the bottom F‐region patch density is smaller than that of the surrounding region in the noon sector, these observations suggest that the major plasma source for patches is the dayside solar EUV produced plasma transported into the polar cap region, which is consistent with the statistical UT dependence found in both modeling (Sojka et al., 1994) and observation (David et al., 2016). However, we also need to realize that the RISR‐C is deep in the center of the polar cap, where it takes time for the patch to drift from the dayside cusp region to reach, and the electron temperature can change along the trajectory. Therefore, a careful numerical modeling study is required to quantify the cooling rate inside the patch and evaluate whether the electron temperature is a good and sufficient indicator of the origin of the patch plasma.
Although statistically the polar cap patches have lower electron temperature than that of the surrounding regions, “hot” patches, which have higher electron temperature than the surrounding regions, have also been identified using DMSP observations at ~850 km (Zhang et al., 2017). It is found that this type of patch is associated with local particle precipitation and convection flow shear, and might be produced when the traditional patches convect into the particle precipitation region.
4.4 DYNAMIC EVOLUTION OF POLAR CAP PATCHES
Polar cap patches are density enhancements in the F region and topside ionosphere, and thus they should convect in the polar region at the ExB convection velocity. Thomas et al. (2015) carefully compared the patch drifting speed obtained from 630 nm airglow images and that measured by the SuperDARN radars. They confirmed that the horizontal motion of the optical patches is consistent with the background plasma convection, and thus patches can be used as tracers of polar cap convection. In their case, the convection flow speed within the patch exceeded 500 m/s. Because the ionospheric convection is controlled by the solar wind and IMF condition, as well as magnetospheric dynamics (in particularly on the nightside), the transport of patch across the polar cap can thus be of considerable complexity, including substantial rotation as observed in Oksavik et al. (2010).
Optical images of patches provide clues of 2‐D horizontal structure of the patch. Hosokawa et al. (2010) provide a good example of using the optical signature of patches as a tracer for mesoscale plasma convection flow. They combined observations from a redline imager located at Resolute Bay with plasma convection flow estimates derived from SuperDARN line‐of‐sight plasma velocity measurements. In their case study, they showed the bifurcation of a patch as it transited the polar cap. The bifurcation occurred as the patch moved into a region where the plasma convection field diverged. However, the number of imagers is limited and they can provide images only under certain conditions, such as in the dark and under clear sky. Furthermore, the imagers are biased to patches that are at a sufficiently low enough altitude to allow for the chemistry to generate the observed emissions. Perry and St. Maurice (2018) provide evidence confirming the notion that higher altitude patches generate dimmer emissions than lower altitude patches. Thus, it is conceivable that high‐altitude patches and their dim optical signatures may go undetected by some imagers.
The fast‐growing number of GNSS receivers in the high‐latitude regions provides another unprecedented opportunity to reveal the horizontal morphology of polar cap density structures and to visually track their dynamic motion on a regular basis. Taking advantage of the fact the patches can be used as convection flow tracer, studies have been trying to use it to infer the underlying coupling processes between the magnetosphere and ionosphere (Nishimura et al., 2014; Zhang et al., 2013a; Zou et al., 2015; Zhang et al., 2016a). For example, using the 2‐D GPS TEC maps with superposition of the SuperDARN convection patterns (Zhang et al., 2013a) reported direct observations of the full evolution of patches during a geomagnetic storm, including entering the polar cap from the dayside cusp and turning into a blob on the nightside after exiting the polar cap. Therefore, through observing the life cycle of a patch in the ionosphere, the timescale for the global magnetospheric convection can also be inferred. A follow‐up study by (Zhang et al., 2015) using global TEC maps revealed that a complete magnetospheric convection cycle, that is, Dungey cycle, takes about 3 hours.
While the number of ground‐based GNSS receivers continues to increase and TEC maps of better spatial resolution can be obtained, their imaging capability is limited by the inhomogeneous distribution of the continents. This limitation can be improved by using tomographic reconstruction and data assimilation techniques (e.g., Aa et al., 2016, 2018; Bust & Datta‐Barua, 2014; Bust & Mitchell, 2008; Gardner et al., 2014, 2018; Mannucci et al., 1998; Yin et al., 2008; Yin et al., 2017), which are able to take in heterogeneous ionospheric data sets and combine with a physics‐based or empirical models to best describe the ionosphere density and TEC distributions. For instance, by combing measurements from ground‐based GNSS receivers and LEO satellites, such as COSMIC (Constellation Observing System for Meteorology, Ionospheric, and Climate) and GRACE (Gravity Recovery And Climate Experiment), Yue et al. (2016) obtained the global ionospheric electron density and TEC with the spatial/temporal resolution of 5° in latitude, 10° in longitude, ~30 km around the F2 peak, and 0.5 h in time, during the 17 March 2013 storm. Figure 4.6 shows the comparison between the original ground‐based GNSS TEC map and the reconstructed map combining both ground and space‐based measurements. SED and TOI signatures are clear in both plots, but the reconstructed map is able to provide a much more complete picture. This type of method has a huge potential for filling in the large gaps over ocean and would be of great use for studying the ionospheric density distributions on a global scale. For example, during this storm, this method was able to reveal two separate SED/TOI structures in the conjugate hemispheres. In addition to these tomographic reconstruction and data assimilation techniques, new statistical analysis of these TEC maps also reveals new features of the SED/TOI evolution. Yang et al. (2016) statistically analyzed the exit of the SED/TOI plasma and found that they preferentially exit the nightside polar cap in the premidnight sector but join the dawn convection cell, forming a "hook‐like" high‐density pattern.
Figure 4.6 (a) TEC map of the Northern Hemisphere using ground‐based receivers only; (b) TEC map results from the reanalysis by combining multiple data sources together
(from Yue et al., 2016; Reproduced with permission of John Wiley and Sons).
In the polar cap region, the Earth's magnetic field is nearly vertical. Thus, the ExB convection flow and the horizontal thermospheric wind cannot efficiently lift or descent the F‐region density structures. Indeed, using the TIEGCM model, Liu et al. (2016) studied the 2015 St. Patrick's Day storms and found that the horizontal transport due to the E×B ion drift plays an essential role in moving plasma from the dayside convection throat region to the polar cap to form plume/TOIs. Similar results have been obtained using the Global Ionosphere and Thermosphere model (GITM) in Zou and Ridley (2016). In their study, SED plume/TOIs in the polar cap region have been traced backward to the subauroral region during the 24–25 October 2011 geomagnetic storm. Simulation results suggest that plasma originating from both the dawn and dusk sectors is able to contribute to the formation of the SED plume in the model. However, the ionospheric plasma originating from different local time sectors can have different properties in terms of the F‐region peak density NmF2 and the peak height hmF2. Plasma originating from the dawn sector exhibits slow but steady increase in hmF2, NmF2, and TEC as the flux tube drifting from lower to higher latitudes, while those from the dusk sector experience more steep increase and decrease in hmF2 and TEC because of larger convection flow variations. This suggests that the hmF2, together with the TEC value, might provide a means for understanding the origin of the plasma that contributes to formation of the SED plumes.
Several mechanisms have been proposed for segmenting SED plume/TOI entering the polar cap into a patch