A total of 33 ESR scan experiments are used in this study, each run on separate days. The duration of the individual experiments ranges from 30 to 358 min. Combined, the 33 experiments consists of approximately 100 h of observations. Most of the experiments are centered at magnetic noon (local morning time for Longyearbyen), but the data also includes some pre-noon and post-noon observations. All experiments were run during the boreal winter, with the majority being from the winters of 2014 and 2015 (solar maxima), during 2002 (declining phase) or 2010 (solar minima). Information on the scans modes, orientations, and durations of each experiment are listed in Table . The data from each experiment was inspected visually, and the events with a clear auroral oval at 300 km were included in the data set used for this study.

Figure  shows ESR observations from 4 December 2014. The ESR was run in the taro experiment mode with the radar line of sight moving from 30° elevation pointing South to 30° pointing North, with every elevation step covering 6.4 s. This leads to full N–S scans over approximately 3 min. The shown parameters are the averages between 275–325 km over the different latitudes covered by the radar beam unless otherwise stated. For context, panels (a) and (b) show the IMF By and Bz components, and the AE and PCN indices, respectively. Panels (c)–(g) present parameters Te, Ne,F, Ne,E (E-region, 110–130 km), Ti, and vi, respectively. The velocity vi is the horizontal component of the along-beam observed velocity. It is calculated from the observed velocity and the radar elevation angle. Velocities where the radar beam is almost vertical, from 85 to 95° elevation, are discarded.

The open-closed field boundary based on the 32-meter observations is overlaid in black. The scattered squares colored in red, green or yellow are the OCB location quality flags for each time step. Panel (h) shows the identified regions, as illustrated in Fig. . Regions on closed field lines are colored in grey, the 3° band directly North of a 0.3° buffer in red shows the OCB region, and the polar cap region is drawn in green. The offset from UTC to MLT in Longyearbyen is approximately UTC+3h. The shadow height at the ESR latitude is approximately 300 km at the beginning of the experiment and 120 km at the end of the experiment. The altitude is calculated using solar elevation angles from the PyEphem Python package .

For the case shown in Fig. , the IMF Bz and By are relatively stable and only the magnitude is slowly varying. Bz is positive throughout, save a few short intervals of low magnitude negative Bz in the middle of the experiment. By is negative throughout, except for some minor switches in the beginning of the experiment. The AE index, shown in panel (b), is low during the experiment, but reaches a peak of 200 nT slightly before 09:00 UTC. The PCN index is also relatively low throughout, but it increases slightly after 09:00 UTC.

The OCB is far North in the radar FOV at the beginning of the experiment, and gradually moves equatorward. The farthest South location is at 09:00 UTC, coinciding with the AE peak. After 09:00 UTC, the OCB retracts back towards the North of the radar FOV. The correlation coefficient between the OCB latitude and the AE index is -0.65. The OCB latitude error flags in Fig. c and d show that the boundary position is flagged as uncertain (“red”) where the position is found in a region of missing data, like shortly after 07:00 UTC. Similarly, the location is flagged as uncertain where the error bars in the latitude is large, e.g. shortly after 09:00 UTC.

As expected, Te shown in panel (c) is enhanced along the identified boundary and the gradient from closed to open field lines is generally well defined. During intervals where the gradient is not well defined, the boundary latitude is flagged as uncertain, e.g. at approximately 09:00 UTC. In addition, the enhancements in Te are clear during the entire experiment and span several degrees in latitude. In particular, no differences between the OCB region and the PC are visible here. The enhancements appear to peak between 08:00–09:00 UTC. This collocates with the AE maximum.

Ne in the F-region, shown in panel (d), is relatively stable throughout but there is a region of slightly lower density on closed field lines in the middle of the experiment. Toward the end of the experiment there is an increase in Ne at the Southward edge of the radar FOV which creeps Northward. Ne,E, shown in panel (e), is slightly higher in the first half of the experiment. On open field lines, the density is lower, but the difference is not striking.

Ti, shown in panel (f), exhibits intermittent enhancements. However, the enhancements are not as clear and as persistent in time as the Te enhancements. The region of most clearly and evenly enhanced Ti coincides with the peak Te between 08:00–09:00 UTC. There is also an enhanced region between 07:00–07:30 UTC along the OCB and a region at the equatorward edge of the FOV at 08:00 UTC.

Although the measurements are very noisy, there are some patterns in vi (panel g). For example, there is a region of reverse flow shortly after 07:30 UTC.

Figure  shows ESR observations from 5 December 2015 (in the same format as the case shown in Fig. ).

For the December 2015 experiment, By, shown in panel (a), is consistently positive, except for some short intervals prior to and after 09:00 UTC. Bz is positive until about 07:15 UTC before turning southward, reaching values of up to -10 nT. Bz remains negative until shortly before 10:00 UTC, save for a slight switch at 09:00 UTC. The AE index, shown in panel (b), is low and stable until about 08:00 UTC, growing slowly before a sharp and short-lived peak before 09:00 UTC. While remaining high for the rest of the experiment, it decays slowly toward the experiment end. The PC index increases steadily from 08:00 UTC and reaches its peak at ∼09:15 UTC. The shadow height at the ESR latitude is approximately 340 km at the beginning of the experiment and 115 km at the end of the experiment.

The OCB, which is plotted over the ESR parameters in panels (c)–(g), is almost directly overhead at the beginning of the experiment, before propagating toward the South edge of the FOV following the southward turning of the IMF. However, there is a ∼20 min delay before the OCB starts moving. The OCB latitude is low until about 09:30 UTC before retracting. Around this time the IMF turns northward. As for the 2014 case, the OCB seems to be inversely linked to both the AE and PCN indices, where the largest AE values coincides with the lowest OCB latitudes. In fact, the correlation coefficient between the OCB latitude and the AE index is -0.91.

The enhancements in Te (shown in panel c) on open field lines have a wide peak centered at approximately 08:20 UTC. After 08:20 UTC and until approximately 09:30 UTC, there is a strong N–S Te gradient from the region on CF to the OCB region while there is a softer gradient northward of the Te enhancements into the polar cap. During this interval, there are also several Polar Moving Auroral Forms (PMAFs). These are seen as defined structures of enhanced Te that move poleward (Northward).

Ne is stable throughout, save for some patchy activity and a depletion around 09:00 UTC. In the E-region, shown in panel (e), Ne is highest at the start and the end of the experiment. There is a clear reduction as the IMF turns Southward and the OCB moves southward. Toward the end of the experiment the density again increases.

There is a more clearly defined region of enhanced Ti than for the December 2014 case. From 07:30 UTC to about 09:20 UTC there is a clear band of enhanced Ti, although it is less even than the Te band. Some other regions of enhanced Ti, e.g. before 07:00 UTC and after 09:30 UTC, are also seen. There are several points of interest in vi. Following the turning of the IMF, its magnitude increases and the velocity is clearly poleward. At the location of the highest Te and Ti, and coincident with the 4th PMAF at 10:00 UTC, there are pockets of reversed flow.

Following the southward turning of the IMF at approximately 07:15 UTC, several features are observed in the ESR measurements. For Te, the strength of the S-N gradient across the OCB increases. The enhancements in Ti increase. vi is strongly northward from the southward turning, and until the IMF northward turning. These three features are observed just after the southward turning at 07:15 UTC. The OCB moves equatorward following the southward turning, but the movement only starts ∼20 min after the turning and the features are observed in the ESR parameters.

To summarize, this section has presented two case studies that show different IMF conditions and ionospheric responses. The key difference is that the 2015 case (Fig. ) contains a strong southward turning of the IMF and the ionosphere is more structured, with both patchy Ne and PMAFs.

For completeness, this section presents a statistical analysis of the location of the OCB. The location and movement of the OCB is linked to the opening of magnetic flux at the dayside magnetopause and to the closing of flux in the magnetotail, and its latitude is therefore expected to relate to both the IMF and geomagnetic indices such as AE references therein. This is previously shown in studies based on ASI and satellite e.g.. Here, we therefore investigate the relation of the OCB latitudes obtained by the ISR method to the IMF Bz component, the AE, and PCN indices and to a coupling function .

Figure  shows the boundary latitude obtained by the radar method. The latitudes are shown as a function of dΦMP/dt which represents the change in open magnetospheric flux , the AE and PCN indices, and the IMF Bz component. Only data points where 11:00 < MLT < 13 are included as the cusp is most likely within this interval .

In Fig.  a, the red dots indicate where Bz≤0 nT and the blue dots where Bz>0 nT. In panel (b), blue dots indicate points sorted with respect to the AE index, the red dots with PCN. For all three plots (panels a–c), large points mark data points where |By|<3 nT. Small and more lightly colored points have |By|>3 nT. The right-hand axis shows the number of points per interval on the horizontal axis. Points where the OCB latitude is flagged as uncertain are removed.

Although the distributions of latitudes are widespread and clearly limited by the narrow ESR field of view, some trends are visible. For both the AE index and dΦMP/dt, the variation in the OCB latitude is larger for the smallest values, spanning the entire range of the radar FOV. For larger AE, especially above 500 nT, no OCB latitudes above 75° are detected. The same trend is seen in dΦMP/dt. Larger index values yield a smaller range of latitudes, shifted towards the equatorward edge of the radar FOV. This is true for both large and negligible |By|.

For IMF Bz (panel c) the trend is similar, but weaker. For Bz>0 nT, the spread in latitude is large. For Bz>5 nT and |By|<3 nT, there is a preference for higher latitudes, but the number of points in this range is low. For Bz<0 nT, the latitudes are lower for both small and large |By| nT. For Bz>0 nT and By>3 nT, there are also a number of data points with lower latitudes. This population disappears when considering only |By|<3 nT. Overall, the obtained OCB latitudes behave in the same manner as has been presented in previous studies , although our data set is smaller and there are some limitations concerning the FOV.

For an overview of the variation in Te, Ti, and Ne, this section presents a statistical analysis of the three parameters in the three regions CFL, OCB, and PC.

Using elevation scans makes it possible to observe both latitudinal and altitudinal variation of the ionospheric parameters within ∼3 min. Accordingly, the altitudinal variation can be investigated separately. In this section, we present a quantitative analysis of the variation between the three regions, CFL, OCB, and PC, by constructing average altitude profiles within each region. Prior to constructing the profiles, the data is filtered by a median filter with kernel size 3. This is done for the statistical results in this section and further sections.

Figure  shows an overview of Te, Ne, and Ti shown as altitude profiles in the three regions. The medians for each altitude bin for each region make up the the altitude profiles in the left column. The middle column shows the distributions of each parameter for every 30 km altitude bin for each region. The distributions are centered at the median for each altitude bin. The right hand column shows the number of points contributing to the altitude bin median and distribution in each altitude bin. The distributions for all three parameters include some amount of unreasonable outliers. Therefore, the error estimates for each altitude range are calculated as the 30th and 70th percentiles of the distributions.

From Fig. a, it appears that the Te difference between CFL and along the OCB increase with altitude, especially above 200 km. The increase peaks upwards from 300 km with a difference of 900–1000 K from the closed field lines, corresponding to a factor of 1.5–1.6. Although the error bars are large and the distributions are wide, it appears that Te is hotter along the OCB than in the PC above 200 km. Above 300 km, the median is on average ∼200 K degrees hotter along the OCB than in the PC.

A similar trend is seen in Ti (panel d), where an increase from closed to open field lines above approximately 170 km altitude is seen. From around 260 km and upwards, the increase from CFL to open field lines (OFL) averages to 120 K and varies little with altitude, corresponding to a factor increase by 1.1–1.2. There are no distinguishable differences between the OCB and the PC.

For Ne in panel (g), there is a clear increase along the OCB centered at around 280 km altitude. Between 170–300 km Ne along the OCB reaches densities over 1.3×1011m3, with an increase from CFL to the OCB by a factor 1.1–1.4. In the PC, the density is lower than along the OCB across all altitudes except at the top altitudes. The PC densities are slightly higher than the CFL densities, except at the lowest altitudes.

For a more detailed look at the latitude variation of the parameters, Fig.  shows a superposed epoch analysis of Te (top row), Ti, and Ne (bottom row) with respect to the OCB, allowing to study how the quantities vary with distance from the OCB.

The groups of red, blue and green lines correspond to the ionospheric parameters on altitude intervals 100–150, 200–300 and 300–400 km. The data points are extracted from altitude cuts like the ones in Figs.  and . Each altitude cut is the average over 25 km. The colored-black dashed lines show the median for each altitude range. The horizontal axis shows the latitude offset (AACGM degrees) from the OCB. Each column shows a specific time interval. The left column shows interval 09:00–11:00 MLT, the middle column shows interval 11:00–13:00 MLT, and the right column shows 13:00–15:00 MLT.

As seen in panels (a)–(c), the F-region Te exhibits clear variations with respect to latitude relative to the OCB regardless of the time interval. Most striking is the strength of the Te gradient across the OCB. The gradient when crossing the OCB is significantly steeper for the 11:00–13:00 MLT interval than for the two other intervals. For 11:00–13:00 MLT, Te is enhanced until 4° poleward of the OCB before decreasing. For all three time intervals, the spread in the F-region Te is larger poleward of the OCB. Between 100–150 km altitude, the variation in Te relative to the OCB is negligible expect for a larger spread poleward of the OCB.

For Ti shown in panels (d)–(f), there are also some differences dependent on the relative distance to the OCB. Although Ti also increase poleward of the OCB, no strong latitudinal gradient such as the one seen in Te can be observed.

In the F-region, Ti is slightly enhanced on open field lines. Latitudinal variation apart from this enhancement is hard to discern although there might be some differences between the three MLT ranges. For all three intervals, there is a significant difference in the standard deviation between open and closed field lines. On open field lines, poleward of the OCB, the standard deviation is significantly enhanced. Between 13:00–15:00 MLT there is little to no latitudinal variation in Ti.

The behavior of the F-region Ne (panels g–i) is different between the three intervals. For 09:00–11:00 MLT, the Ne peaks poleward of the OCB and decreases gradually into the polar cap. For 11:00–13:00 MLT there is a less pronounced peak and Ne decreases gradually from the equatorward edge of the FOV to the poleward edge. For 13:00–15:00 MLT there is a stronger difference between 200–300 and 300–400 altitude ranges, and it peaks just poleward of the OCB between 200–300 km altitude. In the E-region, there is a pronounced decrease in Ne on open field lines. For 13:00–15:00 MLT, there is a slight peak in Ne just northward of the OCB. For all three time intervals, the lowest E region Ne values are poleward of the OCB.

To further characterize Ne in the dayside ionosphere, this section presents an analysis of the Ne variation in the E and F regions. The electron densities in the E and F regions depend on different precipitation energies and therefore contain information about the precipitation into the dayside auroral region. Their time variation or relationship can both be used to gather information about said precipitation e.g.. In this section the Ne,E to Ne,F ratio is quantified for the scans used in this study, which may provide information on the ratio of high to low energy precipitation.

Figure  shows the ratio between Ne,E to Ne,F as a function of MLT for the PC (panel a), OCB (panel b) and CFL (panel c). For each timestep, average Ne values in the E and F regions are calculated in the CFL, OCB, and PC regions. To account for the FOV size differences between high and low altitudes, the F region Ne is cropped to match the E-region FOV before the averages are calculated. The colored points show the hourly median. The last binned median is on the interval 13:00–14:00 MLT. This is because the number of points in the next two bins is low, especially for the CFL and PC regions. For completeness, the number of points per MLT hour interval is also plotted using dashed black bars.

Even though the FOV is limited in the E region and the spread in the data is naturally relatively large, some trends with respect to MLT can be perceived. The behavior of the Ne ratio is different between the three regions.

On CFL, the ratio peaks pre-noon at values up to 0.6. After 11:00 MLT the ratio varies little around approximately 0.5. On open field lines (OCB and PC) the ratio is almost constant with MLT, and any MLT variation is far within the error bars. The ratios averages at almost 0.3.

Larger ratios are explained by either larger E region densities or lower F region densities. In general, larger ratios are more likely on CFL than on open field lines regardless of MLT range.

The case examples show aspects of the dynamic behavior of the dayside auroral region. The regions of enhanced Te clearly display the presence of low energy precipitation. The 2015 event (Fig. ) exhibits a southward turning and shows signatures of pulsed magnetopause reconnection. Indeed, after 08:00 UTC, several PMAFs appear. PMAFs are typically related to flux transfer events in the dayside magnetopause references therein. The intensity of the PMAFs and the time intervals between them vary, which is consistent with previous observations . Strong latitudinal Te gradients such as the one seen after ∼07:30 UTC in Fig. d, has also been linked to the cusp . There is also a strong enhancement of the poleward flow velocity in vi, which can be linked to reconnection. In the 2014 case in Fig. , where the IMF is positive throughout, less structured behavior and fewer signatures associated with magnetopause reconnection are seen.

In the 2015 case in Fig. , the OCB moves equatorward following the southward turning of the IMF, which can be explained by reconnection at the magnetopause which expands the region of open field lines equatorward. After the IMF turns northward, the OCB relaxes poleward. In the 2014 case, the OCB also moves equatorward, which can be explained by growing convection cell size due to the considerable By component . This is also seen in the OCB statistics in Fig. c, where both Bz and By affect the latitude.

Between the two presented cases, there seems to be less structured behavior in all four ionospheric parameters for positive Bz, especially in Te and Ne. The two cases differ most after the southward turning, where we observe highly structured Te, Ne, and an increased vi in the 2015 case (Fig. ). Similar trends were observed by in field aligned ESR observations. This is also seen in several of the other experiments in Table , especially in cases with well-defined intervals of stable southward or northward IMF. Although common, not all of the experiments match this description, which helps illustrate the dynamic nature of the dayside auroral ionosphere, and challenges in predicting its behavior. It is also worth mentioning that directly comparing the ionosphere to IMF observations at Lagrange point 1 can be influenced by the variation of the spacecraft relative position to the Sun–Earth (S–E) line and the time delay from the bow shock to the ionosphere . Therefore, exact characteristics of the solar wind and IMF that reach the magnetopause can be uncertain. Although this effect is not large, small uncertainties concerning IMF effects can be introduced. For instance, for the 2015 case (Fig. ), the NASA Magnetospheric Multiscale Mission (MMS) satellites crossed the S–E line during the experiment (seen in MMS orbit plot from https://lasp.colorado.edu/mms/sdc/public/plots/#/historical-orbit, last access: 8 April 2026). The MMS observed IMF exhibits overall similar characteristics to that of the OMNI dataset, including the timing of the Southward turning; however, some small variations can be perceived (seen in MMS quick look plots at https://lasp.colorado.edu/mms/sdc/public/plots/#/quicklook, last access: 8 April 2026). A more detailed analysis of the effect of such variations between IMF observations at different positions along the S–E line and ESR measurements could be interesting, but it is outside the scope of this study.

Figures  and shows statistical results considering Te, Ne, and Ti. Before these results are discussed, four aspects of the statistical analysis are addressed in the following four paragraphs.

First, Te is the main parameter used in the method for obtaining the OCB latitude and for separating open from closed field lines. Therefore, a separation between open and closed field lines and higher Te on open field lines is expected. However, Te is the parameter exhibiting the strongest latitudinal differences between open/closed field lines when viewing the experiment observations separately (e.g. Figs.  and ). In the superposed epoch analysis, we also note that the OCB is typically placed where the Te gradient is strongest. Further, every timestep where the latitude of the OCB is uncertain is not included in the statistics.

Secondly, for this study, the regions of interest were defined to primarily separate the ionosphere on open field lines and closed field lines, and to distinguish the region closest to the OCB from the remaining open field lines. The region closest to the OCB contains possible cusp passes and possible reconnection signatures due to its proximity to the OCB. It is also likely that cusp passes are classified in our PC region if the cusp is more than 3° wide in latitude, which is not uncommon . In addition, our OCB region likely contains ionosphere in the polar cap. The size of the cusp is dynamic and separating it from the remaining open field lines is not straightforward. However, the altitude profiles do not vary significantly when varying the separation latitude (OCB +3°) between the OCB and PC regions.

Next, the velocity vi was omitted from the statistical analyses. Due to the nature of elevation scans, the observed line of sight vi is a mixture of vertical and horizontal velocities depending on the current elevation angle. When calculating the horizontal N–S vi, data is lost overhead the radar and emphasis is therefore put on the edges on the FOV. This is especially critical for the E-region, where the FOV is narrow. In addition, depending on the ESR position relative to the convection pattern, the magnitude of N–S and E–W velocities will vary . Since the data set includes experiments along both the geographic and geomagnetic meridians, the interpretation of what parallel velocity that is observed is slightly different (magnetic or geographic N–S).

Finally, attempts to quantitatively assess vi proved difficult due to a large number of unreasonably high values in the E-region and a high sensitivity to the treatment and filtering of the observations prior to quantification. This sensitivity indicates that any possibly obtained trends in vi would be too uncertain. On the other hand, this sensitivity was not observed in the remaining ionospheric parameters.

Despite the challenges above mentioned, Fig.  showed differences between the ionospheric parameters on open (OCB, PC) and closed field lines along the altitudinal column. The differences are, as expected, most clear in Te. However, possible trends are also seen in Ne and Ti.

Simultaneous signatures in both Ne and Te have been used to identify the cusp region in ISR data, e.g. by thresholds such as Te>3000 K and Ne>3×1011 m-3 above 300 km . These values are significantly larger than the average values in the statistical altitude profiles in Fig. . There is enhanced Ne along the OCB where the cusp precipitation is expected to enhance Ne at 280 km, but values of 3×1011 m-3 are rare and accounts only for 4.3 % of all Ne values in the OCB region and 2.6 % in the PC at 300 km. In addition, these percentages are elevated by days were the electron density is generally high. use a threshold of 1800 K for Te when identifying the dayside auroral region. This threshold agrees well with the separation between Te on CFL and along the OCB in our dataset. The ESR data based electron heating model in suggests that electrons heated by precipitation on open field lines are separated from cool electrons on closed field lines at approximately 2000 K. This also fits the presented separation.

In their model study, predict strongly enhanced Te reaching up to 1000 K peaking at around 300 km due to cusp precipitation. The expected Te are significantly lower for boundary layer precipitation, only a few ∼100 K. Our observed enhancements of up to 1000 K is in agreement with these values, which might indicate that a large fraction of our observations contain cusp precipitation. Almost 65 % of the data points in this study is within the interval 10:30–13:30 MLT which is the most probable MLT range for cusp observation .

In their study, observe altitude profiles of the ionospheric parameters within the cusp, the polar cap and equatorward of the dayside aurora with the ESR looking field aligned. However, their altitude profiles are not measured simultaneously and are separated by up to 30 min and contains only a few minutes of data each. They observe Te enhanced in the cusp region that are significantly higher than the temperatures in Fig.  at above 3500 K at 400 km altitude. They observe lower temperatures in the polar cap than in the cusp, at even lower values than the sub-auroral Te. Although the magnitudes differ, the results presented in Fig.  are similar and support these previous observations.

For the latitude variation, several points of interest are seen in Fig. . The enhancements in Te in panels (a)–(c) are wide with respect to latitude poleward of the OCB in the F-region. The width of the enhancements are slightly larger than what is expected of the cusp . For the interval 11:00–13:00 MLT, the equatorward Te gradient is the strongest. This is a cusp signature , consistent with strong anti-sunward flow of cool plasma recombining with heated plasma below the cusp. The fact that there is little to no visible latitude effects in the E-region Te is explainable by considering that the low energy precipitation that heats the F-region electrons on open field lines does not reach E-region altitudes.

The increased variation in Ti on open field lines in Fig. d–f matches the observations in the presented case examples, i.e. that the enhancements in Ti are more sporadic than Te. Enhancements in Ti is also not necessarily coincident with Te enhancements, which is also observed by . Increases in Ti are heavily linked to enhanced Joule heating rates that are linked to increased flows e.g., meaning that vi is likely enhanced where Ti is high. The increased variability in Ti on open field lines would be worthwhile a more detailed look in a future study.

The investigation of Ne in the E-region compared to the F-region (Fig. ) showed that there are slight MLT trends on CFL. The ratio decreases after 11:00 MLT. A small ratio signifies that the E-region is less dense or the F region density is larger. This scenario can be explained by low energy precipitation into a dark ionosphere, consistent with dayside precipitation from either the cusp or mantle . Low energy precipitation does not penetrate into the E-region, which is depleted due to a lack of ionization in the dark ionosphere. The precipitation rather increases the density in the F-region. This process is used for identification of the cusp e.g., where the cusp is found in regions with a depleted E region coincident with an enhanced F region Ne. use conditions on both the E and F region densities when identifying the cusp.

Although the error estimates and variation per MLT bin in our analysis are large, the trends are as expected. Denser E-regions are more likely on closed field lines especially pre-noon due a larger amount of higher energy precipitation on closed field lines pre-noon than compared to post-noon . In addition, there is a larger amount of ionization due to solar EUV post-noon (09:00 UTC). The effect of EUV ionization is largest in the F-region, but after noon, the shadow height is below 100 km altitude in the southern part of the ESR FOV. In Fig.  there is an increase of Ne at the equatorward edge of the FOV after 09:00 UTC that grows poleward. The F-region Ne can also be affected by structures such as patches, which can influence the comparisons.