitude of the occultation tangent point. One can see that the medium-scale disturbances (curve a) are found primarily in the equatorial region and, to a lesser degree, in the high-latitude region. The minima are located near 45° geographic latitude.
 As the double crest feature in Figure 4 (curve a) shows, the occurrence of the low latitude irregularities is closely related to the equatorial anomaly phenomenon. A more detailed analysis of the data indicate that they occur most often and most intense in a band along the magnetic equator and strongly intensify in the evening local time sector.
 More interesting is how the small-scale irregularities are distributed. The latitudinal distribution of this type of irregularities also shows a double-crest in the low-latitude zone which is obviously associated with the equatorial anomaly. In addition to this one can see two very well pronounced maxima in the polar caps. While the medium-scale irregularities distribution does not display any significant dependency on season, small-scale ones show an asymmetric summer-winter hemisphere behavior. The disturbances in the polar regions are very persistent, as they were observed during all three years of analyzed CHAMP observations. Surprisingly, the intensity of ionospheric irregularities in the winter hemisphere is somewhat (or even significantly) higher than in the summer hemisphere despite the much lower ionization in the polar night (see Figure 4).
 To get a better understanding of the occurrence of the small-scale irregularities at high latitudes, we created polar maps of their intensity. As we do not have sufficient data for producing a map at any given moment of time, we selected data obtained under similar solar illumination conditions and then averaged them. Thus for generating the map shown in Figure 5a we selected occultations from a four month period near the 2002 winter solstice, between 2200 UT and 0200 UT of every day. One can see that the activity is much higher on the night side. There are some high activity spots which are obviously associated with the auroral oval.
 The high wave activity in the polar region was previously studied by other techniques. Innis and Conde  reported high vertical winds in the polar regions, found by the Wind and Temperature Spectrometer on the Dynamics Explorer 2 satellite. They suggested that these winds are caused by atmospheric gravity waves (AGW) generated in the lower thermosphere. The AGW might be generated in the auroral oval region by electromagnetic effects like Joule heating and Lorentz forcing [Hocke and Schlegel, 1996]. The short-wave activity dependency from the geomagnetic Ap index derived from our data (not shown here) and a similar dependency discussed by Innis and Conde, supports this view. Of course, there are numerous physical mechanisms of the generation and motion of ionospheric irregularities. Particle precipitation, flux transfer events (FTE) and thermospheric upwelling [Lühr et al., 2004] play a role in disturbing the polar ionosphere. However, AGW can propagate far away from the region where they were generated, and most likely they cause most of the small-scale ionospheric irregularities observed.
 If the interpretation of these irregularities as the ionospheric response to AGW is correct, then it should be taken into account that the electron density only follows air density changes. Thus in case of high electron density even slight changes of the thermospheric density would cause strong deviations in the TEC. To estimate the intensity of neutral atmosphere waves, it is more effective to use a relative measure. This can be realized by normalizing the irregularity measure (spectral density) to the corresponding absolute TEC value. Consequently, all effects due to the background plasma density are removed.
 To estimate the absolute values of TEC we have compared the differential carrier phases with the code phase measurements of GPS which provide absolute but noisy TEC data as described by Heise et al. . The needed instrumental biases for TEC calibration were taken also from this paper.
 The results of relative spectral densities mapping are shown in Figure 6. First of all, the disturbance pattern is clearly associated with the auroral oval in the nighttime sector. The diurnal rotation is very pronounced: as it can be seen by the position of the terminator, the day-side activity drops while it peaks during night. The terminator lies approximately along the same isoline (0.2 km in units used in the figures) in all cases.
 This behavior can be naturally explained by the ion drag phenomenon, i.e. by the damping of thermospheric waves due to the interaction with the ionospheric plasma mobility of which is restricted by the geomagnetic field. The ion drag is strong where the electron density is high, and when ionization increases after sunrise, it effectively damps thermospheric gravity waves. It is worth to note that there is a sector with lower activity in Figure 6 after sunset where the electron density is still high. The dissipation of AGW due to thermospheric wind shear, which Innis and Conde believe to be the primary cause of wave extinction, also could play a role.
 It has been shown that GPS radio occultation measurements onboard LEO satellites provide a powerful tool for studying ionospheric irregularities on global scale. Key ionospheric parameters including the level of ionospheric perturbations for different spatial scales may easily be monitored globally. The occultation measurements bear great potential for detecting and analyzing the spatial structure of plasma irregularities and are able to provide new findings for an improved understanding of the ionospheric perturbations.