counterintuitive, but as shown in Liu et al. (2016) and Zou et al. (2014), the evolution of SED plume or TOI depends on the interplay between the convection electric field and thermospheric winds. There are cases in which the SED plume/TOI do not extend into the polar cap during geomagnetic storms. This is further complicated by the fact that global‐scale thermospheric composition change with increased molecular species can occur during the negative storm phase and, thus, extremely low ionospheric densities may occur while the geospace is still under perturbed condition. Therefore, the relationship between TOI/patch and geomagnetic activities may not be described by a simple linear relation.
It has been well known that the IMF direction and magnitude largely control the ionospheric convection pattern and their variations can segment large‐scale high‐density structures into smaller‐scale patches (Anderson et al., 1988; Lockwood & Carlson, 1992; Rodger et al., 1994; Valladares et al., 1996, 1998; Zhang et al., 2013a). Studies have been performed trying to understand the patch occurrence rate dependence on the IMF conditions. Spicher et al. (2017) found that patches occur more often in the Northern Hemisphere postnoon/prenoon sector for negative/positive By condition, while the trend is mirrored in the Southern Hemisphere. This result is consistent with the cusp location dependence on the IMF By, confirming that the dynamics in the cusp region is responsible for the patch segmentation. The superposed epoch analysis carried out by Noja et al. (2013) shows that enhanced IMF Bz preceded the patches, suggesting that enhanced convection is important for the patch formation. In the Jin et al. (2018) paper, ESR was selected to observe patches within 3 hours surrounding the noon MLT in order to minimize the time between their formation near the dayside cusp and their detection at ESR. This study confirmed the preference of patch formation during southward IMF Bz, and also revealed the IMF By influence on the patch location.
In this section, we briefly reviewed the recent results about the statistical occurrence rate of polar cap patches and its dependence on season, UT, geomagnetic activity, and IMF conditions. In the Southern Hemisphere, the seasonal occurrence rate of patch differs depending on its identification mechanisms, whether it is based on in situ density measurement or integrated TEC. This discrepancy suggests that caution is needed when identifying the patches using the traditional doubling electron density method. Criteria reflecting that patch is a high‐density structure should be included as well, such as a requirement of the patch density being higher than the average background density. Also, the occurrence rate of TOI/patch shows no clear relationship with the geomagnetic activity level indicated by Kp. More detailed analysis is needed in the future to take into consideration the storm phases and thermosphere composition changes. In addition, further statistical studies are needed to understand the most probable IMF conditions right at the time when the patches are produced at the dayside cusp region, in order to single out the major segmenting mechanism of patch.
4.3 PLASMA CHARACTERISTICS WITHIN THE POLAR CAP PATCHES
In the last decade, two advanced modular incoherent scatter radars (AMISRs) have been installed deep in the polar cap at Resolute Bay, Canada, named Resolute Bay ISR‐North face (RISR‐N) and Canadian face (RISR‐C). These two ISRs provide new opportunities for in‐depth investigation of the patch plasma characteristics, such as altitude profiles of key plasma parameters (e.g., Dahlgren et al., 2012a and 2012b; Gillies et al., 2016; Lamarche & Makarevich, 2017; Perry & St. Maurice, 2018; Ren et al., 2018).
Using a special 25‐beam imaging mode of RISR‐N, Dahlgren et al. (2012a,b) revealed, for the first time, the 3‐D density structure of a patch and its temporal evolution. Figure 4.4 shows a volumetric image of the patch studied in Dahlgren et al. (2012b) with measured 630 nm redline emission shown at the bottom. They identified up to 10% density variability even though the patch acted as a closed system with no additional plasma transported horizontally into the patch. A comprehensive discussion is provided in Dahlgren et al. (2012a) trying understand the source of this density variation, including field‐aligned motion and local precipitation, but none of them seem to be supported by observations and, thus, they concluded that internal plasma structuring is responsible for the density variability and plasma irregularities develop rapidly as the patch drifts across the polar cap. Perry et al. (2015) later postulated that the density variations may be a signature of several patches with scale sizes below the spatial resolution of the radar system.
Figure 4.4 Volumetric image of a patch using RISR‐N data on 11 December 2009 at 22:12:36–22:13:46 UT. The horizontal slices show the electron density at 220, 250, 280, 310, and 340 km altitude. The contemporary 630.0 nm all‐sky image is projected onto the 200 km plane. The locations of the radar beams at each altitude slice are indicated as black circles
(from Dahlgren et al., 2012b; Reproduced with permission of John Wiley and Sons).
Using the most field‐aligned beams from RISR‐C, Ren et al. (2018) automatically identified over 400 patches in 2016 and statistically constructed the patch electron density, electron and ion temperatures, vertical flow and flux profiles in the noon, dusk, midnight, and dawn sectors. Figure 4.5 selectively presents the median profiles of those parameters at the center of patches (blue curves) in the noon and midnight sectors, compared with the sector (red curves) and overall (black curves) median profiles. As expected, the patch median density is higher than the sector median, and the F‐region peak density (hmF2) decreases by ~42% from the dayside (~3.6 x 1011 m‐3) to the nightside (~2.1 x 1011 m‐3). In the noon sector, the bottom F‐region density inside the patch is actually lower than that of the surrounding region, consistent with the SED density profile observation shown in Zou et al. (2013). As suggested in Zou et al. (2013), these SED plasma can be lifted to higher altitudes due to a combination of poleward convection flow and equatorward thermospheric wind. When the production in the lower F region is not fast enough to replenish the transported plasma, this lifting motion will result in a lower F‐region density than the surrounding region. This observation has also been reproduced in the numerical simulation performed by (Zou & Ridley, 2016).
Figure 4.5 Polar cap patch median profiles compared with sector median and all‐sector median profiles. From left to right, plasma density, electron temperature, ion temperature, and ion flux are shown. The ion flux profiles are based on measurements from the RISR‐C vertical beam
(modified based on Ren et al., 2018; Reproduced with permission of John Wiley and Sons).
In addition, Ren et al. (2018) found that the median electron temperature at the center of the patch is suppressed by as much as ~380 K compared with the sector median temperature in the noon sector. The patch median ion temperature is very close to the background ion temperature in all sectors except ~100 K higher in the noon sector, and the ion vertical flows and fluxes within patches are typically downward. Plasma temperature is often used to infer whether local particle precipitation is present in the patch. It can also shed light on the patch generation mechanisms, since higher electron temperature would be expected if local precipitation is responsible for