Comment on “A New Aspect of Ionospheric E Region Electron Density Morphology” by Yen‐Hsyang Chu, Kong‐Hong Wu, and Ching‐Lun Su
Jiuhou Lei,Xinan Yue,William Schreiner
DOI: https://doi.org/10.1029/2009ja015234
2010-01-01
Journal of Geophysical Research
Abstract:[1] Chu et al. [2009] have recently reported a new aspect of ionospheric E region electron density morphology from the Constellation Observing System for Meteorology, Ionosphere and Climate (COSMIC) radio occultation (RO) measurements. They claimed that “…there are three geomagnetic latitude regions where striking enhancements of the E region electron density occur…The off-equator E region electron density enhancements are closely connected with the bottomside of the F region equatorial anomaly crests, where the component of the electron density parallel to the magnetic field line is maximum…” However, it has been long recognized that the variations of the E region electron density are dominated by the photochemical processes [Rishbeth and Garriott, 1969], and thus the changes of the E region electron density are mainly governed by the solar zenith angle rather than the magnetic field line. As discussed subsequently, the so-called “new” phenomenon in the equatorial E region described by Chu et al. [2009], which shows the imprint of the equatorial ionization anomaly (EIA) in the F region, is potentially associated with the retrieval error of RO. [2] Recently, Yue et al. [2010] carried out a simulation study to investigate the error distribution of RO electron density profiles retrieved from the Abel inversion. Occultation events observed by the COSMIC satellites are simulated during equinox by calculating the integrated total electron content (TEC) along the COSMIC occultation paths with the “true” electron density from an empirical model. The retrieval errors are then calculated by comparing the retrieved density profiles with the “true” profiles. The details about the simulation are given by Yue et al. [2010]. As shown in Figure 1, the retrieved electron density underestimates the true electron density in the region surrounding the EIA crest (±10°–30° latitude) while overestimating near the equator (±10°) and in the north and south of the EIA crests (±30°–50°). Two artificial plasma depletions are seen clearly underneath the EIA peaks in the retrieved electron density (Figure 1b) but not in the true electron density (Figure 1a). These features in Figure 1b are reflected in the COSMIC data [see Chu et al., 2009, Figure 12]. As a result, three obvious peaks and two troughs are present in the absolute retrieval error along latitude, especially in the E region, albeit only one peak in the “true” E region electron densities located at the equatorial region (Figure 1a). Thereby, three peaks in daytime E region electron densities along the magnetic latitude seen in the COSMIC data [see Chu et al., 2009, Figures 1, 6, and 9] are resulting from the RO Abel inversion mainly due to the spherical symmetry assumption used in the retrieve [e.g., Lei et al., 2007; Schreiner et al., 1999]. [3] In addition, ionospheric horizontal gradients in the zonal direction also can result in errors in the RO retrieval process [Yue et al., 2010]. The GPS ray may go through different local time zones for a given universal time. In the same way, an artificial wave 4 structure along the longitude direction is expected to occur in the E layer densities at the constant local time frame as long as the wave 4 structure is present in the F2 region ionosphere [Lin et al., 2007, and references therein]. In this case, the longitudinal variation in the E region density will be out of phase with that in the F2 region, although in reality the nonmigrating tides propagating from lower atmosphere may result in the longitudinal variation in neutral atmosphere in the E region and then reflect in the E region electron density. Thus, the longitudinal variation of E layer electron density presented by Chu et al. [2009, Figures 9–11 and 13] is also contaminated by the effect of RO retrieval errors. [4] In closing, several of the principal conclusions of Chu et al. [2009] are incorrect, given that these authors are not aware of the limitation of the retrieval RO electron density profiles at low latitudes and low altitudes. However, in their original reply, Chu et al. [2010] insisted that their results are valid because their comparison between the ionosonde NmE and COSMIC observations seems to be inconsistent with the simulations of Yue et al. [2010]. To assess their statement, we computed the relative deviation of noontime NmE between the global ionosonde data and COSMIC observations. Our results in Figure 2 are in agreement with the simulations (see Figure 1) qualitatively, except in the equatorial region where ionosonde observations are spare. Although Chu et al. [2010] also indicated a larger error over the subtropical regions, they showed that the COSMIC-retrieved NmE is systematically larger than that from ionosonde measurement at all latitudinal zones; this is probably due to the fact that they discarded a larger portion of the COSMIC profiles before the ionosonde-COSMIC comparison. Additionally, the results in Figure 2 of Chu et al. [2010] have no coherent pattern to support their argument that electron density enhancements over the off-equator E region are associated with plasma diffusion from the EIA in the F region. [5] X. Yue and W. S. Schreiner received support under NSF grant 0723439. The ionosonde data used for this analysis are available in the American National Geophysical Data Center (NGDC-NOAA) database. [6] Bob Lysak thanks the reviewer for assistance in evaluating this paper.