Discrete aurora and the nightside ionosphere of Mars: an EMM–MEX conjunction of FUV imaging, ionospheric radar sounding, and suprathermal electron measurements

Abstract

Since 2021, a new surge in discrete aurora detections at Mars has been observed by the Emirates Mars Ultraviolet Spectrometer (EMUS) onboard the Emirates Mars Mission (EMM) Hope Orbiter as EMUS started to regularly obtain synoptic auroral images with a high sensitivity. Here we report on a fortuitous conjunction between EMM and Mars Express (MEX) using far ultraviolet (FUV) imaging of discrete aurora by EMM EMUS, in situ measurements of suprathermal electrons by the MEX Analyzer of Space Plasma and Energetic Atoms Electron Spectrometer (ELS), and topside radar sounding of the nightside ionosphere by the MEX Mars Advanced Radar for Subsurface and Ionosphere Sounding (MARSIS). In this event, EMM EMUS imaged a clear discrete aurora signature around moderately strong crustal magnetic fields on the nightside near the dusk terminator, 11 min before which MEX MARSIS measured a prominent local enhancement of the peak electron density in the nightside ionosphere and MEX ELS observed an in situ enhancement of suprathermal electrons at the corresponding location. A remarkable geographic agreement is found between the enhancements of the aurora, ionosphere, and suprathermal electrons, suggesting that the enhanced ionization and auroral emission are caused concurrently by precipitating suprathermal electrons. Subsequent images indicate that the discrete aurora slightly changed its shape in 15 min and mostly disappeared in a few hours. The MEX MARSIS measurements of the auroral ionosphere display overlapping ionospheric and surface echoes indicative of horizontal gradients of the peak electron density. Analysis of the overlapping echoes implies that the auroral ionosphere and electron precipitation could be highly structured with horizontal spatial scales on the order of several tens of km. MEX MARSIS also observed a non-auroral ionospheric enhancement with a wider spatial extent than the local auroral enhancement, suggesting alternative sources of the enhanced nightside ionosphere such as plasma transport. The comparison between the ionospheric structures measured by MEX MARSIS, suprathermal electron flux measured by MEX ELS, and discrete auroral emission imaged by EMM EMUS underscores the complexity of the auroral and non-auroral nightside ionospheres. This motivates further investigations of their sources, transport, and connections to the magnetotail dynamics of Mars.

Graphical Abstract

Introduction

Since the discovery of aurora at Mars (Bertaux et al. 2005), observations and modeling of the Martian aurora have continued to shed light on the interplay between the upper atmosphere and the space environment of Mars. Among the different types of the Martian aurora (Schneider et al. 2015, 2018; Deighan et al. 2018; Ritter et al. 2018; Chaffin et al. 2022), discrete aurora at Mars represents spatially and temporally confined auroral emission conventionally observed in the crustal magnetic fields of Mars (Acuña et al. 1999) and is typically interpreted as driven by suprathermal electron precipitation (Leblanc et al. 2006, 2008; Brain and Halekas 2013; Gérard et al. 2015; Soret et al. 2016, 2021; Schneider et al. 2021). The occurrence of discrete aurora in the strong crustal field region ($60^{\circ }\hbox {S}$–$30^{\circ }\hbox {S}$, $150^{\circ }\hbox {E}$–$210^{\circ }\hbox {E}$) shows clear dependence on the local time and the polarity of the interplanetary magnetic field (IMF) $B_y$ component (Schneider et al. 2021), while the auroral occurrence outside the strong crustal field region depends strongly on the solar wind dynamic pressure (Girazian et al. 2022). These statistical trends are favorably compared with in situ observations of auroral electrons with energies of 50–2000 eV (Brain et al. 2006; Xu et al. 2020, 2022a). The local-time and IMF-$B_y$ dependences of the auroral occurrence in the strong crustal field region can be explained in terms of magnetic reconnection between the draped IMF and crustal magnetic fields (Johnston et al. 2023; Bowers et al. 2023a). Most recently, high-sensitivity and wide-coverage observations by the Emirates Mars Mission (EMM) Emirates Mars Ultraviolet Spectrometer (EMUS) have started unraveling new aspects of the Martian discrete aurora, including their prevalence, instantaneous morphology, temporal evolution, and additional categories including the non-crustal field patchy aurora and “sinuous” aurora (Lillis et al. 2022). These observations suggest a close connection between the discrete aurora and magnetotail dynamics at Mars, the latter of which (a) controls the magnetic field lines guiding suprathermal electrons to the atmosphere and (b) potentially drives the acceleration processes of auroral electrons. The dynamically changing shapes and patterns of the discrete aurora of Mars are likely driven by a variety of dynamic processes that are known to operate in the Martian magnetotail (e.g., Dubinin and Fraenz 2015; Halekas et al. 2021). To name a few, in situ particle and field measurements have revealed the presence of magnetic reconnection (Eastwood et al. 2008, 2012; Halekas et al. 2009; Harada et al. 2015, 2017, 2020; Hara et al. 2017; Wang et al. 2023), energy-dispersed ions and electrons (Halekas et al. 2015; Harada et al. 2016a; Zhang et al. 2023a), tail current sheet flapping (DiBraccio et al. 2015, 2017; Zhang et al. 2023b), and plasma waves (Espley et al. 2004, 2005; Ruhunusiri et al. 2015; Harada et al. 2016b, 2019; Fowler et al. 2021; Teng et al. 2023) in the magnetotail of Mars.

Until recently, the nightside ionosphere of Mars was a relatively unexplored region as opposed to the extensively studied dayside ionosphere (e.g., Withers 2009). In the absence of photoionizing solar radiation, the primary sources of the Martian nightside ionosphere can be either electron impact ionization or transport from the dayside ionosphere (e.g., Fox et al. 1993). Radar sounding by the Active Ionospheric Sounding mode of the Mars Advanced Radar for Subsurface and Ionosphere Sounding (MARSIS) onboard Mars Express (MEX) and radio occultation observations by the Mars Express Radio Science Experiment indicate that the Martian nightside ionosphere is patchy and sporadic, exhibiting spatial inhomogeneity often associated with the crustal magnetic field distribution and temporal variability in response to the changing upstream solar wind conditions and space weather events (Safaeinili et al. 2007; Gurnett et al. 2008; Němec et al. 2010, 2011, 2014; Withers et al. 2012; Duru et al. 2011; Cartacci et al. 2013; Diéval et al. 2014; Harada et al. 2018a; Qin et al. 2021; Lester et al. 2022). In situ observations by the Mars Atmosphere and Volatile Evolution (MAVEN) mission provide a wealth of information on the nightside ionosphere such as the ion composition, global and local structures, seasonal and long-term variability, and species-by-species response to the suprathermal electron precipitation (Fowler et al. 2015; Girazian et al. 2017a, b, 2021; Adams et al. 2018; Cui et al. 2019; Mayyasi et al. 2019; Qin et al. 2022; Xu et al. 2022b). Despite the growing body of knowledge derived from these observations, the relationship between the spatiotemporal variability of the nightside ionosphere and the magnetotail dynamics at Mars remains poorly understood.

In this letter, we present a fortuitous conjunction of topside radar sounding of the Martian nightside ionosphere by MEX MARSIS, suprathermal electron measurements by the Analyzer of Space Plasma and Energetic Atoms (ASPERA-3) Electron Spectrometer (ELS) (Barabash et al. 2006) also onboard MEX, and far ultraviolet (FUV) imaging of discrete aurora by EMM EMUS. The quasi-simultaneous observations of the peak electron density, suprathermal electrons, and the auroral emission enable us to directly link the nightside ionospheric structures and suprathermal electron precipitation to the two-dimensional visualization of energy deposition from the magnetotail to the upper atmosphere represented by discrete aurora.

Observations

Figure 1 provides an overview of the EMM–MEX conjunction event of discrete aurora on 14 January 2022 (MEX Orbit 22790). MEX conducted active ionospheric sounding and suprathermal electron measurements around the periapsis as it traveled northward on the nightside, passing over moderately strong crustal magnetic fields near the equator (Fig. 1e, f, and g). Figure 1b shows the MEX MARSIS observations during 20:05–20:45 UTC in the radargram format. We observe the horizontal trace around the apparent altitude of 0 km representing reflection from the surface of Mars along with sporadic ionospheric echoes between 0–200 km. These are typical signatures of MEX MARSIS sounding of the nightside ionosphere (Němec et al. 2010, 2011; Diéval et al. 2014). Figure 1a shows in situ measurements of suprathermal electrons by MEX ASPERA-3 ELS in the same time interval, indicating a distinct enhancement of local suprathermal electron flux during 20:24–20:25 UTC as discussed in detail later in this section. Meanwhile, EMM EMUS obtained a raster-scanned image from 20:30:27 to 20:44:09 UTC as shown in Fig. 1g. Here we derive the 130.4 nm brightness by integrating differential radiance in a wavelength range of 129–132 nm and generate a geographic emission map of a $1^{\circ } \times 1^{\circ }$ resolution by averaging contributions from overlapping pixels in a similar manner to Lillis et al. (2022). As we are interested in auroral emission, we only show nightside pixels obtained at $>100^{\circ }$ solar zenith angles. In Fig. 1g, we observe a clear discrete aurora signature around $0^{\circ }$ latitudes and $50^{\circ }\hbox {E}$–$90^{\circ }\hbox {E}$ longitudes extending from the dusk terminator. The auroral emissions are located around the moderately strong crustal magnetic fields but appear to be offset from the radial field region as sometimes seen for “crustal field aurora” near the dusk terminator (Lillis et al. 2022). As shown by the MEX orbit track in Fig. 1g, MEX traveled directly above the discrete aurora. A detailed investigation of the acquisition times of individual pixels indicates that the nearest pixel of the EMM EMUS image to the sub-MEX position on the discrete aurora was taken at 20:35:39 UTC as indicated by the white label in Fig. 1d, only 11 min after MEX MARSIS conducted radar and in situ measurements at the corresponding location at 20:24:37 UTC as labeled in Fig. 1g. The close proximity in time of the MEX MARSIS ionospheric sounding, MEX ELS in situ measurements, and EMM EMUS auroral imaging makes this event an excellent case for a direct comparison between the auroral emission pattern, ionospheric properties, and suprathermal electrons.

Here we briefly examine the surrounding magnetic field conditions of this discrete aurora. As shown in Fig. 1g, the discrete aurora was observed on the duskside and located northeast of the radially outward crustal field region (red contours), suggesting a locally northward and eastward crustal field direction (corresponding to “Bth” $<0$ and “Bph” $>0$ in Fig. 1f). The magenta dots in Fig. 1f indicate the local magnetic field strength at the MEX position estimated from the electron cyclotron echoes recorded in MARSIS ionograms (Gurnett et al. 2008; Akalin et al. 2010). We observe that the local magnetic field exceeds 30 nT around 20:24:37 UTC at the MEX altitude of $\sim$400 km above the discrete aurora. This is significantly stronger than both the model crustal field strength at 400 km altitude (black line in Fig. 1f) and the draped field strength of $\sim$10–20 nT observed at higher altitudes before 20:15 UTC and after 20:36 UTC (magenta dots in Fig. 1f). This implies an enhanced pileup of draped IMF, compression and deformation of crustal fields, or combinations thereof. The upstream solar wind was not directly measured by MAVEN due to the orbit configuration, but the IMF clock angle proxy in the dayside magnetosheath (Dong et al. 2019) infers an IMF condition of generally $+B_y$ and mixed $B_z$ polarities for this event (not shown). Because of the varying IMF clock angle proxy, we cannot make a conclusive statement about the shear angle between the draped IMF and underlying crustal fields (Bowers et al. 2023b) and its implications for magnetic reconnection potentially driving discrete aurora (Johnston et al. 2023; Bowers et al. 2023a).

Fig. 1

Overview of the EMM–MEX conjunction event of discrete aurora observations at Mars. a Energy–time spectrogram of suprathermal electrons obtained by MEX ASPERA-3 ELS. MEX MARSIS echo intensities as a function of b apparent altitude and time (radargram) showing the average intensity in a frequency range of 1.4-$-$1.6 MHz and c radio frequency and time (spectrogram) showing the maximum intensity in an apparent altitude range of 50–200 km along with the maximum plasma frequency ($f_{\mathrm{p(max)}}$) of the ionospheric echo denoted by the white lines. The magenta labels of “(2a)”–“(2d)” indicate the acquisition timings of the ionograms shown in Fig. 2. d EMM EMUS 130.4 nm brightness around the sub-MEX orbit track. Note that the horizontal axis in this panel corresponds to the MEX latitude and does not represent time. e MEX solar zenith angle. f Crustal magnetic field at the MEX longitude and latitude at a 400 km altitude computed from Morschhauser et al. (2014)’s model. The magenta dots show the magnetic field strength extracted from the electron cyclotron echoes recorded in MARSIS ionograms (Akalin et al. 2010). The latitude and east longitude of MEX are shown along the time axis. g Image of 130.4 nm emission obtained by EMM EMUS and the MEX orbit on a geographic longitude and latitude map. The gray curve indicates the solar zenith angle of 90$^{\circ }$. The red and blue contours show the outward and inward radial component of the crustal magnetic field at a 400 km altitude computed from Morschhauser et al. (2014)’s model, denoting -40, -20, -10, 10, 20, and 40 nT from blue to red. h Zoom-in comparison of the peak electron density ($N_{\mathrm{e(max)}}$) measured by MEX MARSIS, the integrated electron energy flux over 50–2000 eV measured by MEX ASPERA-3 ELS, and 130.4 nm emission measured by EMM EMUS as a function of latitude along the sub-MEX orbit track

Fig. 2

Representative ionograms obtained by MEX MARSIS. Echo intensity as a function of time delay and radio frequency obtained when MEX was located at a south of the discrete aurora, b immediately above the discrete aurora, c north of the discrete aurora, and d distant from the discrete aurora

Figure 1c shows the MEX MARSIS observations in the spectrogram format with a focus on the ionospheric echoes. The enhanced echo intensities over a continuous range of frequencies represent the ionospheric echoes in this format, while the stripe-like patterns arise from local plasma oscillations (e.g., Gurnett et al. 2008) and are not of interest here. The white lines in Fig. 1c mark the upper edge of the ionospheric echo signals, namely the maximum frequency of the ionospheric echoes ($f_{\mathrm{p(max)}}$), extracted by examining individual ionograms. Specifically, for a given ionogram, potential signals with spectral densities exceeding $10^{-15}$ (V/m)$^2$/Hz are automatically marked by our software. Then, we visually identify an ionospheric echo trace and record the maximum frequency. Some representative ionograms are shown in Fig. 2 with the identified $f_{\mathrm{p(max)}}$ of the ionospheric echoes indicated by the white labels. For comparison, east–west cuts of the 130.4 nm emission map around the sub-MEX orbit track are shown in Fig. 1d. Around 20:24:30 UTC, we observe a sharp enhancement of $f_{\mathrm{p(max)}}$ up to 1.5 MHz (Fig. 1c) and intense suprathermal electron fluxes at energies below $\sim 600$ eV (Fig. 1a) remarkably coinciding with the location of the discrete aurora emission (Fig. 1d). A zoom-in comparison of the radar ionospheric measurements, in situ suprathermal electrons, and FUV auroral emission along the sub-MEX orbit track is shown in Fig. 1h. Here the peak electron density, $N_{\mathrm{e(max)}}$, is converted from $f_{\mathrm{p(max)}}$ using the relation of $f_{\textrm{p}} = 8980 \sqrt{N_{\textrm{e}}}$, where $f_{\textrm{p}}$ is the electron plasma frequency in Hz and $N_{\textrm{e}}$ is the electron density in $\hbox {cm}^{-3}$. We observe the enhanced $N_{\mathrm{e(max)}}$ of $2.9 \times 10^4$ $\hbox {cm}^{-3}$, which is quite high for the nightside ionosphere of Mars (cf. $N_{\mathrm{e(max)}} < 5 \times 10^3$ $\hbox {cm}^{-3}$ for 90% of the time on the nightside as estimated by Němec et al. (2010)). The red line in Fig. 1h shows the integrated electron energy flux over 50–2000 eV, which is used as an empirical measure to select auroral electron events (Xu et al. 2022a). The observed peak energy flux reaches $3.1\times 10^{11}$ eV/$\hbox {cm}^2$/s/sr, far exceeding the empirical threshold of $1.1\times 10^{10}$ eV/$\hbox {cm}^2$/s/sr for auroral emission detectable by MAVEN imaging ultraviolet spectrograph instrument (Xu et al. 2022a). In Fig. 1h, the local enhancements of $N_{\mathrm{e(max)}}$ (black), suprathermal electron energy flux (red), and auroral emission (green) overlap with each other for the most part and have similar widths of $\sim 3^{\circ }$–$4^{\circ }$ in latitude corresponding to $\sim$180–250 km in horizontal distance, yet we observe a small $\sim 1^{\circ }$ shift between the peaks, which will be discussed later in this section. The striking similarity between the enhancements of $N_{\mathrm{e(max)}}$, suprathermal electrons, and auroral brightness provides compelling evidence that the enhanced nightside ionosphere is caused by precipitating suprathermal electrons that also give rise to the discrete aurora.

We point out that the nightside ionosphere is also enhanced in a region far from the discrete aurora. Specifically, MARSIS measured consistently high $f_{\mathrm{p(max)}}$ of $\sim$2 MHz from 20:28 to 20:31 UTC after the data gap (Fig. 1c) while we observe suprathermal electron depletions (Fig. 1a) and virtually no auroral emission at the corresponding locations (Fig. 1d). Such suprathermal electron depletions observed at altitudes above the electron exobase on the nightside suggest closed magnetic field line topology with both footpoints on the nightside (Mitchell et al. 2001; Brain et al. 2007; Steckiewicz et al. 2015). This non-auroral enhanced ionosphere is observed $\sim 15^{\circ }$–$25^{\circ }$ north of the discrete aurora above a weak crustal magnetic field region (Fig. 1f) with a wider spatial extent compared to the local auroral enhancement. The absence of auroral emission and suprathermal electrons suggest that this ionospheric enhancement is not caused by immediate electron impact ionization by precipitating electrons, and by elimination, we can infer that plasma transport (e.g., low-altitude ion transport by collisional coupling with neutral winds (Adams et al. 2018)) could be implicated in forming the enhanced nightside ionosphere observed by MARSIS above the weak crustal field region. Alternatively, MEX ELS may have missed suprathermal electron precipitation due to its limited field of view.

We now investigate detailed ionospheric echo properties in the ionograms shown in Fig. 2. Figure 2b shows the ionogram obtained directly above the discrete aurora (see the label “(2b)” in Fig. 1c). A notable feature of this “auroral ionosphere” ionogram is that the maximum frequency of the ionospheric echo, $f_{\mathrm{p(max)}}$, is significantly higher than the minimum frequency of the surface echo, $f_{\mathrm{s(min)}}$ as opposed to ideal expectation for a horizontally stratified ionosphere for which $f_{\mathrm{p(max)}}$ is equal to $f_{\mathrm{s(min)}}$. These ionospheric and surface echoes with overlapping frequencies can be explained by a vertically reflected surface echo and an obliquely reflected ionospheric echo from a slant iso-density surface of the ionosphere (Duru et al. 2010). The vertical critical frequency, $f_{\mathrm{s(min)}}$, is determined by the peak density at the nadir ionosphere, and the oblique $f_{\mathrm{p(max)}}$ is determined by the peak density of the slant reflection surface at the off-nadir ionosphere. In other words, the overlapping ionospheric and surface echoes are indicative of horizontal gradients of the peak electron density. For the ionograms obtained at locations adjacent to the discrete aurora (Figs. 2a and c), the overlapping feature is not as prominent as the auroral ionosphere. The overlapping ionospheric and surface echoes are also observed for the non-auroral enhanced ionosphere (Fig. 2d).

Based on analysis of the ionogram obtained above the discrete aurora (Fig. 2b) and on a geometric consideration of the overlapping ionospheric and surface echoes as shown in Fig. 3, we can estimate the horizontal gradient of the peak electron density of the auroral ionosphere. The observed properties of the presumably oblique ionospheric echo indicate the maximum electron plasma frequency of $f_{\mathrm{p(max)}} = 1.52$ MHz and the apparent range to the electron density peak of $R' = 258$ km. Here $R'$ represents the virtual distance to the target under the assumption that the radio wave propagates at the speed of light, and the actual range, R, is smaller than $R'$ because of the dispersion of the radar pulse propagating through the ionized medium with group velocities slower than the speed of light. The observed minimum frequency of the surface echo, $f_{\mathrm{s(min)}}$, indicates that the sub-MEX vertical critical frequency is 1.02 MHz. Since the peak altitude of the nightside ionosphere is typically $\sim 150$ km in the presence of precipitating suprathermal electrons (Fillingim et al. 2007; Lillis et al. 2009; Girazian et al. 2017a), we assume an ionospheric peak altitude of $h_{\textrm{iono}} \sim 150$ km. We note that the peak altitude of Martian discrete aurora is expected to be generally around $\sim 135$ km (Schneider et al. 2021; Soret et al. 2016, 2021). With these observations and assumptions, the off-nadir angle of the oblique echo, $\theta$, and the horizontal distance from the sub-MEX point to the oblique echo reflection point at the peak altitude, d, can be estimated as $\theta <\sim 18^{\circ }$ and $d <\sim 80$ km (Fig. 3). This suggests that the overlapping ionospheric and surface echoes result from the peak electron density varying from $1.3 \times 10^4$ $\hbox {cm}^{-3}$ to $2.9 \times 10^4$ $\hbox {cm}^{-3}$ within a horizontal distance of $<\sim 80$ km. This distance is comparable to, or even smaller than, the effective spatial resolution of the EMM EMUS image (the nearest EMUS pixel to the sub-MEX position at the MARSIS auroral ionospheric measurement has $4.9^{\circ }$ longitude $\times$ $0.91^{\circ }$ latitude widths, approximately corresponding to 300 km $\times$ 55 km at 150 km altitude), implying that there could be spatial inhomogeneity of aurora unresolved by EMM EMUS for this event.

Another point to note is that the ionospheric echoes recorded in Fig. 2a–d have relatively broad widths in time delay even though the transmitted pulse has a very short duration corresponding to the width of one time-delay bin. These signatures are referred to as “diffuse echoes” in the literature and are interpreted as a superposition of multiple echoes reflected at oblique angles from ionospheric irregularities (Gurnett et al. 2008; Harada et al. 2018b). Diffuse echoes are commonly observed for the nightside ionosphere (Harada et al. 2018b) as expected for generally large fluctuations of electron and ion densities in the nightside ionosphere (Girazian et al. 2017b; Park 2024).

Fig. 3

Schematic illustration of a geometric consideration of the overlapping ionospheric and surface echoes. See text for details

Fig. 4

Images of 130.4 nm emission of the same region taken by EMM EMUS. The images were taken at a1 $\sim$11 min, b1 $\sim$26 min, and c1 $\sim$2.6 h after the MEX MARSIS observation of the peak electron density enhancement. For reference, the times at which individual pixels were measured are indicated in panels (a2, b2, c2)

Finally, we examine the time evolution of the discrete aurora by comparing the EMM EMUS images obtained at different times (Fig. 4). Figure 4a1 shows a close-up of the EMM EMUS image shown in Fig. 1g, and  4b1 and 4c1 shows subsequent images of the same region obtained by EMM EMUS available for this event. Figure 4a2, b2, and c2 demonstrates how the pixel acquisition times vary over the maps. The nearest pixels to the sub-MEX position at the MARSIS auroral ionospheric measurement were taken at 20:51:28 UTC for Fig. 4b1 and at 23:03:18 UTC for Fig. 4c1. We can see that the discrete auroral pattern slightly varies its shape over $\sim$15 min from Fig. 4a1 to Fig. 4b1. Such small yet visible variations on $\sim$10 min time scales may explain the aforementioned small shift between the $N_{\mathrm{e(max)}}$ and auroral emission enhancements (the black and green lines in Fig. 1h), which were observed 11 min apart. Also, the in situ measurements of suprathermal electrons at the MEX altitude do not necessarily map exactly to the nadir auroral emission depending on the magnetic field configuration (Qin et al. 2021), which could explain the shift between the enhancements of the suprathermal electrons and auroral emission (the red and green lines in Fig. 1h). In Fig. 4c1 ($\sim$2 h later), the overall auroral emission intensity is greatly weakened and the once-coherent structure of clear discrete aurora has fragmented into small patches of very weak emission, suggesting that this discrete aurora pattern had almost disappeared by this time. The dynamics of the Martian discrete aurora could be attributed to temporally varying magnetic field and plasma conditions in the magnetotail resulting from the dynamic solar wind–Mars interaction.

Conclusions and implications

We analyze the quasi-simultaneous FUV, radar, and in situ electron observations relevant to discrete aurora at Mars. In this event, the FUV image of a clear discrete aurora on the duskside of Mars was obtained by EMM EMUS just 11 min after MEX MARSIS conducted topside radar sounding of the nightside ionosphere and MEX ELS measured the enhancement of suprathermal electrons. We observe a local and distinct enhancement of the ionospheric peak electron density (up to $2.9 \times 10^4$ $\hbox {cm}^{-3}$) and in situ electron energy flux (up to $3.1\times 10^{11}$ eV/$\hbox {cm}^2$/s/sr in the 50–2000 eV energy range) spatially coinciding well with the discrete auroral emission, demonstrating that the locally enhanced nightside ionosphere and discrete aurora at Mars are caused by a common agent: precipitating suprathermal electrons. The subsequent auroral images of the same region indicate that the discrete aurora was dynamic, exhibiting time-varying patterns and intensities. The auroral emission slightly changed its shape over 15 min and its overall intensity mostly diminished in a few hours. Based on the properties of the overlapping ionospheric and surface echoes measured by MEX MARSIS, we can infer that the peak electron densities vary by a factor of 2 or more within a horizontal distance of $<\sim 80$ km. The observed inhomogeneity of the auroral ionosphere suggests that the auroral electron precipitation could be highly structured with horizontal spatial scales on the order of several tens of km. Meanwhile, a broader region of enhanced electron densities was observed with suprathermal electron depletions and without any prominent auroral emission around $15^{\circ }$–$25^{\circ }$ north of the discrete aurora, implying a non-auroral source of the nightside ionosphere in this region (e.g., plasma transport from the dayside ionosphere via collisional neutral-ion coupling at low altitudes (Adams et al. 2018)). To examine the nightside ionospheric sources, in situ ion composition measurements will be highly informative (Girazian et al. 2017b), and further investigations should be conducted to reveal the sources and behaviors of the non-auroral nightside ionosphere of Mars. Both the auroral and non-auroral ionospheres display diffuse echoes indicative of ionospheric irregularities. These observations highlight the complexity of sources, transport, and dynamics of the Martian nightside ionosphere and suggest that a successful model of the nightside ionosphere would need to include both electron impact ionization and day-to-night plasma transport, the former of which can vary on short time and small spatial scales resulting from the highly dynamic magnetotail of Mars.