- Email: [email protected]
Martian ionosphere observed by Mars Express. 1. Influence of the crustal magnetic fields
Planetary and Space Science 124 (2016) 62–75
Contents lists available at ScienceDirect
Planetary and Space Science journal homepage: www.elsevier.com/locate/pss
Martian ionosphere observed by Mars Express. 1. Influence of the crustal magnetic fields E. Dubinin a,n, M. Fraenz a, D. Andrews b, D. Morgan c a
Max-Planck-Institute for Solar System Research, 37077 Goettingen, Germany Swedish Institute of Space Physics, Uppsala, Sweden c University of Iowa, Iowa city, Iowa, USA b
art ic l e i nf o
a b s t r a c t
Article history: Received 16 October 2015 Accepted 11 February 2016 Available online 21 February 2016
We present multi-instrument observations of the effects of the crustal magnetic field on the Martian ionosphere at different altitudes and solar zenith angles by Mars Express. Total electron content (TEC) at solar zenith angles 551 Z SZA Z 1051 over the ionosphere with crustal sources increases with the strength of the magnetic field. A similar trend is observed in a dependence of the local electron density in the upper ionosphere on the crustal magnetic field. On the nightside, at SZA Z 1101, the opposite trend of TEC increase with decrease in the magnetic field value is observed. A dependence on the magnetic field inclination also varies between the day and night sides. TEC decreases for vertical field inclination at 901 Z SZA Z 701 and increases at SZA Z 1101. This effect becomes stronger for larger magnetic field values. A different dependence of the local electron densities in the upper ionosphere at small and high SZA is observed too. An ionospheric exhaust for vertical field inclination in the regions with strong crustal sources is probably caused by escape to space along open field lines which arise due to reconnection that is confirmed by the case studies. The existence of such localized ionospheric depressions is also observed by the in-situ plasma observations. In contrast, on the nightside downward plasma transport and electron precipitation along the field lines produce patches of enhanced ionization. & 2016 Elsevier Ltd. All rights reserved.
Keywords: Mars Crustal magnetic field Ionosphere Mars Express
1. Introduction The first detection of the Martian ionosphere with a peak value of the electron number density of about 105 cm 3 at an altitude of about 125 km was made by Mariner-4 using the method of radio occultation (Kliore et al., 1965). Later, similar measurements on other space missions to Mars have extensively sampled the Martian ionosphere (Kliore, 1992; Hinson et al., 1999; Pätzold et al., 2005). In particular, a lower-altitude density peak or shoulder at h 110 km was also identified. The first in-situ ionospheric measurements were carried out by Viking-1 and 2 landers (Hanson et al., 1977) and provided us with the height versus density profiles of the main ionospheric species (O2þ , CO2þ and O þ ). Models of the ionosphere (see e.g. Shinagawa and Bougher, 1999; Krasnopolsky, 2002; Fox et al., 1993, 2004; Schunk and Nagy, 2009; Mendillo et al., 2011; Chaufray et al., 2014) have significantly increased our knowledge of the ionospheric structure and its dynamics. The bulk of the ionosphere at Mars is created by the ionization of the major atmospheric neutrals CO2. The main ionization n
Corresponding author. Tel.: þ 49 551 384979129. E-mail address: [email protected] (E. Dubinin).
http://dx.doi.org/10.1016/j.pss.2016.02.004 0032-0633/& 2016 Elsevier Ltd. All rights reserved.
process is photoionization by the solar EUV irradiance (10–90 nm). A lower peak at 90–100 km is produced by soft X-rays ð r10 nmÞ. However, due to the photochemistry reactions, the dominant ion species at hZ 200 km become molecular ðO2þ Þ and atomic ðO þ Þ oxygen ions. A balance between photoionization and recombination determines the peak electron density at altitudes of about 130 km. At altitudes above 180 km the ionosphere is no longer in photochemical equilibrium and diffusion and transport processes become important. Due to recombination of O2þ ions the ionospheric density rapidly decreases at the nightside. The recent observations by Mars Global Suveyor (MGS) and Mars Express (MEX) spacecraft have strongly increased the database of the ionospheric measurements providing us a lot of new information. Reviews of these observations are given in Haider et al. (2011), Orosei et al. (2015), and Withers et al. (2015)). In addition to the radio occcultation experiment, MEX carries the Mars Advanced Radar for Subsurface and Ionospheric Sounding (MARSIS). The MARSIS radar has two modes of operation: the Active Ionospheric Sounder (AIS) mode (Gurnett et al., 2005) and the SubSurface (SS) mode (Picardi et al., 2005). In the AIS mode the radar transmits stepped pulses from 0.1 to 5.4 MHz. The remote ionospheric measurements are based on the reflection of the
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
transmitted waves in the ionosphere where the wave frequency pffiffiffiffiffi meets the local plasma frequency (f p ne , where ne is the electron number density). Such a method allows to retrieve the altitude profile of the ionospheric electron number density down to the altitude of the main ionospheric peak ðh 130 kmÞ (Gurnett et al., 2005; Morgan et al., 2008). The sounding measurements have shown that the electron density in the main ionospheric layer at 140 km is in reasonable agreement with the Chapman model while at higher altitudes ( 200 km) a possible second layer probably associated with O þ ions appears. Besides that, when the transmitted frequency passes through the local electron frequency, strong electrostatic oscillations at fp are excited in the ionosphere and can be measured giving us an information about the local ionospheric density (Gurnett et al., 2005; Duru et al., 2008; Andrews et al., 2013). In the subsurface (SS) mode of the MARSIS operation, the radar transmits wideband (1 MHz) pulses at four centered frequencies (1.3, 3, 4 and 5 MHz). Traversing the ionosphere the waves are distorted by a frequency-dependent change of the phases due to dependence of the wave speed on the frequency. The distortions R1 depend on the total electron content TEC ¼ 0 nð zÞ dz which can be retrieved from the echo wave signal (Safaeinili et al., 2007; Mouginot et al., 2008). Safaeinili et al. (2003) have analyzed the TEC data obtained from June 2005 to September 2006. Assuming that the signal distortion is caused mostly by the ionospheric content below 200 km where the approach of the photochemical equilibrium is valid, and using the Chapman model (Chapman, 1931) Safaeinili et al. (2003) have derived a mean scale height of 11 .5 km and the mean value of the peak number density of 2:1 105 cm 3 at the subsolar point (solar zenith angle SZA ¼ 01). While the ionosphere at h r200 km is in photochemical equilibrium, deviations from the Chapman model become essential at larger altitudes. Duru et al. (2008) have shown that the ionospheric densities above 300 km varies with solar zenith angle not according to Chapman theory as cos ðSZAÞ1=2 implying a dominance of horizontal transport processes over photochemistry. Since the ionospheric thermal pressure above 300 km is insufficient to endure the dynamic pressure exerted by the solar wind, the ionosphere occurs magnetized and dynamics of the upper ionosphere is controlled the by the solar wind (Shinagawa and Cravens, 1989; Dubinin et al., 2008). The nightside ionosphere of Mars occurs highly variable. It is generally thought that plasma transport from the dayside and ionization of the atmosphere by precipitating electrons are the most important sources of the nightside ionosphere (Haider et al., 1992; Fox et al., 1993; Fillingim et al., 2010; Lillis et al., 2009, 2011). Analyzing the radio occultation measurements at the nightside of Mars made onboard MEX Withers et al. (2012a) have shown that transport from the dayside is a main source at SZA r 1151 but at higher SZA electron precipitation probably dominates. The existence of strong localized sources of crustal magnetic field at Mars (Acuna et al., 1999) introduces new important features. Analyzing the radio occultation measurements by MGS, Ness et al. (2000) have observed that scale heights of the electron number density were much larger above regions of locally vertical magnetic field. Ness et al. (2000) have assumed that reconnection between the crustal magnetic field and draping IMF may lead to penetration of solar wind electrons to low altitudes and heating of the ionosphere electrons. Increase in the electron temperature can decrease the efficiency of the dissociative recombination of the major ion species ðO2þ Þ and increase in the density scale height (Krymskii et al., 2003). Nielsen et al. (2007) have reported about cases of the peak density enhancements in spatially limited regions with strong crustal magnetic field. Neither solar events or precipitation of
63
energetic electrons were observed during these events. Nielsen et al. (2007) have suggested that plasma instabilities driven by solar wind interaction in minimagnetospheres of Mars might be responsible for electron heating and density enhancement. Safaeinili et al. (2007) have observed that at the nightside ionosphere of Mars, TEC is higher over the region with vertical crustal magnetic field although the enhancements in TEC were also measured in the areas where there was no evident connection with crustal sources. Analyzing MARSIS ionograms in AIS mode Němec et al. (2010) have observed that the occurrence rate of the nightside ionosphere is more than 4 times larger in regions with nearly vertical crustal magnetic field. The measurements of low energy ions carried by the Ion Mass Analyzer (IMA) onboard MEX have shown that the dayside ionosphere over the regions with strong crustal magnetization occurs more inflated (Dubinin et al., 2012). These results were confirmed by the measurements of the local plasma density retrieved from the frequencies of electron plasma oscillations (Andrews et al., 2013). Andrews et al. (2015) have provided an empirical model of effects of the crustal magnetic field on plasma density in the topside ionosphere. Cartacci et al. (2013) have analyzed variations of TEC in the nightside ionosphere based on MARSIS data from 2006 to 2011. To quantify the TEC variations they did not used average values as was done by Safaeinili et al. (2007) but 10th order polynomial fits. Cartacci et al. (2013) have also observed that the increase in TEC (positive δ (TEC)) was often related to the regions with vertical crustal fields. Fillingim et al. (2010) and Lillis et al. (2011) have modeled effects of precipitating electrons into the crustal field area on the nightside. These studies have found a large variability in peak ionization rate between different geographic regions. According to Fillingim et al. (2010) ionization patches produced by the precipitating electrons with densities up to 3 104 cm 3 at an altitude of about 140 km appear. In this paper we study properties of the Martian dayside and nightside ionosphere observed by MEX spacecraft by combining observations by the MARSIS radar and ASPERA-3 particle spectrometer. The measurements of the total electron content (TEC) provide us the information about the ionosphere near the main ionospheric peak. The in-situ measurements by MARSIS of the electron number density at altitudes from 300 km to 1400 km tell us about the characteristics of the topside ionosphere. These data are complemented by the measurements of low-energy ion and electron fluxes by ASPERA-3. We focus on effects of the planetary crustal magnetic field on the ionosphere of Mars.
2. Observations 2.1. Instrumentation We use the MARSIS-TEC data from June 2005 to August 2010 derived using the method by Mouginot et al. (2008). Note also that Cartacci et al. (2013) have applied another method (‘contrast method’) of compensating the distortions in the phase shift caused by the Martian ionosphere. This method allowed processing the MARSIS data in the nightside ionosphere with an accuracy of 10% . Local electron number densities can be derived by measuring the spacing between plasma frequency harmonics generated by distortion in the preamplifier due to the large amplitude of the emitted wave (Duru et al., 2008). Andrews et al. (2013) have developed a method for the automatic retrieval of local plasma density from a large data set of the MARSIS data in the AIS mode. In this paper, we use this database of local density derived from the observations between June 2005 and May 2014.
64
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
The MARSIS data are complemented by the ASPERA-3 measurements. The Ion Mass Analyzer (IMA) is a part of the ASPERA-3 instrument (Barabash et al., 2006). It determines the composition, energy and angular distribution of ions with energy r 30 keV. A new patch uploaded on May 1, 2007 expanded the energy range down to 10 eV. However, in the ionosphere, IMA is able to detect ions with lower energy as the spacecraft potential in this region usually varies in the range from 5 to 10 V. This shifts the ion energy spectrum to higher energy channels. At Ei 4 50 eV, IMA measures fluxes with a time resolution of 192 s and with the field of view of 901 3601. At lower energies ðEi r 50 eVÞ the measurements are carried out without elevation scanning that decreases field-of-view to 41 3601 but increases time resolution to 12 s. The electron sensor (ELS) measures electron fluxes in the energy range of 5 eV–20 keV with a field-of-view of 41 3601 and a time resolution of 4 s. 2.2. Presentation of the data Fig. 1a shows the values of TEC as a function of the solar zenith angle (SZA). The color bar presents the number of samples in each bin (1° bins in SZA). One can see that the sampling at low zenith angles is rather poor. Variations in TEC are caused by several factors. One of them is the varying distance between Sun and Mars that affects the solar UV flux which impacts Mars. Fig. 1b presents the same data set with color corresponding to the Sun-Mars distance. It is observed that TEC increases with decrease of the distance although some anomal features, for example, at SZA 60° and 100°, are observed too, which might be related to other factors controlling the ionospheric conditions. Fig. 1c compares the TEC measurements with solar irradiance. We used here the data of the Solar EUV Monitor (SEM) on the Solar Heliospheric Observatory (SOHO) which measures EUV and soft X-ray solar irradiance since 1996. SEM measures absolute EUV fluxes in the band 0.1–50 nm. Solar fluxes from the SOHO spacecraft were scaled in the intensity to the distance of Mars and rotated to Mars position. Note that the SEM data are the source of the S10.7 solar irradiance proxy which is widely used for analysis of TEC. A dependence of TEC on solar irradiance in Fig. 1c is evident. To investigate dependence of TEC on other factors we need to normalize TEC values to EUV flux and solar zenith angle. The dependence of TEC on solar zenith angle follows well the Chapman (Chapman, 1931) theory. Fig. 2 shows the mean values of TEC
R1 and its standard deviations. A Chapman curve TEC ¼ no 0 exp p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi h ho dh ¼ no H 2π e cos ðSZAÞ is shown 1=2 1 h Hho secðSZAÞe H by the green curve. Here h is the altitude, ho is the altitude of the peak density, H is the scale height, SZA is the solar zenith angle, and no is the peak electron density at SZA ¼ 0. A reasonable fitting between the data and the Chapman curve (no ¼ 1:7 105 cm 3 and H ¼11.5 km) calculated without taking into account of grazing incidence of solar radiation is observed for the dayside ionosphere ðSZA r951Þ except for the region around SZA 501 where a dip in the TEC occurs. Therefore we use the given Chapman function to normalize TEC values to SZA. Note that the value of H is close to the value obtained by Safaeinili et al. (2007) from a smaller dataset though the value of no is slightly lower. To normalize TEC values to EUV flux we study variations of TEC normalized to SZA on solar irradiance (Fig. 3). Here we used the SEM data from SOHO in two passbands (0.1–50 nm and 27– 34 nm) (see panels (a) and (b)) and the data from Solar EUV Experiment (SEE) of the Thermosphere, Ionosphere, Mesosphere, Energetics and Dynamics (TIMED) satellite in the range of 27–34 nm (panel (c)) and in the passband about 30.4 nm HeII
Fig. 2. Mean values of TEC and its standard deviations. Dashed green line shows a Chapman curve (no ¼ 1:7 105 cm 3 , H ¼ 11:5) fitted to the data. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
Fig. 1. TEC as a function of SZA. (a) Color shows a number of the measurements in each bin, (b) color shows the distance (in AU) between Sun and Mars, (c) color shows solar irradiance measured by the SEM instrument onboard SOHO and translated and rotated to the Mars position. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
65
Fig. 3. TEC values normalized to SZA and its standard deviations as a function of solar irradiance measured on the Earth orbit by different instruments on different satellites and translated to Mars: (a) SEM/SOHO data in the passband 0.1–50 nm, (b) SEM/SOHO data in the passband 27–34 nm, (c) SEE/TIMED data in the 30.4 HeII line. Red lines show power-law fitting to the data. Exponents of fitted curves are also given. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
Fig. 4. Comparison between TEC-Solar irradiance dependencies using the measurements of solar radiation by SEE/TIMED and SDO.
line (panel (d)). Mean values of TEC and its standard deviations are shown in black. Red lines are power-law fitting curves. Colored bins present the number of samples. Power exponents which characterize the slope of the curves occur different and generally differ from the Chapman predicted value k ¼0.5. The difference is partly related to the uncertainties in EUV measurements made by different instruments (see e.g. Peterson et al., 2012; Wieman et al., 2014) and different spectral efficiency in production of photoelectrons. Note that Lillis et al. (2010) have compared TEC on Mars with solar EUV flux in the 30.4 nm HeII line for the period between June 2005 and September 2006 and found that it varies as power law with exponent equal to 0.54. Only with launch of Solar Dynamic Observatory (SDO) high spectral resolution observations of solar irradiance in the passband 1–100 nm become available since April 2010. Fig. 4 compares TEC values as function of solar irradiance measured by TIMED (30.4 nm line) and SDO in the period between August 2010 and August 2012. The power-law exponent 0.52 for the relationship between TEC and solar flux by SDO is in approximate agreement with Chapman theory while for TIMED we observe a somewhat higher value of the exponent (0.62). In this
paper we use the EUV data only to normalize empirically TEC values and do not focus on analysis of exact relationship between TEC and solar fluxes keeping in mind that there are always uncertainties associated with a phase shift of EUV irradiance measured near Earth and at Mars (see also Girazian and Withers, 2015). 2.3. Effect of crustal field on TEC Fig. 5 shows TEC values normalized by EUV flux as a function of solar zenith angle (SZA). Colored bins in panel (a) depict the median values of the latitude at which the measurements were done. A remarkable asymmetry between the northern hemisphere and the southern hemisphere, where the most strong crustal sources are placed, is seen at SZA 601–1001. TEC at the dayside increases while moving from the northern to the southern latitudes. Colored bins in panel (b) show the median values of the crustal magnetic field at altitude of 150 km derived from the Cain's model (Cain et al., 2003). We observe that at SZA 601–1001 lower TEC values are over the regions with weak crustal field ðB r 10–20 nTÞ. On the nightside, the picture is more controversial.
66
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
Fig. 5. TEC values normalized to EUV fluxes as functions of SZA. (a) Color shows the mean planetocentric latitude of the observations in each bin, (b) color shows the value of the crustal magnetic field at altitude of 150 km derived from Cain's model. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
Over the regions with strong crustal field we observe higher as well as very low TEC values. Let us consider variations in TEC with crustal field in more detail. Fig. 6 shows the mean values of TEC on the dayside, at solar zenith angles SZA Z 601, where statistics of sampling is higher (see Fig. 1), as functions of the total value of the field (Bt), the radial (Br) and the horizontal (Bh) components, respectively. The standard deviations are shown only on Bt TEC plots to avoid overloading of Fig. 6. The trend of lower TEC in the ionosphere with low strength of the crustal field is observed. Such a trend is seen at B r 50 nT and not only for the total value of the magnetic field but also for the vertical and horizontal components that conceal a dependence on the magnetic field inclination. The two bottom rows show the dependences of TEC on the angle of the field inclination at the altitude of 150 km. The data are sorted by the value of the crustal field at 400 km – weak ð r 20 nTÞ and strong ð 4 20 nTÞ crustal fields, respectively. Note that the value 20 nT which separates here two different regimes of the crustal magnetization is rather arbitrary. It is observed that at SZA Z 751 a trend of a decrease in TEC at vertical orientation of the field (Mag. inclination 01 or 180°) appears. At smaller SZA we do not see a dependence of TEC on the field direction. At B Z 50 nT variability significantly increases exposing higher as well as lower TEC values although a trend of a slight increase in TEC with magnetic field value is still visible. On average, the increase in TEC can reach 20%. A large variability observed at high magnetic fields is partly due to a poorer sampling. Similar trends in TEC are observed on the nightside close to the terminator ðSZA r 1051Þ (Fig. 7) since altitudes near 150 km at which a typical ionospheric density peak is located remain sunlit at SZA r1071. At larger SZA we observe a different picture (Fig. 8). TEC increases in regions with small ðB r 50 nTÞ values of the crustal field without a visible relation to the values of the vertical (Br) component. The right panel on the middle shows a dependence of TEC on magnetic inclination angle without separation between regions with weak and strong crustal fields. One can see that TEC somewhat increases in the regions with a dominance of the vertical component. The bottom panels depict such a dependence in the ionosphere with weak (B r20 nT at 400 km) and strong
(B 4 20 nT at 400 km) field. We also increase the range of SZA (110°–140°) for better statistics. In the ionosphere with strong crustal field the dependence on the field line inclination becomes more pronounced. At vertical direction of the field lines, TEC increases up to 2 times. In the regions with weak crustal field dependence of the field orientation becomes weaker and almost vanishes. 2.4. Effect of crustal field on upper ionosphere Since upward diffusion and horizontal transport become more important with increase in altitude and these motions should be affected by the crustal sources we expect a more distinct relation between the ionosphere above 200 km and crustal fields. The upper panels in Fig. 9 show the local electron number density derived from analysis of waves excited by MARSIS pulses as a function of altitude above 300 km and crustal field (Bt ; Br ; Bh and magnetic inclination angle) derived from the MGS magnetic field map at 400 km (Connerney et al., 2001). The MARSIS data were processed for the period of 2005– 2014. Lower panels show the mean values of the electron density and the standard deviations measured on the dayside, at altitudes 350–450 km and 450–600 km, respectively. On the nightside MARSIS has operated mostly in SS mode and the local measurements are rather scarce. It is seen that the electron density generally increases with increase in the magnetic field strength at all altitudes sampled by MEX (see also Dubinin et al., 2012; Andrews et al., 2013, 2015). Such a trend is also observed in the density variations with variations in the vertical and horizontal components of the crustal field. In regions with strong magnetic field the density increases approximately by a factor 3. Such a behavior is similar to the relationship between TEC and crustal field. The dependence on the field line inclination at altitudes o 350 km contains an enhancement in the number density at angles near 90° that was also typical for TEC variations at large SZA. At higher altitudes the opposite trend develops. Higher values of the density are observed at the vertical orientation of the field lines. We also see that on the color maps of Fig. 9 the number density in some bins at high values of the vertical field component ðBr Þ is lower than in bins at lower field values, i.e. at very large magnetic fields the electron density again decreases. The latter might be
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
67
Fig. 6. Mean values of TEC at different SZA on the dayside ionosphere as functions of the total value of the crustal magnetic field (model), the vertical component Br, the horizontal component Bh, and inclination angle of the magnetic field.
68
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
Fig. 7. Similar to Fig. 6 plots for SZA ¼ 901–1051.
Fig. 8. Mean values of TEC at different SZA on the nightside ðSZA 41101Þ as a function of crustal magnetic field parameters at altitude of 150 km.
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
69
Fig. 9. Upper panels show maps of local electron number density at different altitudes as function of crustal field parameters measured at altitude of 400 km by MGS. Lower panels present mean values of the electron number density and its standard deviations as function of Bt, Br, Bh and the inclination angle at altitudes of 350–450 km and 450– 600 km, respectively. (For interpretation of the references to color in this figure, the reader is referred to the web version of this paper.)
related to more easy escape of the ionospheric plasma in regions where the crustal field can directly interact with the magnetosheath plasma opening access of solar wind plasma to the ionospere and escape of the ionospheric plasma into space. This is confirmed by the case studies. Fig. 10a shows an example of the measurements made on a single MEX orbit crossing the region of strong crustal field. The top panel presents the local electron density. Second and third panels depict the vertical component of the crustal field derived from the Cain's model and the energy–time spectrogram of electron fluxes measured by the ELS/ASPERA-3 instrument, respectively. Sharp drops in the local electron density are observed in the regions where the magnetosheath electrons intrude into the ionosphere marking an openness of the field lines. It is interesting to note that these density depressions occur rather stable in time. Fig. 11 shows two sets of observations on several subsequent orbits crossing two almost identical regions of strong crustal fields in the southern hemisphere (longitude 187° – left panels and 180° – right panels). It seems that some dips in the density survive from orbit to orbit over several days.
We do not know how deep in altitude these density depressions extend and whether they are visible in the total electron content. Since MARSIS operates in AIS and SS modes in turn we cannot compare TEC with local number densities at higher altitudes. However we can tentatively compare TEC data with the ELS measurements. Fig. 10b compares TEC variations with the electron fluxes and the radial component of the crustal magnetic field on the MEX orbit sampling the same region of the ionosphere as one in Fig. 10a. We observe similar features of the electron intrusion as in Fig. 10a, accompanied by small dips in TEC in the areas where there were depressions in the local electron density (Fig. 10a). However, it is difficult to confirm a potential connection between such depressions in TEC and the number density dips at altitudes above 300 km without a correct subtraction of the average TEC values. The local measurements of the density at lower ð r 300 kmÞ altitudes can clarify a connection between lower and upper ionosphere. It is worth noting that the existence of the ionospheric density dips is also confirmed by the ELS/ASPERA-3 measurements. Two narrow energy bands at 20 eV on the energy-time spectrogram of electron fluxes shown in Fig. 12 are associated with photoelectron
70
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
Fig. 10. (a) From top to bottom: the electron number density, the vertical component of the model crustal field, the spectrogram of electron fluxes. Dips in the density are observed in the regions where sheath electrons intrude into the ionosphere. (b) Comparison of TEC variations with the radial field component and electron fluxes on the MEX orbit sampling the same area as in Fig. 11a.
Fig. 11. Data obtained on the subsequent orbits sampling the same regions of the crustal magnetic field. Some of the density dips are repetitively observed on each orbit implying their rather stable subsistence. Electron density is in black. Components (total, radial and horizontal) of the model magnetic field are in color. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
peaks due to absorption of the strong HeII solar line at 30.4 nm in the carbon dioxide dominated atmosphere (Frahm et al., 2006). Peaks are shifted to low energies on the spectrograms by a negative spacecraft potential. The crimson colored curve depicts the number of counts in the energy range 14–22 eV. We observe that depressions in the intensity of the ‘CO2 photoelectron lines’ correlate with the dips in the total electron density measured by MARSIS implying a possible ionospheric escape from such areas. Note that the data in Fig. 9 were summed over all SZA and therefore variations with SZA were missed. A map of the local number density at h r 350 km and 350 o h r450 km at different SZA as a function of Br and Bh measured by MGS at 400 km is shown in Fig. 13. An increase in ne with increase in Br is observed at all SZA although at Br 4100 200 nT the number densities are less than at Br r 100 nT. A similar trend at all SZA is observed for the dependence on the horizonal magnetic field Bh. If we look on the relationship between the local number density in the upper
ionosphere and the magnetic field inclination at different SZA (Fig. 13 lower panels) we see that at smaller SZA the number density is higher where the crustal field is more vertical. Such a dependence vanishes at higher SZA even with the opposite, though weak, trend of decreasing density in the regions with vertical field inclination, i.e. similar to what we observe for TEC variations at SZA Z 701.
3. Discussion and conclusions 3.1. Dayside ionosphere Systematic measurements of TEC on the dayside were made only at SZA Z 551. At smaller SZA the data are rather occasional. It is shown that TEC values normalized to solar radiation fluxes over the ionosphere with crustal sources increase with the strength of
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
the magnetic field. This trend is already observed at low values of the crustal field ( r 50 nT at altitude of 150 km) and remains in the regions with stronger magnetic fields although the variability in TEC significantly increases. Similar trends observed in relationships between TEC and crustal fields, and between local number densities in the upper
Fig. 12. Same as in Fig. 11a. Dips in the ionospheric number density are also displayed as depressions in the fluxes of CO2 photoelectrons (crimson colored curve on the bottom panel). (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
71
ionosphere ð Z300 kmÞ and crustal fields motivated us to evaluate the contribution to TEC from different altitude layers. Contribution from altitudes above 300 km can be evaluated using the empirical model for the upper ionosphere by Němec et al. (2011) and Andrews et al. (2015). Both models have used the same expoðh ho Þ
nential dependence ne ðh; SZAÞ ¼ neo ðSZAÞe HðSZAÞ , where neo is the number density at ho ¼325 km and the scale height H are parametrized values of SZA (Andrews et al., 2015). Taking, for example, SZA ¼ 601 yields neo ¼ 1044 (1892) cm 3 and H¼ 110.4 (176) km and contribution to TEC of 1:15 1014 ð3:2 1014 Þ m 2 for areas with weak and strong crustal fields, respectively. This is only 2% of the typical TEC values in these regions (see Fig. 6). Thus the observed TEC variations (up to 20%) related to the crustal field probably occur at altitudes below 325 km. MARSIS soundings have shown that the height profile of the ionosphere on the dayside often deviates from the Chapman profile at altitudes just few tens of km above the main ionospheric peak forming layers at altitudes from 180 km to 300 km (Kopf et al., 2008). The multifacetal shaping of the dayside ionosphere which is very different from the classical Chapman one was also emphasized by Withers et al. (2012b). Integrating the ne(h) profile, presented by Kopf et al. (2008, Figure 5), in the altitude range between 135 and 391 km yields to the column density 6:96 1015 m 2 . Yields in the altitude intervals of 180–391 km and 225–381 km are 2.2 10 15 m 2 and 0.74 10 15 m 2, i.e. 31.6% and 10.6%, respectively. Thus variations in TEC related to crustal fields on the dayside probably appear at h Z 225 km. At altitudes above 200 km and magnetic field B 30–50 nT, the frequency of ion-neutral collisions νin is less than the ion gyrofrequency ωBi, and the frequency of electron-neutral collisions is much less than the electron gyrofrequency ΩBe.
Fig. 13. Upper panels (a, b): maps of the local electron density at h r 350 km (a) and 350 o h r 450 km (b) at different SZA as function of the Br and Bh field components of the crustal field at 400 km. Lower panels (c, d) show maps of the local electron number density as a function of the magnetic field inclination. Curves depict dependences of ne on the field inclination at SZA ¼ 201 451 and SZA ¼ 751 851 at altitudes h r 350 km and 350 o h r 450 km, respectively.
72
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
Correspondingly, the ambipolar diffusion of plasma across the A , where DA is magnetic field is strongly reduced: D ? 1 þ ωBi ΩDBe =νin νen the coefficient of the ambipolar diffusion without magnetic field or along the field lines. Below 200 km, ions become demagnetized ðνin 4 ωBi Þ while electrons remain magnetized down to the altitude of the main ionospheric peak. With increase in the value of the magnetic field the ion demagnetization occurs at lower altitudes. Ion demagnetization can enlarge the diffusion even across the magnetic field. Moving freely upward, along the closed crustal magnetic field lines, electrons produce an excess of the negative charge. The resulting ambipolar electric field will push nonmagnetized ions upward across the magnetic field. The coefficient of ambipolar diffusion in this case will be determined by ion dynamics D ? 1 þ ωDBiA =νin DA . Despite of this local altitude effect, a horizontal crustal magnetic field generally, strongly reduces the mobility of ions and electrons and therefore one could expect an increase in the electron content with increase in the magnetic field strength. If the field is stronger, then the locally produced plasma and, in particularly, O þ ions originating at higher altitudes are less exposed to vertical and horizontal motions and their column density increases. Note, that even in the unmagnetized Martian ionosphere inclusion of the ion dynamics essentially modifies the ion height profiles above 170–180 km (Chaufray et al., 2014). A dependence on the field inclination should also appear since the ionospheric plasma can more easily rise up along the field lines with a subsequent horizontal transport driven by solar wind. This might be a cause of decrease in TEC for vertical field inclination at SZA Z 751. On the other hand, downward ion fluxes pushed by solar wind pressure could produce the opposite trend of TEC increase at the regions of vertical field inclination. Such a dependence could be expected rather at smaller SZA where the effect of solar wind is stronger. Unfortunately, we cannot examine this effect on TEC at low solar zenith angles because of a poor sampling (see Fig. 1a). Fig. 14 shows a map of the radial velocity of cold oxygen ions ðEi r50 eVÞ measured by IMA/ASPERA-3 as a function of SZA and altitude. We used the whole dataset from June 2007 to June 2015. Since the ion measurements of the cold component are carried out without electrostatic scanning in the elevation angle we select only the observations made for the intervals when the normal vector to Mars was in the plane ( 710°) of the IMA field of view. Since the measurements of the low-energy ions are also affected by the spacecraft potential we have chosen the average value of the potential value to be 8 V that is in general agreement with the ELS measurements of the peak energy of the ‘CO2 photoelectrons’. We observe that at lower SZA downward plasma motions
Fig. 14. Map of the planetocentric radial velocity of low-energy oxygen ions measured by IMA/ASPERA-3. A reversal of the flow direction is observed.
which could increase TEC, dominate while at higher SZA the flow occurs outward from the planet. Note also that a divergence of the ion fluxes can give rise to sharp gradients in the ionospheric number densities (‘ionopause’ signature (Duru et al., 2009)). 3.2. Nightside ionosphere On the nightside ðSZA Z 1101Þ we observe an opposite trend. At Bt r 50 nT at 150 km TEC increases with decrease in the magnetic field value. In this range TEC also almost does not vary with the variations in the Br-component. Such a behavior at small crustal fields can be easily interpreted assuming that the downward plasma transport is the main supplier of the nightside ionosphere. In the regions with weak crustal fields the downward transport is more efficient as compared to areas with strong crustal sources where transport and electron precipitation operate mainly along the field lines. This explains why increase in TEC for vertical field inclination becomes more profound with increase in the field strength (see Fig. 8). This trend was also noticed by Cartacci et al. (2013). Fig. 15 shows examples of observations made by MEX on four orbits on the nightside. Two top (bottom) panels show examples of the measurements made in the ionosphere with strong (weak) crustal field. Upper panels depict TEC values. Second and third panels present three components of the model crustal field at 150 km and angle of the magnetic field inclination, respectively. Lower panels show energy–time spectrograms of electron fluxes. Spikes in TEC correspond to the regions with a strong vertical component Br (blue lines) and are often accompanied by spikes of the magnetosheath/plasma sheet electrons penetrating into the ionosphere (top panels). It is reasonable to assume that these crustal field lines reconnect with the draped IMF and become open. Variability in TEC in such regions might be partly caused by sporadic processes of field reconnection. In the regions with weak crustal field (bottom panels), spikes in TEC are still observed. Although the value of the crustal field in these regions is small the appearance of spikes generally prefers the vertical orientation of the magnetic field. In this case, the assumption of open field lines is questionable and another mechanism of plasma transport seems to be operating. The observations by ASPERA-3 confirm the important role of plasma transport on the nightside. Because of the lack of the measurements with a proper attitude of the IMA sensor at low altitudes on the nightside we use another method. To overcome the problem of narrow field of view for detection of low-energy ions we use the method suggested by Nilsson et al. (2012) and Fränz et al. (2015) to construct the mean 3D-distribution in a given spatial bin around Mars by collecting many different 2Dobservations over long time in the Martian ionosphere. A detailed description of the derivation of the mean 3D – distribution functions and their moments is given in Fränz et al. (2015). Fig. 16a shows the dependence of the vertical velocity of O þ ions measured by IMA/ASPERA-3 instrument as a function of the altitude and the magnetic inclination angle. Although the uncertainty in the velocity values related to the effect of the spacecraft potential remains, it is seen that downward motions of low-energy plasma transported from the dayside prevail in the regions with vertical field orientation. An increase in the partial density of oxygen ions (still reduced as compared to the MARSIS observations due to a missing of the cold core population) at the vertical field orientation is also observed (Fig. 16b). Precipitation of the magnetosheath/plasma sheet electrons along the field lines is another important source supplying the nightside Martian ionosphere. Fig. 17a shows that TEC on the nightside ðSZA ¼ 1101–1401Þ increases with increase in the energy flux of electrons ðEe ¼ 30–500 eVÞ measured locally by ELS/
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
73
Fig. 15. Examples of MARSIS and ASPERA-3 observations made at the nightside. From top to bottom are TEC, three components of the crustal magnetic field (model), the inclination of the field line and energy-time spectrogram of electron fluxes. Upper (lower) panels show the TEC measurements made over the region with strong (weak) crustal field.
Fig. 16. (a) Map of the vertical velocity of oxygen ions at the nightside ðSZA ¼ 1101–1401Þ as a function of the angle of the magnetic field inclination and altitude. (b) Map of the ion density.
74
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
Fig. 17. (a) TEC on the nightside as a function of the energy flux of the electrons ðEe ¼ 30 500 eVÞ measured by ELS/ASPERA at the point of the radar operation. (b) TEC as a function of the electron fluxes on the dayside.
ASPERA-3. Note that on the dayside, at high SZA we often observe the opposite trend (Fig. 17b) due to losses along the field lines. Because of the lack of magnetic field observations at low altitudes we did not evaluate the role of the magnetic field of solar wind origin which permeates the ionosphere of Mars. This field is mainly horizontal and transported by the convection of the ionospheric plasma. It is reasonable to assume that in the regions with weak crustal sources the magnetic field is governed by the draped horizontal IMF. In contrast, the role of crustal field increases with their strength. Presently, there are no reliable measurements of the ionospheric motions and adequate models for the plasma flow in the magnetized ionosphere of Mars. A twodimensional model for the magnetized ionosphere of Venus by Shinagawa (1996) predicts a planetward motion (down to altitude of 180 km) at low SZA ðSZA r 501–601Þ and upward convective flows at higher SZA. Correspondingly, if this model is applicable to Mars (see Fig. 14) one could expect an increase (decrease) in TEC at low (high) SZA, respectively, in the regions with weak crustal sources where the ionospheric flows driven by solar wind dominate. In conclusion, we note that the coordinated measurements by MARSIS and ASPERA-3 and the measurements by the plasma instruments LPW, STATIC, SWEA and NGIMS, onboard MAVEN will provide a powerful tool for further studies of the Martian ionosphere and its response to different factors. The in-situ observations at altitudes not surveyed by MEX will allow us to identify the links between different ionospheric regions and their role in the ionospheric dynamics.
Acknowledgments Authors (ED, MF) wish to acknowledge support from DFG for supporting this work by Grant WO 910/1-1 and DLR by grants 50QM99035, FKZ 50 QM 0801 and 50QM1302. DJA was supported by Grants from Swedish National Space Board (DNR 162/14) and the Swedish Research Council (DNR 621-2014-5526).
References Acuna, M.H., et al., 1999. Global distribution of crustal magnetization discovered by the Mars Global Surveyor MAG/ER experiment. Science 284 (5415), 790–793. Andrews, D.J., et al., 2013. Determination of local plasma densities with the MARSIS radar: asymmetries in the high-altitude Martian ionosphere. J. Geophys. Res. 118, 6188–6196. http://dx.doi.org/10.1002/jgra.50593. Andrews, D.J., et al., 2015. Control of the topside Martian ionosphere by crustal magnetic fields. J. Geophys. Res. 120, 3042–3058. http://dx.doi.org/10.1002/ 2014JA020703. Barabash, S., et al., 2006. The analyzer of space plasmas and energetic atoms (ASPERA-3) for the Mars Express mission. Space Sci. Rev. 126, 113–164. http: //dx.doi.org/10.1007/s11214-006-9124-8.
Cain, J.C., Ferguson, B.B., Mozzoni, D., 2003. An n¼ 90 internal potential function of the Martian crustal magnetic field. J. Geophys. Res. 108, 5008. http://dx.doi.org/ 10.1029/2000JE001487. Cartacci, M., et al., 2013. Mars ionosphere total electron content analysis from MARSIS subsurface data. Icarus 223, 423–437. Chapman, S., 1931. Absorption and dissociative or ionizing effect of monochromatic radiation in an atmosphere on a rotating earth. Proc. Phys. Soc. (London) 43, 1047–1055. Chaufray, J.-Y., et al., 2014. Three-dimensional Martian ionosphere model: II. Effect of transport processes due to pressure gradients. J. Geophys. Res. 119, 1614–1636. http://dx.doi.org/10.1002/2013JE004551. Connerney, J.E.P., et al., 2001. The global magnetic field of Mars and implications for crustal evolution. Geophys. Res. Lett. 28, 4015–4018. http://dx.doi.org/10.1029/ 2001GL013619. Dubinin, E., et al., 2008. Plasma environment of Mars as observed by simultaneous MEX-ASPERA-3 and MEX-MARSIS observations. J. Geophys. Res. 113, A10217. http://dx.doi.org/10.1029/2008JA013355. Dubinin, E., et al., 2012. Upper atmosphere of Mars is not axially symmetrical. Earth Planets Sp. 64, 113–120. http://dx.doi.org/10.5047/eps.2011.05.022. Duru, F., et al., 2008. Electron densities in the upper ionosphere of Mars from the excitation of electron plasma oscillations. J. Geophys. Res. 113, A07302. http: //dx.doi.org/10.1029/2008JA013073. Duru, F., et al., 2009. Steep, transient density gradients in the marian ionosphere similar to the ionopause at Venus. J. Geophys. Res. 114, A12310. http://dx.doi. org/10.1029/2009JA014711. Fillingim, M.O., et al., 2010. Localized ionization patches in the nighttime ionosphere of Mars and their electrodynamic consequences. Icarus 206, 112–119. http://dx.doi.org/10.1016/j.icarus.2009.03.005. Fox, J.L., 2004. Advances in the aeronomy of Venus and Mars. Adv. Sp. Res. 33, 132–139. Fox, J.L., Brannon, J.F., Porter, H.S., 1993. Upper limits to the nightside ionosphere of Mars. Geophys. Res. Lett. 20, 1339–1342. http://dx.doi.org/10.1029/93GL01349. Frahm, R., Winningham, J.D., Sharber, J.R., et al., 2006. Carbon dioxide photoelectron energy peaks at Mars. Icarus 182, 371. Fränz, M., et al., 2015. Cold ion escape from the Martian ionosphere. Planet. Space Sci. 119, 92–102. http://dx.doi.org/10.1016/j.pss.2015.07.012. Girazian, Z., Withers, P., 2015. An empirical model of the extreme ultraviolet solar spectrum as a function of F 10:7 . J. Geophys. Res. 120, 6779–6794. http://dx.doi. org/10.1002/2015JA021436. Gurnett, D.A., et al., 2005. Radar soundings of the ionosphere of Mars. Science 310, 1929–1933. http://dx.doi.org/10.1126/science.1121868. Haider, S.A., et al., 1992. Calculated ionization rates, ion densities, and airglow emission rates due to precipitating electrons in the nightside ionosphere of Mars. J. Geophys. Res. 97, 10,637–10,641. http://dx.doi.org/10.1029/92JA00317. Haider, S.A., Mahajan, K.K., Kallio, E., 2011. Mars ionosphere: a review of experimental results and modeling studies. Rev. Geophys. 49, RG4001. http://dx.doi. org/10.1029/2011RG000357. Hanson, W.B., Satani, S., Zuccaro, D.R., 1977. The Martian ionosphere as observed by the Viking retarding potential analyzers. J. Geophys. Res. 82, 4351–4363. http: //dx.doi.org/10.1029/JS082i028p04351. Hinson, D., et al., 1999. Initial results from radio occultation measurements with Mars Global Surveyor. J. Geophys. Res. 104, 26997–27012. Kliore, A., et al., 1965. Occultation experiment: results of the first direct measurement of Mars'atmosphere and ionosphere. Science 149, 1243–1248. Kliore, A., 1992. Radio occultation observations of the ionospheres of Mars and Venus, in Venus and Mars: Atmospheres, Ionospheres and Solar Wind Interactions. In: Luhmann, J., Tatrallyay, M., Pepin, R. (Eds.), Geophysical Monograph Series, 66. AGU, Washington, DC, pp. 265–276. http://dx.doi.org/10.102 9/GM066p0265. Kopf, A.J., et al., 2008. Transient layers in the topside ionosphere of Mars. Geophys. Res. Lett. 35, L17102. http://dx.doi.org/10.1029/2008GL034948. Krasnopolsky, V.A., 2002. Mars' upper atmosphere and ionosphere at low, medium and high solar activities: implication for evolution of water. J. Geophys. Res. 107 (E12), 5128. http://dx.doi.org/10.1029/2001JE001809. Krymskii, A., et al., 2003. Effect of crustal magnetic fields on the near terminator ionosphere at Mars: comparison of in situ magnetic field measurements with
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
the data of radio science experiments on board Mars Global Surveyor. J. Geophys. Res. 108, http://dx.doi.org/10.1029/2002JA009662. Lillis, R.J., et al., 2009. The nightside ionosphere of mars: modeling the effects of crustal magnetic fields and electron pitch angle distributions on electron impact ionization. J. Geophys. Res. 114, E11009. http://dx.doi.org/10.1029/ 2009JE003379. Lillis, R.J., et al., 2010. Total electron content in the Mars ionosphere: temporal studies and dependence on solar EUV flux. J. Geophys. Res. 115, A11314. http: //dx.doi.org/10.1029/2010JA015698. Lillis, R.J., Fillingim, M.O., Brain, D.A., 2011. Three-dimensional structure of the Martian nightside ionosphere: predicted rates of impact ionization from Mars Global Surveyor magnetometer and electron reflectometer measurements of precipitating electrons. J. Geophys. Res. 116, A12317. http://dx.doi.org/10.1029/ 2011JA016982. Mendillo, M., et al., 2011. Modelling Mars' ionosphere with constrains from sameday observations by Mars Global Surveyor and Mars Express. J. Geophys. Res. 116, A11303. http://dx.doi.org/10.1029/2011JA016865. Morgan, D.D., et al., 2008. Variation of the martian ionospheric electron density from Mars Express radar soundings. J. Geophys. Res. 113, A09303. http://dx.doi. org/10.1029/2008JA013313. Mouginot, J., et al., 2008. Correction of the ionospheric distortion on the MARSIS surface sounding echoes. Planet. Sp. Sci. 56, 917–926. Němec, F., et al., 2010. Night side ionosphere of mars: radar soundings by the Mars Express spacecraft. J. Geophys. Res. 115, E12009. http://dx.doi.org/10.1029/ 2010JE003663. Němec, F., et al., 2011. Dayside ionosphere of mars: empirical model based on data from the MARSIS instrument. J. Geophys. Res. 116, E07003. http://dx.doi.org/ 10.1029/2010JE003789. Ness, N.F., et al., 2000. Effects of magnetic anomalies discovered at Mars on the structure of the Martian ionosphere and solar wind interaction as follows from radio occultation experiments. J. Geophys. Res. 105, 15991–16004. http://dx. doi.org/10.1029/1999JA000212. Nielsen, E., et al., 2007. Vertical sheets of dense plasma in the topside Martian ionosphere. J. Geophys. Res. 112, E02003. http://dx.doi.org/10.1029/ 2006JE002723. Nilsson, H., et al., 2012. Ion distributions in the vicinity of Mars: signatures of heating and acceleration processes. Earth Planets Sp. 64, 135–148. http://dx.doi. org/10.5047/eps.2011.04.011.
75
Orosei, R., et al., 2015. Mars advanced radar for subsurface and ionospheric sounding (MARSIS) after nine years of operation: a summary. Planet. Sp. Sci. 112, 98–114. http://dx.doi.org/10.1016/j.pss.2014.07.010. Pätzold, M., et al., 2005. A sporadic third layer in the ionosphere of mars. Science 5749, 837–839. Peterson, W.K., et al., 2012. Solar EUV and XUV energy input to thermosphere on solar rotation time scales derived from photoelectron observations. J. Geophys. Res. 117, A05320. http://dx.doi.org/10.1029/2011JA017382. Picardi, G., et al., 2005. Radar soundings of the subsurface of mars. Science 310, 1925–1928. Safaeinili, A., et al., 2003. Impact of mars ionosphere on orbital radar sounder operation and data processing. Planet. Sp. Sci. 51 (7–8), 505–515. Safaeinili, A., et al., 2007. Estimation of the total electron content of the martian ionosphere using radar sounder surface echoes. Geophys. Res. Lett. 34, L23204. http://dx.doi.org/10.1029/2007GL032154. Schunk, R.W., Nagy, A.F., 2009. Ionospheres: Physics, Plasma Physics, and Chemistry. Cambridge University Press, New York. Shinagawa, H., 1996. A two-dimensional model of the Venus ionosphere 2. Magnetized ionosphere. J. Geophys. Res. 101, 26921–26930. Shinagawa, H., Cravens, T.E., 1989. A one-dimensional multispecies magnetohydrodynamic model of dayside ionosphere of mars. J. Geophys. Res. 94, 6506–6516. http://dx.doi.org/10.1029/JA094iA06p06506. Shinagawa, H., Bougher, S.W., 1999. A two-dimensional MHD model of the solar wind interaction with Mars. Earth Planets Sp. 51, 55–62. Wieman, S.R., Didkovsky, L.V., Judge, D.L., 2014. Resolving differences in absolute irradiance measurements between the SOHO/CELIAS/SEM and the SDO/EVE. Sol. Phys. 289, 2907–2925. http://dx.doi.org/10.1007/s11207-014-0519-5. Withers, P., et al., 2012a. Observations of the nightside ionosphere of Mars by the Mars Express radio science experiment (MaRS). J. Geophys. Res. 117, A12307. http://dx.doi.org/10.1029/2012JA018185. Withers, P., et al., 2012b. A clear view of the multifaceted dayside ionosphere of mars. Geophys. Res. Lett. 39, L18202. http://dx.doi.org/10.1029/2012GL053193. Withers, P., Morgan, D., Gurnett, D.A., 2015. Variations in peak electron densities in the ionosphere of mars over a full solar cycle. Icarus 251, 5–11. http://dx.doi. org/10.1016/j.icarus.2014.08.008.
Contents lists available at ScienceDirect
Planetary and Space Science journal homepage: www.elsevier.com/locate/pss
Martian ionosphere observed by Mars Express. 1. Influence of the crustal magnetic fields E. Dubinin a,n, M. Fraenz a, D. Andrews b, D. Morgan c a
Max-Planck-Institute for Solar System Research, 37077 Goettingen, Germany Swedish Institute of Space Physics, Uppsala, Sweden c University of Iowa, Iowa city, Iowa, USA b
art ic l e i nf o
a b s t r a c t
Article history: Received 16 October 2015 Accepted 11 February 2016 Available online 21 February 2016
We present multi-instrument observations of the effects of the crustal magnetic field on the Martian ionosphere at different altitudes and solar zenith angles by Mars Express. Total electron content (TEC) at solar zenith angles 551 Z SZA Z 1051 over the ionosphere with crustal sources increases with the strength of the magnetic field. A similar trend is observed in a dependence of the local electron density in the upper ionosphere on the crustal magnetic field. On the nightside, at SZA Z 1101, the opposite trend of TEC increase with decrease in the magnetic field value is observed. A dependence on the magnetic field inclination also varies between the day and night sides. TEC decreases for vertical field inclination at 901 Z SZA Z 701 and increases at SZA Z 1101. This effect becomes stronger for larger magnetic field values. A different dependence of the local electron densities in the upper ionosphere at small and high SZA is observed too. An ionospheric exhaust for vertical field inclination in the regions with strong crustal sources is probably caused by escape to space along open field lines which arise due to reconnection that is confirmed by the case studies. The existence of such localized ionospheric depressions is also observed by the in-situ plasma observations. In contrast, on the nightside downward plasma transport and electron precipitation along the field lines produce patches of enhanced ionization. & 2016 Elsevier Ltd. All rights reserved.
Keywords: Mars Crustal magnetic field Ionosphere Mars Express
1. Introduction The first detection of the Martian ionosphere with a peak value of the electron number density of about 105 cm 3 at an altitude of about 125 km was made by Mariner-4 using the method of radio occultation (Kliore et al., 1965). Later, similar measurements on other space missions to Mars have extensively sampled the Martian ionosphere (Kliore, 1992; Hinson et al., 1999; Pätzold et al., 2005). In particular, a lower-altitude density peak or shoulder at h 110 km was also identified. The first in-situ ionospheric measurements were carried out by Viking-1 and 2 landers (Hanson et al., 1977) and provided us with the height versus density profiles of the main ionospheric species (O2þ , CO2þ and O þ ). Models of the ionosphere (see e.g. Shinagawa and Bougher, 1999; Krasnopolsky, 2002; Fox et al., 1993, 2004; Schunk and Nagy, 2009; Mendillo et al., 2011; Chaufray et al., 2014) have significantly increased our knowledge of the ionospheric structure and its dynamics. The bulk of the ionosphere at Mars is created by the ionization of the major atmospheric neutrals CO2. The main ionization n
Corresponding author. Tel.: þ 49 551 384979129. E-mail address: [email protected] (E. Dubinin).
http://dx.doi.org/10.1016/j.pss.2016.02.004 0032-0633/& 2016 Elsevier Ltd. All rights reserved.
process is photoionization by the solar EUV irradiance (10–90 nm). A lower peak at 90–100 km is produced by soft X-rays ð r10 nmÞ. However, due to the photochemistry reactions, the dominant ion species at hZ 200 km become molecular ðO2þ Þ and atomic ðO þ Þ oxygen ions. A balance between photoionization and recombination determines the peak electron density at altitudes of about 130 km. At altitudes above 180 km the ionosphere is no longer in photochemical equilibrium and diffusion and transport processes become important. Due to recombination of O2þ ions the ionospheric density rapidly decreases at the nightside. The recent observations by Mars Global Suveyor (MGS) and Mars Express (MEX) spacecraft have strongly increased the database of the ionospheric measurements providing us a lot of new information. Reviews of these observations are given in Haider et al. (2011), Orosei et al. (2015), and Withers et al. (2015)). In addition to the radio occcultation experiment, MEX carries the Mars Advanced Radar for Subsurface and Ionospheric Sounding (MARSIS). The MARSIS radar has two modes of operation: the Active Ionospheric Sounder (AIS) mode (Gurnett et al., 2005) and the SubSurface (SS) mode (Picardi et al., 2005). In the AIS mode the radar transmits stepped pulses from 0.1 to 5.4 MHz. The remote ionospheric measurements are based on the reflection of the
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
transmitted waves in the ionosphere where the wave frequency pffiffiffiffiffi meets the local plasma frequency (f p ne , where ne is the electron number density). Such a method allows to retrieve the altitude profile of the ionospheric electron number density down to the altitude of the main ionospheric peak ðh 130 kmÞ (Gurnett et al., 2005; Morgan et al., 2008). The sounding measurements have shown that the electron density in the main ionospheric layer at 140 km is in reasonable agreement with the Chapman model while at higher altitudes ( 200 km) a possible second layer probably associated with O þ ions appears. Besides that, when the transmitted frequency passes through the local electron frequency, strong electrostatic oscillations at fp are excited in the ionosphere and can be measured giving us an information about the local ionospheric density (Gurnett et al., 2005; Duru et al., 2008; Andrews et al., 2013). In the subsurface (SS) mode of the MARSIS operation, the radar transmits wideband (1 MHz) pulses at four centered frequencies (1.3, 3, 4 and 5 MHz). Traversing the ionosphere the waves are distorted by a frequency-dependent change of the phases due to dependence of the wave speed on the frequency. The distortions R1 depend on the total electron content TEC ¼ 0 nð zÞ dz which can be retrieved from the echo wave signal (Safaeinili et al., 2007; Mouginot et al., 2008). Safaeinili et al. (2003) have analyzed the TEC data obtained from June 2005 to September 2006. Assuming that the signal distortion is caused mostly by the ionospheric content below 200 km where the approach of the photochemical equilibrium is valid, and using the Chapman model (Chapman, 1931) Safaeinili et al. (2003) have derived a mean scale height of 11 .5 km and the mean value of the peak number density of 2:1 105 cm 3 at the subsolar point (solar zenith angle SZA ¼ 01). While the ionosphere at h r200 km is in photochemical equilibrium, deviations from the Chapman model become essential at larger altitudes. Duru et al. (2008) have shown that the ionospheric densities above 300 km varies with solar zenith angle not according to Chapman theory as cos ðSZAÞ1=2 implying a dominance of horizontal transport processes over photochemistry. Since the ionospheric thermal pressure above 300 km is insufficient to endure the dynamic pressure exerted by the solar wind, the ionosphere occurs magnetized and dynamics of the upper ionosphere is controlled the by the solar wind (Shinagawa and Cravens, 1989; Dubinin et al., 2008). The nightside ionosphere of Mars occurs highly variable. It is generally thought that plasma transport from the dayside and ionization of the atmosphere by precipitating electrons are the most important sources of the nightside ionosphere (Haider et al., 1992; Fox et al., 1993; Fillingim et al., 2010; Lillis et al., 2009, 2011). Analyzing the radio occultation measurements at the nightside of Mars made onboard MEX Withers et al. (2012a) have shown that transport from the dayside is a main source at SZA r 1151 but at higher SZA electron precipitation probably dominates. The existence of strong localized sources of crustal magnetic field at Mars (Acuna et al., 1999) introduces new important features. Analyzing the radio occultation measurements by MGS, Ness et al. (2000) have observed that scale heights of the electron number density were much larger above regions of locally vertical magnetic field. Ness et al. (2000) have assumed that reconnection between the crustal magnetic field and draping IMF may lead to penetration of solar wind electrons to low altitudes and heating of the ionosphere electrons. Increase in the electron temperature can decrease the efficiency of the dissociative recombination of the major ion species ðO2þ Þ and increase in the density scale height (Krymskii et al., 2003). Nielsen et al. (2007) have reported about cases of the peak density enhancements in spatially limited regions with strong crustal magnetic field. Neither solar events or precipitation of
63
energetic electrons were observed during these events. Nielsen et al. (2007) have suggested that plasma instabilities driven by solar wind interaction in minimagnetospheres of Mars might be responsible for electron heating and density enhancement. Safaeinili et al. (2007) have observed that at the nightside ionosphere of Mars, TEC is higher over the region with vertical crustal magnetic field although the enhancements in TEC were also measured in the areas where there was no evident connection with crustal sources. Analyzing MARSIS ionograms in AIS mode Němec et al. (2010) have observed that the occurrence rate of the nightside ionosphere is more than 4 times larger in regions with nearly vertical crustal magnetic field. The measurements of low energy ions carried by the Ion Mass Analyzer (IMA) onboard MEX have shown that the dayside ionosphere over the regions with strong crustal magnetization occurs more inflated (Dubinin et al., 2012). These results were confirmed by the measurements of the local plasma density retrieved from the frequencies of electron plasma oscillations (Andrews et al., 2013). Andrews et al. (2015) have provided an empirical model of effects of the crustal magnetic field on plasma density in the topside ionosphere. Cartacci et al. (2013) have analyzed variations of TEC in the nightside ionosphere based on MARSIS data from 2006 to 2011. To quantify the TEC variations they did not used average values as was done by Safaeinili et al. (2007) but 10th order polynomial fits. Cartacci et al. (2013) have also observed that the increase in TEC (positive δ (TEC)) was often related to the regions with vertical crustal fields. Fillingim et al. (2010) and Lillis et al. (2011) have modeled effects of precipitating electrons into the crustal field area on the nightside. These studies have found a large variability in peak ionization rate between different geographic regions. According to Fillingim et al. (2010) ionization patches produced by the precipitating electrons with densities up to 3 104 cm 3 at an altitude of about 140 km appear. In this paper we study properties of the Martian dayside and nightside ionosphere observed by MEX spacecraft by combining observations by the MARSIS radar and ASPERA-3 particle spectrometer. The measurements of the total electron content (TEC) provide us the information about the ionosphere near the main ionospheric peak. The in-situ measurements by MARSIS of the electron number density at altitudes from 300 km to 1400 km tell us about the characteristics of the topside ionosphere. These data are complemented by the measurements of low-energy ion and electron fluxes by ASPERA-3. We focus on effects of the planetary crustal magnetic field on the ionosphere of Mars.
2. Observations 2.1. Instrumentation We use the MARSIS-TEC data from June 2005 to August 2010 derived using the method by Mouginot et al. (2008). Note also that Cartacci et al. (2013) have applied another method (‘contrast method’) of compensating the distortions in the phase shift caused by the Martian ionosphere. This method allowed processing the MARSIS data in the nightside ionosphere with an accuracy of 10% . Local electron number densities can be derived by measuring the spacing between plasma frequency harmonics generated by distortion in the preamplifier due to the large amplitude of the emitted wave (Duru et al., 2008). Andrews et al. (2013) have developed a method for the automatic retrieval of local plasma density from a large data set of the MARSIS data in the AIS mode. In this paper, we use this database of local density derived from the observations between June 2005 and May 2014.
64
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
The MARSIS data are complemented by the ASPERA-3 measurements. The Ion Mass Analyzer (IMA) is a part of the ASPERA-3 instrument (Barabash et al., 2006). It determines the composition, energy and angular distribution of ions with energy r 30 keV. A new patch uploaded on May 1, 2007 expanded the energy range down to 10 eV. However, in the ionosphere, IMA is able to detect ions with lower energy as the spacecraft potential in this region usually varies in the range from 5 to 10 V. This shifts the ion energy spectrum to higher energy channels. At Ei 4 50 eV, IMA measures fluxes with a time resolution of 192 s and with the field of view of 901 3601. At lower energies ðEi r 50 eVÞ the measurements are carried out without elevation scanning that decreases field-of-view to 41 3601 but increases time resolution to 12 s. The electron sensor (ELS) measures electron fluxes in the energy range of 5 eV–20 keV with a field-of-view of 41 3601 and a time resolution of 4 s. 2.2. Presentation of the data Fig. 1a shows the values of TEC as a function of the solar zenith angle (SZA). The color bar presents the number of samples in each bin (1° bins in SZA). One can see that the sampling at low zenith angles is rather poor. Variations in TEC are caused by several factors. One of them is the varying distance between Sun and Mars that affects the solar UV flux which impacts Mars. Fig. 1b presents the same data set with color corresponding to the Sun-Mars distance. It is observed that TEC increases with decrease of the distance although some anomal features, for example, at SZA 60° and 100°, are observed too, which might be related to other factors controlling the ionospheric conditions. Fig. 1c compares the TEC measurements with solar irradiance. We used here the data of the Solar EUV Monitor (SEM) on the Solar Heliospheric Observatory (SOHO) which measures EUV and soft X-ray solar irradiance since 1996. SEM measures absolute EUV fluxes in the band 0.1–50 nm. Solar fluxes from the SOHO spacecraft were scaled in the intensity to the distance of Mars and rotated to Mars position. Note that the SEM data are the source of the S10.7 solar irradiance proxy which is widely used for analysis of TEC. A dependence of TEC on solar irradiance in Fig. 1c is evident. To investigate dependence of TEC on other factors we need to normalize TEC values to EUV flux and solar zenith angle. The dependence of TEC on solar zenith angle follows well the Chapman (Chapman, 1931) theory. Fig. 2 shows the mean values of TEC
R1 and its standard deviations. A Chapman curve TEC ¼ no 0 exp p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi h ho dh ¼ no H 2π e cos ðSZAÞ is shown 1=2 1 h Hho secðSZAÞe H by the green curve. Here h is the altitude, ho is the altitude of the peak density, H is the scale height, SZA is the solar zenith angle, and no is the peak electron density at SZA ¼ 0. A reasonable fitting between the data and the Chapman curve (no ¼ 1:7 105 cm 3 and H ¼11.5 km) calculated without taking into account of grazing incidence of solar radiation is observed for the dayside ionosphere ðSZA r951Þ except for the region around SZA 501 where a dip in the TEC occurs. Therefore we use the given Chapman function to normalize TEC values to SZA. Note that the value of H is close to the value obtained by Safaeinili et al. (2007) from a smaller dataset though the value of no is slightly lower. To normalize TEC values to EUV flux we study variations of TEC normalized to SZA on solar irradiance (Fig. 3). Here we used the SEM data from SOHO in two passbands (0.1–50 nm and 27– 34 nm) (see panels (a) and (b)) and the data from Solar EUV Experiment (SEE) of the Thermosphere, Ionosphere, Mesosphere, Energetics and Dynamics (TIMED) satellite in the range of 27–34 nm (panel (c)) and in the passband about 30.4 nm HeII
Fig. 2. Mean values of TEC and its standard deviations. Dashed green line shows a Chapman curve (no ¼ 1:7 105 cm 3 , H ¼ 11:5) fitted to the data. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
Fig. 1. TEC as a function of SZA. (a) Color shows a number of the measurements in each bin, (b) color shows the distance (in AU) between Sun and Mars, (c) color shows solar irradiance measured by the SEM instrument onboard SOHO and translated and rotated to the Mars position. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
65
Fig. 3. TEC values normalized to SZA and its standard deviations as a function of solar irradiance measured on the Earth orbit by different instruments on different satellites and translated to Mars: (a) SEM/SOHO data in the passband 0.1–50 nm, (b) SEM/SOHO data in the passband 27–34 nm, (c) SEE/TIMED data in the 30.4 HeII line. Red lines show power-law fitting to the data. Exponents of fitted curves are also given. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
Fig. 4. Comparison between TEC-Solar irradiance dependencies using the measurements of solar radiation by SEE/TIMED and SDO.
line (panel (d)). Mean values of TEC and its standard deviations are shown in black. Red lines are power-law fitting curves. Colored bins present the number of samples. Power exponents which characterize the slope of the curves occur different and generally differ from the Chapman predicted value k ¼0.5. The difference is partly related to the uncertainties in EUV measurements made by different instruments (see e.g. Peterson et al., 2012; Wieman et al., 2014) and different spectral efficiency in production of photoelectrons. Note that Lillis et al. (2010) have compared TEC on Mars with solar EUV flux in the 30.4 nm HeII line for the period between June 2005 and September 2006 and found that it varies as power law with exponent equal to 0.54. Only with launch of Solar Dynamic Observatory (SDO) high spectral resolution observations of solar irradiance in the passband 1–100 nm become available since April 2010. Fig. 4 compares TEC values as function of solar irradiance measured by TIMED (30.4 nm line) and SDO in the period between August 2010 and August 2012. The power-law exponent 0.52 for the relationship between TEC and solar flux by SDO is in approximate agreement with Chapman theory while for TIMED we observe a somewhat higher value of the exponent (0.62). In this
paper we use the EUV data only to normalize empirically TEC values and do not focus on analysis of exact relationship between TEC and solar fluxes keeping in mind that there are always uncertainties associated with a phase shift of EUV irradiance measured near Earth and at Mars (see also Girazian and Withers, 2015). 2.3. Effect of crustal field on TEC Fig. 5 shows TEC values normalized by EUV flux as a function of solar zenith angle (SZA). Colored bins in panel (a) depict the median values of the latitude at which the measurements were done. A remarkable asymmetry between the northern hemisphere and the southern hemisphere, where the most strong crustal sources are placed, is seen at SZA 601–1001. TEC at the dayside increases while moving from the northern to the southern latitudes. Colored bins in panel (b) show the median values of the crustal magnetic field at altitude of 150 km derived from the Cain's model (Cain et al., 2003). We observe that at SZA 601–1001 lower TEC values are over the regions with weak crustal field ðB r 10–20 nTÞ. On the nightside, the picture is more controversial.
66
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
Fig. 5. TEC values normalized to EUV fluxes as functions of SZA. (a) Color shows the mean planetocentric latitude of the observations in each bin, (b) color shows the value of the crustal magnetic field at altitude of 150 km derived from Cain's model. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
Over the regions with strong crustal field we observe higher as well as very low TEC values. Let us consider variations in TEC with crustal field in more detail. Fig. 6 shows the mean values of TEC on the dayside, at solar zenith angles SZA Z 601, where statistics of sampling is higher (see Fig. 1), as functions of the total value of the field (Bt), the radial (Br) and the horizontal (Bh) components, respectively. The standard deviations are shown only on Bt TEC plots to avoid overloading of Fig. 6. The trend of lower TEC in the ionosphere with low strength of the crustal field is observed. Such a trend is seen at B r 50 nT and not only for the total value of the magnetic field but also for the vertical and horizontal components that conceal a dependence on the magnetic field inclination. The two bottom rows show the dependences of TEC on the angle of the field inclination at the altitude of 150 km. The data are sorted by the value of the crustal field at 400 km – weak ð r 20 nTÞ and strong ð 4 20 nTÞ crustal fields, respectively. Note that the value 20 nT which separates here two different regimes of the crustal magnetization is rather arbitrary. It is observed that at SZA Z 751 a trend of a decrease in TEC at vertical orientation of the field (Mag. inclination 01 or 180°) appears. At smaller SZA we do not see a dependence of TEC on the field direction. At B Z 50 nT variability significantly increases exposing higher as well as lower TEC values although a trend of a slight increase in TEC with magnetic field value is still visible. On average, the increase in TEC can reach 20%. A large variability observed at high magnetic fields is partly due to a poorer sampling. Similar trends in TEC are observed on the nightside close to the terminator ðSZA r 1051Þ (Fig. 7) since altitudes near 150 km at which a typical ionospheric density peak is located remain sunlit at SZA r1071. At larger SZA we observe a different picture (Fig. 8). TEC increases in regions with small ðB r 50 nTÞ values of the crustal field without a visible relation to the values of the vertical (Br) component. The right panel on the middle shows a dependence of TEC on magnetic inclination angle without separation between regions with weak and strong crustal fields. One can see that TEC somewhat increases in the regions with a dominance of the vertical component. The bottom panels depict such a dependence in the ionosphere with weak (B r20 nT at 400 km) and strong
(B 4 20 nT at 400 km) field. We also increase the range of SZA (110°–140°) for better statistics. In the ionosphere with strong crustal field the dependence on the field line inclination becomes more pronounced. At vertical direction of the field lines, TEC increases up to 2 times. In the regions with weak crustal field dependence of the field orientation becomes weaker and almost vanishes. 2.4. Effect of crustal field on upper ionosphere Since upward diffusion and horizontal transport become more important with increase in altitude and these motions should be affected by the crustal sources we expect a more distinct relation between the ionosphere above 200 km and crustal fields. The upper panels in Fig. 9 show the local electron number density derived from analysis of waves excited by MARSIS pulses as a function of altitude above 300 km and crustal field (Bt ; Br ; Bh and magnetic inclination angle) derived from the MGS magnetic field map at 400 km (Connerney et al., 2001). The MARSIS data were processed for the period of 2005– 2014. Lower panels show the mean values of the electron density and the standard deviations measured on the dayside, at altitudes 350–450 km and 450–600 km, respectively. On the nightside MARSIS has operated mostly in SS mode and the local measurements are rather scarce. It is seen that the electron density generally increases with increase in the magnetic field strength at all altitudes sampled by MEX (see also Dubinin et al., 2012; Andrews et al., 2013, 2015). Such a trend is also observed in the density variations with variations in the vertical and horizontal components of the crustal field. In regions with strong magnetic field the density increases approximately by a factor 3. Such a behavior is similar to the relationship between TEC and crustal field. The dependence on the field line inclination at altitudes o 350 km contains an enhancement in the number density at angles near 90° that was also typical for TEC variations at large SZA. At higher altitudes the opposite trend develops. Higher values of the density are observed at the vertical orientation of the field lines. We also see that on the color maps of Fig. 9 the number density in some bins at high values of the vertical field component ðBr Þ is lower than in bins at lower field values, i.e. at very large magnetic fields the electron density again decreases. The latter might be
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
67
Fig. 6. Mean values of TEC at different SZA on the dayside ionosphere as functions of the total value of the crustal magnetic field (model), the vertical component Br, the horizontal component Bh, and inclination angle of the magnetic field.
68
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
Fig. 7. Similar to Fig. 6 plots for SZA ¼ 901–1051.
Fig. 8. Mean values of TEC at different SZA on the nightside ðSZA 41101Þ as a function of crustal magnetic field parameters at altitude of 150 km.
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
69
Fig. 9. Upper panels show maps of local electron number density at different altitudes as function of crustal field parameters measured at altitude of 400 km by MGS. Lower panels present mean values of the electron number density and its standard deviations as function of Bt, Br, Bh and the inclination angle at altitudes of 350–450 km and 450– 600 km, respectively. (For interpretation of the references to color in this figure, the reader is referred to the web version of this paper.)
related to more easy escape of the ionospheric plasma in regions where the crustal field can directly interact with the magnetosheath plasma opening access of solar wind plasma to the ionospere and escape of the ionospheric plasma into space. This is confirmed by the case studies. Fig. 10a shows an example of the measurements made on a single MEX orbit crossing the region of strong crustal field. The top panel presents the local electron density. Second and third panels depict the vertical component of the crustal field derived from the Cain's model and the energy–time spectrogram of electron fluxes measured by the ELS/ASPERA-3 instrument, respectively. Sharp drops in the local electron density are observed in the regions where the magnetosheath electrons intrude into the ionosphere marking an openness of the field lines. It is interesting to note that these density depressions occur rather stable in time. Fig. 11 shows two sets of observations on several subsequent orbits crossing two almost identical regions of strong crustal fields in the southern hemisphere (longitude 187° – left panels and 180° – right panels). It seems that some dips in the density survive from orbit to orbit over several days.
We do not know how deep in altitude these density depressions extend and whether they are visible in the total electron content. Since MARSIS operates in AIS and SS modes in turn we cannot compare TEC with local number densities at higher altitudes. However we can tentatively compare TEC data with the ELS measurements. Fig. 10b compares TEC variations with the electron fluxes and the radial component of the crustal magnetic field on the MEX orbit sampling the same region of the ionosphere as one in Fig. 10a. We observe similar features of the electron intrusion as in Fig. 10a, accompanied by small dips in TEC in the areas where there were depressions in the local electron density (Fig. 10a). However, it is difficult to confirm a potential connection between such depressions in TEC and the number density dips at altitudes above 300 km without a correct subtraction of the average TEC values. The local measurements of the density at lower ð r 300 kmÞ altitudes can clarify a connection between lower and upper ionosphere. It is worth noting that the existence of the ionospheric density dips is also confirmed by the ELS/ASPERA-3 measurements. Two narrow energy bands at 20 eV on the energy-time spectrogram of electron fluxes shown in Fig. 12 are associated with photoelectron
70
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
Fig. 10. (a) From top to bottom: the electron number density, the vertical component of the model crustal field, the spectrogram of electron fluxes. Dips in the density are observed in the regions where sheath electrons intrude into the ionosphere. (b) Comparison of TEC variations with the radial field component and electron fluxes on the MEX orbit sampling the same area as in Fig. 11a.
Fig. 11. Data obtained on the subsequent orbits sampling the same regions of the crustal magnetic field. Some of the density dips are repetitively observed on each orbit implying their rather stable subsistence. Electron density is in black. Components (total, radial and horizontal) of the model magnetic field are in color. (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
peaks due to absorption of the strong HeII solar line at 30.4 nm in the carbon dioxide dominated atmosphere (Frahm et al., 2006). Peaks are shifted to low energies on the spectrograms by a negative spacecraft potential. The crimson colored curve depicts the number of counts in the energy range 14–22 eV. We observe that depressions in the intensity of the ‘CO2 photoelectron lines’ correlate with the dips in the total electron density measured by MARSIS implying a possible ionospheric escape from such areas. Note that the data in Fig. 9 were summed over all SZA and therefore variations with SZA were missed. A map of the local number density at h r 350 km and 350 o h r450 km at different SZA as a function of Br and Bh measured by MGS at 400 km is shown in Fig. 13. An increase in ne with increase in Br is observed at all SZA although at Br 4100 200 nT the number densities are less than at Br r 100 nT. A similar trend at all SZA is observed for the dependence on the horizonal magnetic field Bh. If we look on the relationship between the local number density in the upper
ionosphere and the magnetic field inclination at different SZA (Fig. 13 lower panels) we see that at smaller SZA the number density is higher where the crustal field is more vertical. Such a dependence vanishes at higher SZA even with the opposite, though weak, trend of decreasing density in the regions with vertical field inclination, i.e. similar to what we observe for TEC variations at SZA Z 701.
3. Discussion and conclusions 3.1. Dayside ionosphere Systematic measurements of TEC on the dayside were made only at SZA Z 551. At smaller SZA the data are rather occasional. It is shown that TEC values normalized to solar radiation fluxes over the ionosphere with crustal sources increase with the strength of
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
the magnetic field. This trend is already observed at low values of the crustal field ( r 50 nT at altitude of 150 km) and remains in the regions with stronger magnetic fields although the variability in TEC significantly increases. Similar trends observed in relationships between TEC and crustal fields, and between local number densities in the upper
Fig. 12. Same as in Fig. 11a. Dips in the ionospheric number density are also displayed as depressions in the fluxes of CO2 photoelectrons (crimson colored curve on the bottom panel). (For interpretation of the references to color in this figure caption, the reader is referred to the web version of this paper.)
71
ionosphere ð Z300 kmÞ and crustal fields motivated us to evaluate the contribution to TEC from different altitude layers. Contribution from altitudes above 300 km can be evaluated using the empirical model for the upper ionosphere by Němec et al. (2011) and Andrews et al. (2015). Both models have used the same expoðh ho Þ
nential dependence ne ðh; SZAÞ ¼ neo ðSZAÞe HðSZAÞ , where neo is the number density at ho ¼325 km and the scale height H are parametrized values of SZA (Andrews et al., 2015). Taking, for example, SZA ¼ 601 yields neo ¼ 1044 (1892) cm 3 and H¼ 110.4 (176) km and contribution to TEC of 1:15 1014 ð3:2 1014 Þ m 2 for areas with weak and strong crustal fields, respectively. This is only 2% of the typical TEC values in these regions (see Fig. 6). Thus the observed TEC variations (up to 20%) related to the crustal field probably occur at altitudes below 325 km. MARSIS soundings have shown that the height profile of the ionosphere on the dayside often deviates from the Chapman profile at altitudes just few tens of km above the main ionospheric peak forming layers at altitudes from 180 km to 300 km (Kopf et al., 2008). The multifacetal shaping of the dayside ionosphere which is very different from the classical Chapman one was also emphasized by Withers et al. (2012b). Integrating the ne(h) profile, presented by Kopf et al. (2008, Figure 5), in the altitude range between 135 and 391 km yields to the column density 6:96 1015 m 2 . Yields in the altitude intervals of 180–391 km and 225–381 km are 2.2 10 15 m 2 and 0.74 10 15 m 2, i.e. 31.6% and 10.6%, respectively. Thus variations in TEC related to crustal fields on the dayside probably appear at h Z 225 km. At altitudes above 200 km and magnetic field B 30–50 nT, the frequency of ion-neutral collisions νin is less than the ion gyrofrequency ωBi, and the frequency of electron-neutral collisions is much less than the electron gyrofrequency ΩBe.
Fig. 13. Upper panels (a, b): maps of the local electron density at h r 350 km (a) and 350 o h r 450 km (b) at different SZA as function of the Br and Bh field components of the crustal field at 400 km. Lower panels (c, d) show maps of the local electron number density as a function of the magnetic field inclination. Curves depict dependences of ne on the field inclination at SZA ¼ 201 451 and SZA ¼ 751 851 at altitudes h r 350 km and 350 o h r 450 km, respectively.
72
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
Correspondingly, the ambipolar diffusion of plasma across the A , where DA is magnetic field is strongly reduced: D ? 1 þ ωBi ΩDBe =νin νen the coefficient of the ambipolar diffusion without magnetic field or along the field lines. Below 200 km, ions become demagnetized ðνin 4 ωBi Þ while electrons remain magnetized down to the altitude of the main ionospheric peak. With increase in the value of the magnetic field the ion demagnetization occurs at lower altitudes. Ion demagnetization can enlarge the diffusion even across the magnetic field. Moving freely upward, along the closed crustal magnetic field lines, electrons produce an excess of the negative charge. The resulting ambipolar electric field will push nonmagnetized ions upward across the magnetic field. The coefficient of ambipolar diffusion in this case will be determined by ion dynamics D ? 1 þ ωDBiA =νin DA . Despite of this local altitude effect, a horizontal crustal magnetic field generally, strongly reduces the mobility of ions and electrons and therefore one could expect an increase in the electron content with increase in the magnetic field strength. If the field is stronger, then the locally produced plasma and, in particularly, O þ ions originating at higher altitudes are less exposed to vertical and horizontal motions and their column density increases. Note, that even in the unmagnetized Martian ionosphere inclusion of the ion dynamics essentially modifies the ion height profiles above 170–180 km (Chaufray et al., 2014). A dependence on the field inclination should also appear since the ionospheric plasma can more easily rise up along the field lines with a subsequent horizontal transport driven by solar wind. This might be a cause of decrease in TEC for vertical field inclination at SZA Z 751. On the other hand, downward ion fluxes pushed by solar wind pressure could produce the opposite trend of TEC increase at the regions of vertical field inclination. Such a dependence could be expected rather at smaller SZA where the effect of solar wind is stronger. Unfortunately, we cannot examine this effect on TEC at low solar zenith angles because of a poor sampling (see Fig. 1a). Fig. 14 shows a map of the radial velocity of cold oxygen ions ðEi r50 eVÞ measured by IMA/ASPERA-3 as a function of SZA and altitude. We used the whole dataset from June 2007 to June 2015. Since the ion measurements of the cold component are carried out without electrostatic scanning in the elevation angle we select only the observations made for the intervals when the normal vector to Mars was in the plane ( 710°) of the IMA field of view. Since the measurements of the low-energy ions are also affected by the spacecraft potential we have chosen the average value of the potential value to be 8 V that is in general agreement with the ELS measurements of the peak energy of the ‘CO2 photoelectrons’. We observe that at lower SZA downward plasma motions
Fig. 14. Map of the planetocentric radial velocity of low-energy oxygen ions measured by IMA/ASPERA-3. A reversal of the flow direction is observed.
which could increase TEC, dominate while at higher SZA the flow occurs outward from the planet. Note also that a divergence of the ion fluxes can give rise to sharp gradients in the ionospheric number densities (‘ionopause’ signature (Duru et al., 2009)). 3.2. Nightside ionosphere On the nightside ðSZA Z 1101Þ we observe an opposite trend. At Bt r 50 nT at 150 km TEC increases with decrease in the magnetic field value. In this range TEC also almost does not vary with the variations in the Br-component. Such a behavior at small crustal fields can be easily interpreted assuming that the downward plasma transport is the main supplier of the nightside ionosphere. In the regions with weak crustal fields the downward transport is more efficient as compared to areas with strong crustal sources where transport and electron precipitation operate mainly along the field lines. This explains why increase in TEC for vertical field inclination becomes more profound with increase in the field strength (see Fig. 8). This trend was also noticed by Cartacci et al. (2013). Fig. 15 shows examples of observations made by MEX on four orbits on the nightside. Two top (bottom) panels show examples of the measurements made in the ionosphere with strong (weak) crustal field. Upper panels depict TEC values. Second and third panels present three components of the model crustal field at 150 km and angle of the magnetic field inclination, respectively. Lower panels show energy–time spectrograms of electron fluxes. Spikes in TEC correspond to the regions with a strong vertical component Br (blue lines) and are often accompanied by spikes of the magnetosheath/plasma sheet electrons penetrating into the ionosphere (top panels). It is reasonable to assume that these crustal field lines reconnect with the draped IMF and become open. Variability in TEC in such regions might be partly caused by sporadic processes of field reconnection. In the regions with weak crustal field (bottom panels), spikes in TEC are still observed. Although the value of the crustal field in these regions is small the appearance of spikes generally prefers the vertical orientation of the magnetic field. In this case, the assumption of open field lines is questionable and another mechanism of plasma transport seems to be operating. The observations by ASPERA-3 confirm the important role of plasma transport on the nightside. Because of the lack of the measurements with a proper attitude of the IMA sensor at low altitudes on the nightside we use another method. To overcome the problem of narrow field of view for detection of low-energy ions we use the method suggested by Nilsson et al. (2012) and Fränz et al. (2015) to construct the mean 3D-distribution in a given spatial bin around Mars by collecting many different 2Dobservations over long time in the Martian ionosphere. A detailed description of the derivation of the mean 3D – distribution functions and their moments is given in Fränz et al. (2015). Fig. 16a shows the dependence of the vertical velocity of O þ ions measured by IMA/ASPERA-3 instrument as a function of the altitude and the magnetic inclination angle. Although the uncertainty in the velocity values related to the effect of the spacecraft potential remains, it is seen that downward motions of low-energy plasma transported from the dayside prevail in the regions with vertical field orientation. An increase in the partial density of oxygen ions (still reduced as compared to the MARSIS observations due to a missing of the cold core population) at the vertical field orientation is also observed (Fig. 16b). Precipitation of the magnetosheath/plasma sheet electrons along the field lines is another important source supplying the nightside Martian ionosphere. Fig. 17a shows that TEC on the nightside ðSZA ¼ 1101–1401Þ increases with increase in the energy flux of electrons ðEe ¼ 30–500 eVÞ measured locally by ELS/
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
73
Fig. 15. Examples of MARSIS and ASPERA-3 observations made at the nightside. From top to bottom are TEC, three components of the crustal magnetic field (model), the inclination of the field line and energy-time spectrogram of electron fluxes. Upper (lower) panels show the TEC measurements made over the region with strong (weak) crustal field.
Fig. 16. (a) Map of the vertical velocity of oxygen ions at the nightside ðSZA ¼ 1101–1401Þ as a function of the angle of the magnetic field inclination and altitude. (b) Map of the ion density.
74
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
Fig. 17. (a) TEC on the nightside as a function of the energy flux of the electrons ðEe ¼ 30 500 eVÞ measured by ELS/ASPERA at the point of the radar operation. (b) TEC as a function of the electron fluxes on the dayside.
ASPERA-3. Note that on the dayside, at high SZA we often observe the opposite trend (Fig. 17b) due to losses along the field lines. Because of the lack of magnetic field observations at low altitudes we did not evaluate the role of the magnetic field of solar wind origin which permeates the ionosphere of Mars. This field is mainly horizontal and transported by the convection of the ionospheric plasma. It is reasonable to assume that in the regions with weak crustal sources the magnetic field is governed by the draped horizontal IMF. In contrast, the role of crustal field increases with their strength. Presently, there are no reliable measurements of the ionospheric motions and adequate models for the plasma flow in the magnetized ionosphere of Mars. A twodimensional model for the magnetized ionosphere of Venus by Shinagawa (1996) predicts a planetward motion (down to altitude of 180 km) at low SZA ðSZA r 501–601Þ and upward convective flows at higher SZA. Correspondingly, if this model is applicable to Mars (see Fig. 14) one could expect an increase (decrease) in TEC at low (high) SZA, respectively, in the regions with weak crustal sources where the ionospheric flows driven by solar wind dominate. In conclusion, we note that the coordinated measurements by MARSIS and ASPERA-3 and the measurements by the plasma instruments LPW, STATIC, SWEA and NGIMS, onboard MAVEN will provide a powerful tool for further studies of the Martian ionosphere and its response to different factors. The in-situ observations at altitudes not surveyed by MEX will allow us to identify the links between different ionospheric regions and their role in the ionospheric dynamics.
Acknowledgments Authors (ED, MF) wish to acknowledge support from DFG for supporting this work by Grant WO 910/1-1 and DLR by grants 50QM99035, FKZ 50 QM 0801 and 50QM1302. DJA was supported by Grants from Swedish National Space Board (DNR 162/14) and the Swedish Research Council (DNR 621-2014-5526).
References Acuna, M.H., et al., 1999. Global distribution of crustal magnetization discovered by the Mars Global Surveyor MAG/ER experiment. Science 284 (5415), 790–793. Andrews, D.J., et al., 2013. Determination of local plasma densities with the MARSIS radar: asymmetries in the high-altitude Martian ionosphere. J. Geophys. Res. 118, 6188–6196. http://dx.doi.org/10.1002/jgra.50593. Andrews, D.J., et al., 2015. Control of the topside Martian ionosphere by crustal magnetic fields. J. Geophys. Res. 120, 3042–3058. http://dx.doi.org/10.1002/ 2014JA020703. Barabash, S., et al., 2006. The analyzer of space plasmas and energetic atoms (ASPERA-3) for the Mars Express mission. Space Sci. Rev. 126, 113–164. http: //dx.doi.org/10.1007/s11214-006-9124-8.
Cain, J.C., Ferguson, B.B., Mozzoni, D., 2003. An n¼ 90 internal potential function of the Martian crustal magnetic field. J. Geophys. Res. 108, 5008. http://dx.doi.org/ 10.1029/2000JE001487. Cartacci, M., et al., 2013. Mars ionosphere total electron content analysis from MARSIS subsurface data. Icarus 223, 423–437. Chapman, S., 1931. Absorption and dissociative or ionizing effect of monochromatic radiation in an atmosphere on a rotating earth. Proc. Phys. Soc. (London) 43, 1047–1055. Chaufray, J.-Y., et al., 2014. Three-dimensional Martian ionosphere model: II. Effect of transport processes due to pressure gradients. J. Geophys. Res. 119, 1614–1636. http://dx.doi.org/10.1002/2013JE004551. Connerney, J.E.P., et al., 2001. The global magnetic field of Mars and implications for crustal evolution. Geophys. Res. Lett. 28, 4015–4018. http://dx.doi.org/10.1029/ 2001GL013619. Dubinin, E., et al., 2008. Plasma environment of Mars as observed by simultaneous MEX-ASPERA-3 and MEX-MARSIS observations. J. Geophys. Res. 113, A10217. http://dx.doi.org/10.1029/2008JA013355. Dubinin, E., et al., 2012. Upper atmosphere of Mars is not axially symmetrical. Earth Planets Sp. 64, 113–120. http://dx.doi.org/10.5047/eps.2011.05.022. Duru, F., et al., 2008. Electron densities in the upper ionosphere of Mars from the excitation of electron plasma oscillations. J. Geophys. Res. 113, A07302. http: //dx.doi.org/10.1029/2008JA013073. Duru, F., et al., 2009. Steep, transient density gradients in the marian ionosphere similar to the ionopause at Venus. J. Geophys. Res. 114, A12310. http://dx.doi. org/10.1029/2009JA014711. Fillingim, M.O., et al., 2010. Localized ionization patches in the nighttime ionosphere of Mars and their electrodynamic consequences. Icarus 206, 112–119. http://dx.doi.org/10.1016/j.icarus.2009.03.005. Fox, J.L., 2004. Advances in the aeronomy of Venus and Mars. Adv. Sp. Res. 33, 132–139. Fox, J.L., Brannon, J.F., Porter, H.S., 1993. Upper limits to the nightside ionosphere of Mars. Geophys. Res. Lett. 20, 1339–1342. http://dx.doi.org/10.1029/93GL01349. Frahm, R., Winningham, J.D., Sharber, J.R., et al., 2006. Carbon dioxide photoelectron energy peaks at Mars. Icarus 182, 371. Fränz, M., et al., 2015. Cold ion escape from the Martian ionosphere. Planet. Space Sci. 119, 92–102. http://dx.doi.org/10.1016/j.pss.2015.07.012. Girazian, Z., Withers, P., 2015. An empirical model of the extreme ultraviolet solar spectrum as a function of F 10:7 . J. Geophys. Res. 120, 6779–6794. http://dx.doi. org/10.1002/2015JA021436. Gurnett, D.A., et al., 2005. Radar soundings of the ionosphere of Mars. Science 310, 1929–1933. http://dx.doi.org/10.1126/science.1121868. Haider, S.A., et al., 1992. Calculated ionization rates, ion densities, and airglow emission rates due to precipitating electrons in the nightside ionosphere of Mars. J. Geophys. Res. 97, 10,637–10,641. http://dx.doi.org/10.1029/92JA00317. Haider, S.A., Mahajan, K.K., Kallio, E., 2011. Mars ionosphere: a review of experimental results and modeling studies. Rev. Geophys. 49, RG4001. http://dx.doi. org/10.1029/2011RG000357. Hanson, W.B., Satani, S., Zuccaro, D.R., 1977. The Martian ionosphere as observed by the Viking retarding potential analyzers. J. Geophys. Res. 82, 4351–4363. http: //dx.doi.org/10.1029/JS082i028p04351. Hinson, D., et al., 1999. Initial results from radio occultation measurements with Mars Global Surveyor. J. Geophys. Res. 104, 26997–27012. Kliore, A., et al., 1965. Occultation experiment: results of the first direct measurement of Mars'atmosphere and ionosphere. Science 149, 1243–1248. Kliore, A., 1992. Radio occultation observations of the ionospheres of Mars and Venus, in Venus and Mars: Atmospheres, Ionospheres and Solar Wind Interactions. In: Luhmann, J., Tatrallyay, M., Pepin, R. (Eds.), Geophysical Monograph Series, 66. AGU, Washington, DC, pp. 265–276. http://dx.doi.org/10.102 9/GM066p0265. Kopf, A.J., et al., 2008. Transient layers in the topside ionosphere of Mars. Geophys. Res. Lett. 35, L17102. http://dx.doi.org/10.1029/2008GL034948. Krasnopolsky, V.A., 2002. Mars' upper atmosphere and ionosphere at low, medium and high solar activities: implication for evolution of water. J. Geophys. Res. 107 (E12), 5128. http://dx.doi.org/10.1029/2001JE001809. Krymskii, A., et al., 2003. Effect of crustal magnetic fields on the near terminator ionosphere at Mars: comparison of in situ magnetic field measurements with
E. Dubinin et al. / Planetary and Space Science 124 (2016) 62–75
the data of radio science experiments on board Mars Global Surveyor. J. Geophys. Res. 108, http://dx.doi.org/10.1029/2002JA009662. Lillis, R.J., et al., 2009. The nightside ionosphere of mars: modeling the effects of crustal magnetic fields and electron pitch angle distributions on electron impact ionization. J. Geophys. Res. 114, E11009. http://dx.doi.org/10.1029/ 2009JE003379. Lillis, R.J., et al., 2010. Total electron content in the Mars ionosphere: temporal studies and dependence on solar EUV flux. J. Geophys. Res. 115, A11314. http: //dx.doi.org/10.1029/2010JA015698. Lillis, R.J., Fillingim, M.O., Brain, D.A., 2011. Three-dimensional structure of the Martian nightside ionosphere: predicted rates of impact ionization from Mars Global Surveyor magnetometer and electron reflectometer measurements of precipitating electrons. J. Geophys. Res. 116, A12317. http://dx.doi.org/10.1029/ 2011JA016982. Mendillo, M., et al., 2011. Modelling Mars' ionosphere with constrains from sameday observations by Mars Global Surveyor and Mars Express. J. Geophys. Res. 116, A11303. http://dx.doi.org/10.1029/2011JA016865. Morgan, D.D., et al., 2008. Variation of the martian ionospheric electron density from Mars Express radar soundings. J. Geophys. Res. 113, A09303. http://dx.doi. org/10.1029/2008JA013313. Mouginot, J., et al., 2008. Correction of the ionospheric distortion on the MARSIS surface sounding echoes. Planet. Sp. Sci. 56, 917–926. Němec, F., et al., 2010. Night side ionosphere of mars: radar soundings by the Mars Express spacecraft. J. Geophys. Res. 115, E12009. http://dx.doi.org/10.1029/ 2010JE003663. Němec, F., et al., 2011. Dayside ionosphere of mars: empirical model based on data from the MARSIS instrument. J. Geophys. Res. 116, E07003. http://dx.doi.org/ 10.1029/2010JE003789. Ness, N.F., et al., 2000. Effects of magnetic anomalies discovered at Mars on the structure of the Martian ionosphere and solar wind interaction as follows from radio occultation experiments. J. Geophys. Res. 105, 15991–16004. http://dx. doi.org/10.1029/1999JA000212. Nielsen, E., et al., 2007. Vertical sheets of dense plasma in the topside Martian ionosphere. J. Geophys. Res. 112, E02003. http://dx.doi.org/10.1029/ 2006JE002723. Nilsson, H., et al., 2012. Ion distributions in the vicinity of Mars: signatures of heating and acceleration processes. Earth Planets Sp. 64, 135–148. http://dx.doi. org/10.5047/eps.2011.04.011.
75
Orosei, R., et al., 2015. Mars advanced radar for subsurface and ionospheric sounding (MARSIS) after nine years of operation: a summary. Planet. Sp. Sci. 112, 98–114. http://dx.doi.org/10.1016/j.pss.2014.07.010. Pätzold, M., et al., 2005. A sporadic third layer in the ionosphere of mars. Science 5749, 837–839. Peterson, W.K., et al., 2012. Solar EUV and XUV energy input to thermosphere on solar rotation time scales derived from photoelectron observations. J. Geophys. Res. 117, A05320. http://dx.doi.org/10.1029/2011JA017382. Picardi, G., et al., 2005. Radar soundings of the subsurface of mars. Science 310, 1925–1928. Safaeinili, A., et al., 2003. Impact of mars ionosphere on orbital radar sounder operation and data processing. Planet. Sp. Sci. 51 (7–8), 505–515. Safaeinili, A., et al., 2007. Estimation of the total electron content of the martian ionosphere using radar sounder surface echoes. Geophys. Res. Lett. 34, L23204. http://dx.doi.org/10.1029/2007GL032154. Schunk, R.W., Nagy, A.F., 2009. Ionospheres: Physics, Plasma Physics, and Chemistry. Cambridge University Press, New York. Shinagawa, H., 1996. A two-dimensional model of the Venus ionosphere 2. Magnetized ionosphere. J. Geophys. Res. 101, 26921–26930. Shinagawa, H., Cravens, T.E., 1989. A one-dimensional multispecies magnetohydrodynamic model of dayside ionosphere of mars. J. Geophys. Res. 94, 6506–6516. http://dx.doi.org/10.1029/JA094iA06p06506. Shinagawa, H., Bougher, S.W., 1999. A two-dimensional MHD model of the solar wind interaction with Mars. Earth Planets Sp. 51, 55–62. Wieman, S.R., Didkovsky, L.V., Judge, D.L., 2014. Resolving differences in absolute irradiance measurements between the SOHO/CELIAS/SEM and the SDO/EVE. Sol. Phys. 289, 2907–2925. http://dx.doi.org/10.1007/s11207-014-0519-5. Withers, P., et al., 2012a. Observations of the nightside ionosphere of Mars by the Mars Express radio science experiment (MaRS). J. Geophys. Res. 117, A12307. http://dx.doi.org/10.1029/2012JA018185. Withers, P., et al., 2012b. A clear view of the multifaceted dayside ionosphere of mars. Geophys. Res. Lett. 39, L18202. http://dx.doi.org/10.1029/2012GL053193. Withers, P., Morgan, D., Gurnett, D.A., 2015. Variations in peak electron densities in the ionosphere of mars over a full solar cycle. Icarus 251, 5–11. http://dx.doi. org/10.1016/j.icarus.2014.08.008.