Electron Dynamics near Diamagnetic Regions of Comet 67P/Churyumov-Gerasimenko
H. Madanian, J.L. Burch, A.I. Eriksson, T.E. Cravens, M. Galand, E. Vigren, R. Goldstein, Z. Nemeth, P. Mokashi, I. Richter, M. Rubin
EElectron Dynamics near Diamagnetic Regions of Comet67P/Churyumov-Gerasimenko
H. Madanian a, ∗ , J. L. Burch a , A. I. Eriksson b , T. E. Cravens c , M. Galand d ,E. Vigren b , R. Goldstein a , Z. Nemeth e , P. Mokashi a , I. Richter f , M. Rubin g a Southwest Research Institute, 6220 Culebra Road, San Antonio, TX 78238, United Sates b Swedish Institute of Space Physics, Uppsala, Sweden c Department of Physics and Astronomy, University of Kansas, Lawrence, KS, USA d Department of Physics, Imperial College London, Prince Consort Road, London, UK e Wigner Research Centre for Physics, Budapest, Hungary f Institut fr Geophysik und extraterrestrische Physik, TU Braunschweig, Braunschweig,Germany g Physikalisches Institut, University of Bern, Bern, Switzerland
Abstract
The Rosetta spacecraft detected transient and sporadic diamagnetic regionsaround comet 67P/Churyumov-Gerasimenko. In this paper we present a sta-tistical analysis of bulk and suprathermal electron dynamics, as well as a casestudy of suprathermal electron pitch angle distributions (PADs) near a dia-magnetic region. Bulk electron densities are correlated with the local neutraldensity and we find a distinct enhancement in electron densities measuredover the southern latitudes of the comet. Flux of suprathermal electrons withenergies between tens of eV to a couple of hundred eV decreases each time thespacecraft enters a diamagnetic region. We propose a mechanism in whichthis reduction can be explained by solar wind electrons that are tied to themagnetic field and after having been transported adiabatically in a decayingmagnetic field environment, have limited access to the diamagnetic regions.Our analysis shows that suprathermal electron PADs evolve from an almostisotropic outside the diamagnetic cavity to a field-aligned distribution nearthe boundary. Electron transport becomes chaotic and non-adiabatic whenelectron gyroradius becomes comparable to the size of the magnetic field linecurvature, which determines the upper energy limit of the flux variation.This study is based on Rosetta observations at around 200 km cometocen- ∗ [email protected] Preprint submitted to Planetary and Space Science July 14, 2020 a r X i v : . [ phy s i c s . s p ace - ph ] J u l ric distance when the comet was at 1.24 AU from the Sun and during thesouthern summer cometary season. Keywords:
Comet 67P/Churyumov-Gerasimenko, Rosetta, Plasmadynamics, Diamagnetic cavity
1. Introduction
A cometary atmosphere is formed through heating and sublimation of iceand other volatiles at the comet surface. This mixture of neutral particlesexpands radially outward and is exposed to solar photons, solar wind chargedparticles, and the interplanetary magnetic field (IMF) (Glassmeier, 2017;Gombosi, 2015; Cravens and Gombosi, 2004). As comets approach perihelion,sublimation and outgassing rates increase and plasma boundaries can beformed in the thicker coma against the impinging solar wind (Mandt et al.,2016). A bow shock is the outer most boundary that may be formed aroundcomets where, due to assimilation of cometary ions into the solar wind, thesupersonic flow slows down to subsonic speeds. Within the bow shock thesolar wind plasma becomes heated. Upstream of comet Halley ' s bow shockat about 10 km from the nucleus, field-aligned backstreaming electrons wereobserved which were reflected from the enhanced magnetic field regions atthe shock (Larson et al., 1992). In addition, some perpendicularly heatedelectrons that escaped the cometary magnetosphere and traveled upstreamalong the magnetic field were present in the distribution at around 90 ◦ pitchangles. Fuselier et al. (1986) used data from the International CometaryExplorer (ICE) spacecraft at comet Giacobini-Zinner and showed similardistributions and increased electron heat flux backstreaming from the regionof enhanced magnetic field near the comet.At closer distances to the nucleus, a boundary that is particularly im-portant for nonmagnetized objects is the diamagnetic cavity boundary. Themagnetometer on the Giotto spacecraft detected the diamagnetic boundaryaround comet Halley at a cometocentric distance of 4500 km during the flybyinbound and the spacecraft exited the cavity at about 4150 km outbound(Neubauer et al., 1986). This global and relatively symmetric diamagneticcavity is formed when the outward ion-neutral drag force in the cometaryatmosphere balances the magnetic pressure gradient in the pile up region(Cravens, 1986; Ip and Axford, 1987). The size of the diamagnetic cavityat comet Halley was much larger than the solar wind proton gyroradius and2agnetohydrodynamic models sufficiently described the stand-off distanceof the diamagnetic boundary and the magnetic field profile across it (Lind-gren et al., 1997; Rubin et al., 2014). The AMPTE (Active MagnetosphericParticle Tracer Explorers) artificial comet experiment in 1984 created a tem-porary diamagnetic cavity ( ∼
60 s long and 70 km in radius) by releasing 2kg of Barium vapor in the solar wind (Bingham et al., 1991; Haerendel et al.,1986; Gurnett et al., 1986). The cavity boundary in this case was formedby electron currents. As Haerendel et al. (1986) describe, photoelectronsof the expanding barium gas are coupled to ions via a polarization electricfield, which further accelerates the ions radially outward. The electron gasinitially reaches a pressure balance with the solar wind magnetic field whileions continue to expand, resulting in an inward polarization electric field.Under these conditions, electrons form a current layer as they undergo E × B drift, leading to a shielding diamagnetic boundary.At comet 67P/Churyumov-Gerasimenko (or 67P for short), similar dia-magnetic regions have been observed, though the formation mechanism forthese events is not yet fully understood. The magnetometer system onboardthe Rosetta spacecraft detected plasma regions with near zero magnetic fieldand relatively small fluctuations (Goetz et al., 2016b). These regions wereobserved within a few hundred kilometers from the comet and appeared tobe highly sporadic and transient. Spacecraft dwell time at each event variedfrom seconds to more than 40 minutes, indicating the very dynamic and vari-able size of these structures (Goetz et al., 2016a; Timar et al., 2017). Aroundthe diamagnetic regions the bulk electron density is closely related to the lo-cal neutral density (Eriksson et al., 2017; Henri et al., 2017; Hajra et al.,2018). Henri et al. (2017) showed that a relation can be established betweenthe electron exobase and the observed diamagnetic boundary distances. Onthe other hand, suprathermal electrons show a peculiar signature in which,at each crossing into the field-free regions, flux of electrons with energiesfrom tens of eV to several hundreds of eV decreases (Madanian et al., 2016a;Nemeth et al., 2016).Rosetta observations near perihelion mostly represent a combination ofshocked, highly perturbed, and heated solar wind plasma, plus electrons andions of cometary origin. Given that comet 67P has no intrinsic magneticfield (Auster et al., 2015), and that solar wind ions have been obscured farupstream (Nilsson et al., 2017), solar wind electrons play a critical role incarrying the IMF through the coma (Plaschke et al., 2018). Furthermore, thespacecraft distance to the comet at this time is only a few hundred kilometers3hich is comparable to or smaller than ion gyroradii. Therefore, studyingthe small scale electron dynamics is crucial in understanding the nature ofthese events. Explicitly, electrons around the comet can originate from threesources: solar wind electrons, photoelectrons, and secondary electrons fromelectron-impact ionization (Galand et al., 2016; Madanian et al., 2016b; Vi-gren et al., 2016; Heritier et al., 2018). Models of electron production aroundcomet 67P at perihelion have shown that without acceleration processes,photoionization is the main source of electron production up to about 70 eV,while at higher energies solar wind electrons become dominant (Madanianet al., 2016a). Unperturbed solar wind suprathermal electrons exhibit dis-tinct non-Maxwellian features; an isotropic component known as halo, and afield-aligned strahl beam propagating usually in the anti-sunward direction(Feldman et al., 1975), though in the turbulent plasma environment of theinner coma and near diamagnetic cavities, these distributions will likely bemodified.In this paper we investigate the electron dynamics near diamagnetic re-gions of comet 67P. We show how bulk electron densities change across dia-magnetic boundaries. We analyze the energy extent of suprathermal electronflux difference between inside and outside the cavities. We provide a detailedcase study on how suprathermal electron pitch angle distributions (PADs)evolve, which has implications for the energy range of flux differences andthe size of the cavities. In section 2, we describe the instruments and dataprocessing method. Our observations are presented in section 3. We discussthe results and review the interpretations in section 4 and finally, provideour conclusions in section 5.
2. Instrumentation and Data Processing Method
We use data from the Magnetometer (MAG) (Glassmeier et al., 2007),the Ion and Electron Sensor (IES) (Burch et al., 2007), the Langmuir Probe(LAP) (Eriksson et al., 2007), and Rosetta Orbiter Spectrometer for Ionand Neutral Analysis / Cometary Pressure Sensor (ROSINA/COPS) (Bal-siger et al., 2007) in our analysis. The first three instruments are part ofthe Rosetta Plasma Consortium (Carr et al., 2007). In the following, wedescribe the IES instrument in detail and provide a brief description of theother instruments. For more details on each instrument, readers are referredto the corresponding instrument papers and references therein. Three elec-tron sensors onboard the Rosetta spacecraft with different detection methods4ave enabled us to study the electron dynamics at different energy rangesthrough the perspective of each instrument. Measurements of LAP and Mu-tual Impedance Probe (MIP) (Trotignon et al., 2007) instruments describethe bulk electron population, while IES measures electrons across a widerange of energies. There is no specific set of criteria to categorize the electronpopulations to our knowledge, and different authors have chosen different en-ergy ranges to label cold, thermal, suprathermal, and in some cases warmelectron populations. In this paper, the term suprathermal electron refers toIES electron measurements with energies above ∼
10 eV, and electrons belowthis energy constitute the bulk population. We use LAP data to estimatebulk electron densities. For electron directional variability analysis, we areonly interested in the suprathermal electrons that have energies exceeding100 eV.
The IES instrument on Rosetta is capable of measuring near full 3-Ddistribution of charged particles (Burch et al., 2007). The IES consists oftwo stacked toroidal electrostatic analyzers that measure electrons and ionswith energies between 4.3 eV and 18 keV. The IES energy resolution is 8percent at each energy bin and the instrument has a 360 ◦ azimuthal by 90 ◦ polar field of view, providing 2 . π solid angle coverage. The electron sensor,which is emphasized in this paper, has a 22 . ◦ azimuthal resolution providedby 16 anodes. It was initially designed to scan the polar coordinate through18 elevation steps with a 5 ◦ resolution. Later on, due to engineering reasonsthe in-flight software was modified to scan 16 elevation steps with 6 ◦ betweeneach step. Note the different angle nomenclature in Burch et al. (2007).The IES instrument was mounted on the corner of the spacecraft provid-ing a perfect pointing during most of the mission for probing the solar wind.For the period near perihelion, IES measurements in every two adjacentenergy channel pairs, elevation step pairs, and azimuthal anode pairs wereaveraged onboard before transmission to fit the available telemetry rate. Inour analysis, we used all individual sectors in the IES field of view (FOV) byassuming that the paired sectors share the averaged value evenly. We shouldnote that soon after the beginning of the mission, two IES anodes (anodes11 and 12) became malfunctional and did not return reliable data while theirneighboring anodes performed nominally. Furthermore, in early April 2015,the instrument showed reduced count intensities in half of the anodes (an-odes 8-15). These anodes shared the same octal amplifier. Further analysis5f data showed that this decrease in amplification efficiency affects primarilylow energy bins that also experienced saturation at high count rates. Highenergy bins ( ∼
100 eV and above) are not affected. We addressed these issuesin our calibration analysis and considered the possible induced uncertaintiesbefore drawing conclusions.To convert IES raw counts to a physical parameter such as differen-tial electron flux, we rely on the instrument geometric factor per sector of G = 3 × − cm str eV/(eV counts/electron) (Burch et al., 2007). Thisgeometric factor must be updated by appropriate correction factors to ac-count for spacecraft blockage and change in the instruments microchannelplate detection efficiency. Measurements inside a 20 minute diamagnetic re-gion on 26 July 2015 were used for calibration. Our assumption here is thatthe electron gas inside the diamagnetic cavity is isotropic. Counts in anodes0–7 and elevation steps 5–15 were averaged over time to obtain a nominalisotropic count value per energy step. These sectors are free from spacecraftblockage, include the Sun viewing FOV, and anodes are considered healthy.Next, for every sector a correction factor was calculated by taking the ra-tio of the sector count to the isotropic count. A similar method of in-flightcalibration was employed by Broiles et al. (2016) in an earlier stage of themission using solar wind data. The particle differential flux in anode i andelevation j measured at energy step k is determined from (Madanian et al.,2016b; Broiles et al., 2016): J ijk = ˙ c ijk G ∗ χ ijk ∗ E k (1)where ˙ c ijk is the count rate, χ ijk is the new correction factor, and E k isthe energy. For tables of χ ijk see Section 7. Employing the new correctionfactors also improved the low amplification rates at low energies. Figures A.1and A.2 in the appendix section show a comparison of the fluxes based ondifferent geometric factors. We consider IES background noise to be smalland we did not subtract a constant background rate from data. Energyspectra are shifted in energy to correct for the spacecraft potential. For anegative (positive) potential, fluxes are shifted to higher (lower) energies.The spacecraft potential can also deflect electron trajectories; however, fortypical spacecraft potentials near perihelion ( ∼ −
10 V), deflection of highenergy electrons (
E > ∼
100 eV) is very low and negligible (Scime et al.,1994). 6 .2. MAG
We used calibrated magnetic field data from the MAG instrument on theRosetta spacecraft (Glassmeier et al., 2007). The MAG instrument has two3-axis fluxgate magnetometers mounted on a 1.5 m boom with 15 cm separa-tion in between them. Entering the magnetic field-free regions in April 2015provided an opportunity to recalibrate the sensors and modify the tempera-ture model for that period (Goetz et al., 2016b). Our analysis is restrictedto time periods for which calibrated magnetic field data are available. Themagnetic field data are low band pass filtered to one second temporal reso-lution.
The LAP instrument consists of two spherical Langmuir probes, LAP1and LAP2, mounted on two booms extending 2.2 and 1.6 m from the space-craft, respectively (Eriksson et al., 2007). The instrument sweeps throughvoltage biases across the probes and the spacecraft ground to retrieve current-voltage curves that are subsequently used to derive plasma density and tem-perature, and spacecraft potential. LAP densities have been cross calibratedby corresponding MIP measurements and the data product has a variabletime resolution less than three minutes.
The two pressure gauges on the ROSINA/COPS instrument provide den-sity measurements of the neutral coma (Balsiger et al., 2007). The nudegauge density measurements have been adjusted for a water dominated comawhich is expected near perihelion (Heritier et al., 2017; Luter et al., 2018;Gasc et al., 2017). Neutral densities in the coma are mostly smooth andchange on the timescale of the comet ' s rotation.7 igure 1: Differential electron flux energy spectra integrated over the entire IES FOV inside(red) and outside (blue) the diamagnetic regions on (a) 30 July 2015 and (b) 3 August 2015.The corresponding spacecraft potentials are annotated on these panels. Error estimatesdue to the counting statistics are smaller than the line widths at most energies and arenot shown. The horizontal green and purple lines on panel (a) highlight energy rangesof dominant cometary and solar wind electrons, respectively. (c) The Rosetta spacecrafttrajectory around the comet. (d) Surface map of the comet illustrating latitudes andlongitudes in the ESA/RMOC frame.
3. Observations
The reduced electron flux inside diamagnetic regions has been discussedin a couple of studies (Madanian et al., 2016a; Nemeth et al., 2016; Timaret al., 2017). Figure 1 shows two examples of IES electron spectra inside and8utside diamagnetic regions, each exhibiting a decreased flux over differentenergies. The top-left panel shows an event on 30 July 2015 and the top-rightpanel shows an event few days later on 3 August 2015. The ordinate axis inthese plots represents the differential electron flux integrated over the entireIES FOV (2 . π solid angle). The red (blue) lines on the top panels show thetime averaged spectra inside (outside) the diamagnetic cavities. The periodinside the diamagnetic cavity on 30 July is from 11:00:52 to 11:11:40 UTCand on 3 August from 17:20:42 to 17:28:03 UTC. The outside periods for30 July and 3 August are selected between 10:52:26 – 11:00:46 UTC and17:12:16 – 17:20:36 UTC, respectively. The horizontal green and purple lineson panel (a) are shown as references to highlight energy ranges of dominantcometary and solar wind electrons (Madanian et al., 2016a). The verticaldashed lines show the energy range in which a flux difference is observed(lower and upper energies).On 30 July, flux of electrons in the ∼ −
350 eV range inside thediamagnetic region has decreased by variable amounts. This energy rangeextends to around 900 eV on 3 August. A characteristic energy indicated by ' Max. drop ' at around 175 eV on panel (b) is the energy at which the highestflux difference is observed. This energy for the event on 30 July is around74 eV. Panel (c) in Figure 1 shows the Rosetta spacecraft trajectory aroundthe comet between 30 July and 6 August 2015. The colorbar representsthe time. The reference frame in this plot is the dynamic body-CenteredSolar Equatorial (CSEQ) frame in which the + x axis is toward the Sun,the + z axis is aligned with the projection of the solar rotation axis on aplane perpendicular to the x axis, and the y axis completes the right-handcoordinate system. The frame ' s origin is the comet ' s center of mass (shownwith a black dot). Rosetta was at around 180 km from the comet on 30 July,and it gradually moved to a distance of 250 km north-east of the comet on6 August. The spacecraft speed with respect to the comet was a few metersper second. We will discuss latitudinal dependence of variables. The latitudeis measured in the ESA/RMOC shape frame (also known as the landmarkcoordinates) illustrated in the surface map of the comet in panel (d). Colorsrepresent different longitudes, while latitudes are annotated on the map.The flux difference across the diamagnetic boundaries creates an energydensity difference between inside and outside plasmas. As seen in Figure 1,the energy range of flux difference varies for different diamagnetic events,and this variability has not been studied so far. In Section 3.1 we providea statistical analysis of a subset of diamagnetic events and in Section 3.2 a9etailed case study for one of these events is presented. We use a subset of diamagnetic events reported in Goetz et al. (2016b)and limit our study to July and August of 2015, when comet activity wasrelatively high and the majority of diamagnetic events were observed. Withthe IES measurement cycle in mind, we down-selected events lasting longerthan 256 s and with at least 512 s separation from another event on atleast one side. These criteria ensure that at least one full IES measurementcycle exist inside the diamagnetic region and that the outside measurementsare not contaminated by shorter events. This brought down the number ofevents from a total of 313 to 62 events. For the list of events see Section7. We used an algorithm to search and compare the IES energy spectrainside and outside each event and record energy bins with reduced electronfluxes. For 31 events we had the option to choose the outside spectrumfrom the trailing or the leading side. For these cases measurements fromthe side with the higher magnetic field strength were selected. For eventsthat showed multiple drops corresponding to multiple energy ranges (i.e.,the inside spectrum would drop below the outside spectrum multiple timesdue to similar overlapping spectra,) the widest energy range was recorded.We present our observations in the context of total energy flux difference,∆ ψ , across boundaries as seen in IES electron spectra and defined by:∆ ψ = E upper (cid:88) k = E lower ( ψ ( E k ) out − ψ ( E k ) in ) × E k (2)10
00 200 300 400
Comet distance (km) ( e V / c m s s t r e V ) (a) 10 Neutral density (cm -3 ) (b) -40 -20 0 20 40 Latitude 0 500 1000 1500 N e (cm -3 ) (c)0 90 180 Cone angle ( ° ) ( e V / c m s s t r e V ) (d) 0 50 100 |B| (nT) (e) 00:00 15:00 30:00 Duration (mm:ss) (f)0 500 1000
Upper E (eV) ( e V / c m s s t r e V ) (g) 0 100 200 E at max. drop (eV) (h) -0.5 0 0.5 1 dN e (i) Figure 2: Distributions of plasma parameters for 62 diamagnetic events included in thestudy. The ordinate axis in all panels shows the parameter ∆ ψ . Panels (a - c) show thedistributions of cometocentric distance, local neutral density, and bulk electron density,respectively. Data in these panels are also color coded by the cometary latitude. Panels (d- f) show distributions of cone angle, magnetic field strength, and event duration. Panels(g - i) show the distributions of upper energy limit of flux variations, energy of the highestflux difference, and bulk electron density difference. The red star marks the event on 30July 2015 which is considered further in Section 3.2. where ψ ( E k ) is the integrated differential electron flux over the IES FOV atenergy E k . ∆ ψ distributions against several other parameters are presented11n Figure 2. The first row in this figure shows ∆ ψ as a function of cometo-centric distance, neutral density and bulk electron density N e measured byLAP instrument, respectively. These panels are also color coded based onthe cometary latitude at each event (see Figure 1, panel (d)). As shown inpanel (a), observations are mostly within 300 km from the comet and neutraldensities varies between 5 × − cm -3 . The neutral densities also showa clear latitudinal dependence. Data in panel (b) shows that the comet issignificantly more active in the southern hemisphere (Hansen et al., 2016;Hassig et al., 2015; Luter et al., 2018). During the perihelion passage, thesouthern hemisphere of the comet receives higher insolation and this periodis in the midst of the southern summer in cometary seasons (Keller et al.,2015).Panel (c) shows LAP electron densities measured at the beginning ofeach diamagnetic crossing. The LAP densities are clustered around 200 cm -3 and 1000 cm -3 . The higher density cluster, corresponding to events over thesouthern latitudes where the comet activity is higher, shows more variations.The lower density events around 200 cm -3 are more contained and show lessvariations. A few points that exhibit the highest ∆ ψ are within this group.Since neutral densities show gradually increase with respect to decreasinglatitude, one would expect to see a gradual increase in LAP electron densitiesat lower latitudes. However, there is a distinct separation in electron densitiesmeasured in the southern versus northern latitudes. This may reflect thatthe bulk radial plasma velocity is higher on the less active side. A reasonfor this could be that ion-neutral collisions, on the less active side, occur lessfrequently and thus are less efficient in hampering ion-acceleration along anambipolar electric field (e.g., Vigren and Eriksson (2017)). In addition, ina simplified view, an equally pronounced outward radial acceleration on themore active side would conflict with momentum conservation. Most eventsin the southern hemisphere occurred between 26 July and 3 August, whenmost of the long-lasting diamagnetic events have been observed.Panels (d - f) show, respectively, distributions of the magnetic field coneangle, magnetic field strength, and event duration. The cone angle definesthe angle between the magnetic field vector and the comet-Sun line. The dis-tribution in panel (d) shows events grouped around 30 ◦ and 150 ◦ cone angleswhich is expected for the observations near perihelion as significant magneticfield draping exists and the spacecraft resides mostly in the terminator planeat this time. Correlation between electron number flux and magnetic fieldmagnitude slightly increases at higher energies when all measurements at12erihelion are included (Madanian et al., 2016a), but the ∆ ψ distributionin panel (e) exhibits no or a very weak correlation with the magnetic fieldstrength. Event durations varied between 257 seconds and 32 minutes. Thelongest event that also shows the highest ∆ ψ is on 7 July 2015 at 09:44:22UTC. The outside spectrum is selected from the trailing side of that eventthe flux difference extends up to 733 eV.The third row in Figure 2 shows ∆ ψ , respectively, versus the upper energylimit of flux difference, energy of the highest flux difference, and relativedifference in bulk electron density between inside and outside plasmas, dN e =( N e out − N e in ) /N e out . The histogram in the background of panel (g) showsthe occurrence rate of the upper energy limit. Although upper limits spreadacross many energies, the distribution suggests that the flux decrease stopsat certain energies more often. The first peak in the histogram at 350–400 eV bin is the most dominant and includes 17 events. Flux differencefor nine events extends up to 650–700 eV (the second highest peak). Wewill revisit this point in Section 3.2. ∆ ψ decreases when the most affectedelectrons are at higher energies which can be observed in panel (h). Inaddition, panel(i) shows that for most events bulk electron density insidethe diamagnetic region decreases, confirming previous findings using MIPdata (Henri et al., 2017), though this decrease shows no apparent relationwith IES flux differences. Suprathermal electrons at 100 or 200 eV travelthrough the plasma at speeds significantly faster than bulk electrons. Theirflux variability occurs on time scales much different than the bulk plasmavariation observed inside diamagnetic regions (Hajra et al., 2018). The diamagnetic cavity event that we consider in this section was shownin Figure 1 panel (a). It is observed at negative latitudes and is one of the 17events for which flux difference extends to ∼
350 eV (see panel (g) in Figure2). Table 1 lists plasma and field parameters around this event.To better understand the nature of the reduced fluxes during the tran-sition into the diamagnetic region we examine the 3D spatial distributionsof high energy suprathermal electrons. Figure 3 shows 2D cuts of electrondistribution variations in the IES FOV for four timestamps before the dia-magnetic event on 30 July 2015. Panel (a) shows the differential electron fluxfor IES anodes (labeled 0-15) averaged around the central elevation plane atthe first timestamp and is labeled as the ”reference” distribution. The colors13 able 1: Plasma and field parameters for diamagnetic cavity event on 30 July 2015 r comet (km) 179.5 D sun (AU) 1.24Neutral density (cm -3 ) 6 . × Latitude -48Cone angle 149.3B (nT) 38.8Duration 00:10:55 (11:00:51 - 11:11:41 UTC)Energy range of reduced flux (eV) 56.1 - 358Energy of max. flux difference(eV) 74.4Inside OutsideLAP bulk electrons density (cm -3 ) 997.3 1164.8are in logarithmic scale and energies between 100 eV and 5 keV are shown.Panels (b - d) show the flux ratios in the next three timestamps (all still out-side the cavity) as compared to the reference distribution. The disconnectionat 3 o ' clock on these panels is an artefact of the plotting software.Relative enhancements (red segments) are observed in anodes 0, 6, 8, and12 of panels (b), (c), and (d); while decreases (blue segments) occur in anodes2, 14, and 15 of panels (b) and (d) and in anodes 4, 6, 12 of panel (c). Fromthis figure we notice directional changes for electrons at different energiesclose to the diamagnetic cavity. It is important to consider these changes inthe electron trajectory with respect to the magnetic field. To better analyzethese spatial changes, we analyze the electron pitch angle distributions.14 ReferenceDistribution4 2 0 2 4
Log Energy (eV)
Log E ne r g y ( e V ) (a) Log E l e c r on f l u x ( / c m s s t r e V ) Log Energy (eV)
Log E ne r g y ( e V ) (b) -1-0.500.51 Log R a t i o s ( F l u x / R e f. ) Log Energy (eV)
Log E ne r g y ( e V ) (c) -1-0.500.51 Log R a t i o s ( F l u x / R e f. ) Log Energy (eV)
Log E ne r g y ( e V ) (d) -1-0.500.51 Log R a t i o s ( F l u x / R e f. ) Figure 3: 2D cuts of the IES FOV showing electron differential flux variations in fourtimestamps between 10:46 and 11:00 UTC before the diamagnetic cavity crossing on 30July 2015. Panel (a) shows the electron differential flux at the first timestamp. Panels (b- d) show the corresponding flux ratios with respect to the distribution in panel (a).
We should note that electron PAD is not an official data product of theIES instrument. Few factors that may complicate derivation of PADs andlimit our ability to interpret them include, (1) low time resolution in IES datadoes not allow to resolve plasma effects such as wave-particle interactions inthe distributions, (2) IES FOV does not cover the full sky and if the magneticfield points toward these gaps in the FOV (i.e. instrument symmetry axis,)part of the distribution will be lost, and (3) IES onboard averaging can re-duce the resolution of the derived PADs. It is not our intention to study finetimescale effects on electrons, but rather we are looking at effects of changingmagnetic field topology and our results prove that PADs at the current res-olution can provide valuable information about those effects. We inspected15he IES FOV for pitch angle coverage and ensured that the magnetic fielddirection during this event is favorable for PAD analysis.The IES time resolution for a full cycle in the current mode is 256 s,resulting in a 2 s sampling time per energy bin. At each energy step, thedeflector plates are biased in a see-saw fashion to conserve power and reducesweep time. We track the time at which different energies and sectors werescanned within a cycle and update the magnetic field vector accordinglybefore calculating the pitch angles. An array consisting of 12 bins, each15 ◦ wide, is used to sort fluxes into the pitch angle space. To account forstraddling of sectors that covered more than one pitch angle bin, sector fluxis distributed across all overlapping bins and the final PADs are normalizedby the sampling rate at each bin.The event on 30 July 2015 at 11:00:51 UTC meets our selection crite-ria. Specifically, we searched for periods of gradual changes in magneticfield strength over a few consecutive IES timestamps, where high amplitudemagnetic field fluctuations were relatively low, as they can modulate the dis-tribution faster than the IES can record and therefore cannot be studied. Forthe event studied in this section, although we do not observe the typical sig-natures of ultra-low frequency (ULF) waves, or circularly polarized whistlerwaves (see panel (a) of Figure 4), we have to assume that wave-particle in-teractions are negligible.Figure 4 shows an overview of magnetic field data and electron PADsacross this event. The top panel in this figure shows the magnetic field com-ponents and magnitude in the CSEQ coordinates. The diamagnetic cavityevent is identified between 11:00:51 and 11:11:41 UTC. The cone angle ( θ cone )is shown in panel (b). The spectrogram in panel (c) shows the FOV inte-grated differential electron flux (cm s eV) -1 as a function of energy in the200-1000 eV range. Flux reductions inside the cavity for this event were pre-viously illustrated in panel (a) of Figure 1, and can also be identified in panel(c). Panels (d - h) show the electron PAD time series at different energiesnormalized by the maximum flux value in each panel. The distributions havebeen averaged over consecutive energy bins to improve the counting statis-tics. The energy ranges are specified in the parentheses. All colorbars are inlogarithmic scale. The white lines overplotted on these panels are contoursof constant magnetic moment, µ m = W ⊥ / | B | , where | B | is the magnetic fieldmagnitude and W ⊥ = 1 / m e V ⊥ is the perpendicular energy of electrons.The pitch angle distributions and contours inside the cavity have no physicalmeaning. 16 a) -50050 B cs eq ( n T ) |B|BxByBz (b) c one ( ° ) (c) E ( e V ) D i ff. f l u x (d) PA ( ° ) ( - ) (e) PA ( ° ) ( - ) (f) PA ( ° ) ( - ) (g) PA ( ° ) ( - ) I n t en s i t y r a t i o s ( f l u x / M a x . f l u x ) (h)
30 Jul 2015 - UTC (hh:mm) PA ( ° ) ( - ) t t t t Cavity
Figure 4: Magnetic field and electron distribution time series around the diamagneticcavity on 30 July 2015. The field free cavity is observed between 11:00:51 and 11:11:41UTC and is marked with a grey box. Panel (a) shows magnetic field components andmagnitude in CSEQ coordinates, panel (b) shows the magnetic field cone angle, panel (c)is the differential electron flux spectrogram in units of log (cm s eV) -1 , and panels (d -h) show electron pitch angle distributions in five different energy ranges. The fluxes arenormalized by the maximum flux value in each panel. The white lines on these panels arethe contours of the constant adiabatic invariant. The vertical dashed-dotted lines markfour IES timestamps before the onset of the diamagnetic cavity. Between 10:53:00 and 11:00:00 UTC, the magnetic field shows, on average,a gradual decrease in the field strength. There are perturbations due to theturbulent plasma environment. The B x component is shown with the blue17olor in panel (a) of Figure 4, and is highly negative throughout this period.In fact, most of the variations in the magnetic field strength originates fromthe B x component while the two other components are relatively quiet. Closeto the diamagnetic region the y component of the field becomes dominantand shows a continuous decline. The magnetic field direction changes fromanti-sunward (cone angle ∼ ◦ ) to a direction perpendicular to the comet-Sun line (cone angle ∼ ◦ ). This period corresponds to four IES timestampsidentified by vertical dashed-dotted lines drawn across all panels and labeledby t , t , t , and t .At 10:45 UTC electrons show a fairly scattered distribution occupyingmost of the pitch angle bins with similar intensities, except for the distri-butions in panels (f) and (g). In the next four timestamps, flux reductionsaround 90 ◦ pitch angles are observed and accompanied by increased fluxesin directions parallel (0 ◦ ) and anti-parallel (180 ◦ ) to the magnetic field. Thisis indicative of a changing distribution from isotropic to field-aligned. Theeffect is particularly evident for 151 −
293 eV electrons, while 306 − −
440 eVelectrons in panel (h) show a different pattern. With the exception of times-tamp ( t ) where an enhanced anti-parallel flux is observed, distributions arerelatively disordered and chaotic with respect to the onset of the diamag-netic cavity, or the adiabatic invariant contours. This implies that the firstadiabatic invariant is only conserved up to a certain energy.In Figure 5 we examine these spectra in a more quantitative way. Inpanels (a - d), differential electron flux at selected energies are plotted versuspitch angle. Each panel corresponds to an IES timestamp marked with ver-tical, dashed lines in Figure 4. The corresponding energies are annotated inpanel (a), and error bars reflect the uncertainty due to the counting statistics.Error estimates for most data points are reasonably low and for a few pointsare larger. Larger error bars do not necessarily indicate that the observationmust be discarded, but rather more measurements are needed to improve theconfidence on the observation.The 99 eV electrons have a maximum in the parallel direction until 10:52UTC. In the next timestamp, the anti-parallel flux increases while the fluxesin ∼ − ◦ pitch angles decrease. Given that between 10:52 and 10:56UTC the magnetic field vector rotates to mostly − y direction, changes in 99eV PADs indicate that a large flux of these electrons travel anti-parallel to18he magnetic field and away from the comet. The 99 eV line also shows asharp peak at around 50 ◦ at 11:00 UTC. The 185 eV electron distribution inpanel (a) shows a rapid fall in the last pitch angle bin. This is most likelydue to the low count rate in that bin, as is evident by the larger error bars.The 185 eV distribution changes into a bidirectional, field-aligned pattern inthe next three timestamps. Similarly, the 202 eV (cyan) and 250 eV (green)electrons start roughly isotropic and evolve into double-peak bidirectionaldistributions, while 90 ◦ electrons become depleted. The net flux in thesedistributions remains almost the same from one timestamp to the next. Inother words, the enhancements in pitch angles near 0 ◦ and 180 ◦ (as seenin panels (c) and (d)) are compensated by the depletions in ∼ ◦ bins.The 358 and 396 eV lines, although moderately changing in time, do notshow any depletion in perpendicular flux. The energies we discussed here aresensitive traces of the magnetic field topology. The depletion of ∼ ◦ pitchangle electrons is consistent with adiabatic transport of electrons in sharplydecreasing magnetic fields.
45 90 135
Pitch Angle D i ff e r en t i a l f l u x ( c m s e V s t r) - t = 10:48 Pitch Anglet = 10:52 (b) 45 90 135 Pitch Anglet = 10:56 (c) 45 90 135 Pitch Anglet = 11:00 (d) Figure 5: Differential electron flux at selected energies (different colors) versus pitch angleat four timestamps prior to the cavity encounter on 30 July 2015. Energies are annotatedin panel (a). Each panel correspond to a timestamp identified by vertical dashed-dottedlines in Figure 4. t show noticeably more depletion around 90 ◦ pitchangles than the neighboring timestamps, but the net fluxes are still higherthan those inside the cavity. The origin of this behavior has not been clearlyidentified at this time.
4. Electron Dynamics, Discussion and Interpretations
The magnetic field in the induced cometary magnetosphere is significantlyhigher than the IMF (50 nT in Figure 4 compared to a typical IMF ∼ ∼
70 eV (Madanian et al., 2016a). Though it should be noted that effects ofacceleration processes were not included in the models. Also the differentialfluxes shown in Figure 5 are higher than the typical flux of solar wind haloelectrons at 1 AU, but lower than the typical strahl component (Andersonet al., 2012; Graham et al., 2017).The suprathermal electrons that interact with the weakened magneticfield near diamagnetic regions are subject to adiabatic cooling and are re-distributed to conserve the magnetic moment (see Figure 4), resulting indecrease of the perpendicular flux and enhancement of fluxes along the mag-netic field. The field-aligned electrons tied to the magnetic field will havelimited access to the diamagnetic regions. For instance, prior to the dia-magnetic event on 30 July 2015 (the four timestamps marked in Figure 4),electrons up to around 350 eV are effectively rearranged to field-aligned dis-tributions. The flux of electrons in these energies is also reduced inside thecavity. Higher energy electrons (
E >
350 eV) do not exhibit field-aligneddistributions in response to the decreasing magnetic field, and their intensityalso remains unchanged during the passage of the diamagnetic cavity.A possible explanation of this behavior could be associated with viola-tion of the first adiabatic invariant beyond certain energies. It has beenshown that in a region of highly curved magnetic field lines where the cur-vature radius is comparable to the electron gyroradius, resonance between20he two leads to a significant scattering in electron gyration and aberra-tion of the first adiabatic invariant (Bchner and Zelenyi, 1989; Young et al.,2008; Zhang et al., 2016). In simple terms, the adiabatic invariant parameter κ = (cid:112) R c /R L , where R c is the curvature radius and R L is the electron gyro-radius, governs the energy threshold of the scattered electrons. Theoreticalstudies have shown that for κ <
5, curvature scattering occurs in almost allpitch angles and any anisotropic electron distribution becomes isotropic andchaotic (Bchner and Zelenyi, 1989). This prediction has also been confirmedby observations near ion diffusion zones in the Earth ' s magnetotail whereelectrons interact with transient diamagnetic plasmas (Young et al., 2008;Wang et al., 2010, 2019).It is natural to estimate the radius of the field line curvatures near thediamagnetic region based on this simple approach and the plasma parametersin Table 1. The gyroradius of a 350 eV electron in a 38 nT magnetic field isaround 1.65 km, which implies a minimum curvature radius of R c ∼
40 km.We speculate that the diamagnetic structure is at least 80 km wide and theshape is not necessarily symmetric.
Our estimate of the field line curvature around the diamagnetic regionsis much smaller than the spacecraft distance to the comet, which suggeststhat the observed structures are in fact smaller than a global diamagneticcavity surrounding the comet. Given the spacecraft ' s relatively stationaryposition around the comet, and recurring diamagnetic encounters with differ-ent durations and sizes, it appears that what Rosetta observed were detachedunmagnetized plasma clouds or filaments that convected over the spacecraft.Figure 6 shows a schematic illustration of a proposed scenario for the inter-action region around the comet (also see Figure 5 in Goetz et al. (2016b) andFigure 6 in Henri et al. (2017)). The solar wind is incident from the right,the neutral coma is shown with a fading blue color, and the extent of thefield free region is marked with the dashed line. The comet is surrounded bya region of enhanced magnetic field. The draped IMF lines are shown withsolid, grey lines. The size of the cycloids next to electrons in this figure isa representation of the pitch angle. Solar wind electrons with larger pitchangles are reflected while small angle electrons reach the inner coma. Theschematic in Figure 6 is a conceptual view to illustrate our findings and doesnot capture all aspects of the interaction such as microstructures within theboundary layer. 21 O + H O + H O + H O + H O + e - e - e - e - e - H Oe - I M F IES e - e - e - e - e - e - H O + D i a m a g n e t i c B o u n d a r y Reflected Electrons
Unmagnetized Plasma Filaments u " B I M F Figure 6: Schematic illustration of a possible interpretation of Rosetta observations of dia-magnetic regions around comet 67P, showing the electron dynamics and the asymmetricalexpansion of unmagnetized plasma structures.
Fluid and hybrid models suggested that a global diamagnetic boundarycan form at around 25 km, much closer to the comet that has been ob-served (Koenders et al., 2015; Rubin et al., 2014). Simple one-dimensionalMHD models (Cravens, 1986; Ip and Axford, 1987), although can predictthe boundary distance at times, are not always consistent with the observa-tions and models are in fact over simplified. For instance, it is assumed thations are in chemical equilibrium, which is not always the case close to theboundary. Therefore, other mechanisms are perhaps in play that give riseto diamagnetic plasmas at the point of the spacecraft. Several authors haveproposed different formation mechanisms (Henri et al., 2017; Huang et al.,2016). Kelvin-Helmholtz instabilities and transient snowplow structures havebeen proposed as processes that can accelerate unmagnetized ions parallel22o the field downstream (Goetz et al., 2016a; Haerendel et al., 1986; Koen-ders et al., 2015). In this picture, the nonuniform neutral production andcometary ion outflow create an inhomogeneous plasma environment favor-able for snowplow structures. Cometary ions are picked up and acceleratedby the solar wind at different rates. Depending on the mass loading rate andsolar wind motional electric field strength and orientation, ions can be accel-erated in transverse directions, causing elongation of the structure boundaryand the underlying outflowing unmagnetized plasma in a particular direction(see the unmagnetized plasma filaments in Figure 6). These structures canalso become detached from the core unmagnetized plasma (Halekas et al.,2016). Studying the evolution of these structures as they form upstream andpropagate to the point of the spacecraft could provide more credence to thishypothesis and is left for a future study.
We focused on events around perihelion when the comet neutral out-gassing rate is significantly high. Yet, it is unlikely that collisional processes,mainly the electron-neutral collisions, undermine our results. Assuming anoutgassing rate of Q = 10 s -1 near perihelion, the total neutral density at r = 200 km from the comet is n ( r ) = Q/ (4 πr u n ) ≈ cm -3 , where weassumed a spherically symmetric coma (Haser, 1957) and a neutral outflowspeed of u n = 1 km/s. At 350 eV, the total (elastic and inelastic) electron-neutral collision cross section for water molecules is ∼ . × − cm , re-sulting in a collisional mean free path, λ mfp , of less than 300 km. For a 50 eVelectron, λ mfp reduces to 100 km. Since these length scales are much largercompared to electron gyroradii, scattering due to electron-neutral collisionsis inefficient at the energies considered. Madanian et al. (2016b) discussedthe effects of an ambipolar electric field that bounds suprathermal electronsto the inner coma, causing them to become further thermalized by collisionalprocesses. The ambipolar electric field is fundamentally parallel to the mag-netic field. However, if an ambipolar electric field was in play to accelerateelectrons to energies as high as 250 eV in the parallel direction, we shouldshould have seen a noticeable shift in energy spectra at lower energy electronsas well. Furthermore, the ambipolar electric field is a transient effect, andonce charge separation occurs, electrons (ions) are decelerated (accelerated)to nullify the ambipolar field and retain charge quasi-neutrality. The factthat we see changes in PADs over several IES timestamps indicates that it isunlikely for high energy electrons to be driven by an ambipolar electric field.23nother observation on the 30 July event is the presence of a peak in the185 eV line at around 140 ◦ in panel (b) or a peak in the 99 eV line at around50 ◦ in panel (d) of Figure 5, which can be interpreted as conic structures.The 202 and 250 eV lines in panels (c) and (d) of that figure, resemblebutterfly distributions. Although several mechanisms (such as interactionwith magnetosonic waves) can produce butterfly distributions, we shouldnote that relatively larger error bars, for instance on the first three points of99 eV line in panel (d), indicate that the distribution may be under-sampledon one side of the peak. Low angular resolution of the available data limitsour ability to interpret finer details and PADs with large error bars must beviewed with caution. The ion sensor on the IES measured noticeable amount of low energy pickup ions (10 < E <
50 eV) with very low but above background counts ofhigher energy ions up to around 3 keV. Near perihelion, Rosetta was insidea solar wind ion cavity (Nilsson et al., 2017) and we did not observe anyparticular consistent pattern in ion energy distributions during the diamag-netic boundary crossings. More comprehensive analyses must be performedin the future using both the IES and the other electrostatic analyzer, the IonComposition Analyzer (ICA) (Nilsson et al., 2007).
5. Conclusions
In this paper, we analyzed the dynamics of bulk and suprathermal elec-trons around comet 67P for a subset of long-lasting diamagnetic events. Forevents observed over the southern hemisphere bulk electron densities arenoticeably higher than those over the northern side, even after taking neu-tral density and distance variations into account. Most events show lowerbulk electron densities inside the diamagnetic region compared to the outsideplasma (Henri et al., 2017). Suprathermal electron fluxes inside the cavitiesshow reductions over different energy ranges, most noticeably extending upto ∼
350 eV for 27 percent of events and up to ∼
700 eV for 14.5 percentof events. We present a first glance at suprathermal electron PADs at closeproximity of an active comet. We propose a mechanism for interpreting PADvariations associated with the changing magnetic field topology, which is asfollows: near diamagnetic regions, suprathermal electrons are transported24diabatically and rearrange from isotropic to field-aligned directions to con-serve the magnetic moment. The field-aligned electrons tied to the magneticfield have limited access to diamagnetic regions, causing the observed de-crease in electron flux between inside and outside the cavities. Electronsbeyond certain energies behave non-adiabatically when their gyroradius iscomparable to the magnetic field line curvature. These electrons do notshow a systematic flux difference. For the diamagnetic event on 30 July 2015at 11:00:51 UTC the size of the diamagnetic region is estimated to be at least80 km. Why flux reductions for many events extend up to ∼ −
400 eV isnot completely understood yet and should be investigated further in futurestudies.We considered fairly long and isolated events when Rosetta spent fewminutes inside each diamagnetic region, but there are shorter diamagneticencounters that we did not discuss. We showed PADs in a generally decay-ing magnetic field environment. Small scale variations of the magnetic fieldcan have an influence on PADs and further measurements and analyses areneeded to verify the results shown in this paper.Understanding the detailed behavior of electrons around comet 67P de-mands fully kinetic models with sufficient knowledge of the electromagneticfields. The magnetic field structure around this comet is not a simple one.During some diamagnetic events the magnetic field showed a gradual decreasein strength before entering the diamagnetic plasma, but a sudden increaseafterward. This asymmetry in the front and back envelopes requires furtherconsiderations. Analysis of suprathermal electron pitch angle distributionsnear perihelion and at other phases of the mission can provide another toolfor studying plasma phenomena at this comet. Through Rosetta observa-tions we now know that a cometary plasma environment is an exciting spaceplasma laboratory that deserves another visit not only to make improvedplasma measurements, but to use new measurement techniques to study theevolution of plasma boundaries around an object with an inhomogeneousatmosphere.
6. Acknowledgments
Rosetta is a European Space Agency (ESA) mission with contributionsfrom its member states and National Aeronautics and Space Administration(NASA). Work at Imperial College London was supported by STFC of UKunder grant ST/N000692/1. EV is grateful for support from SNSA (Dnr2566/14). The work of ZN was supported by the NKP-18-4 New NationalExcellence Program of the Ministry of Human Capacities and the BolyaiJanos Research Scholarship of the HAS. MR acknowledges the support of theState of Bern and the Swiss National Science Foundation (200020-182418).
7. Data Availability
All data used in this study can be accessed via the ESA Planetary Sci-ence Archive ( https://archives.esac.esa.int/psa ) as well as the NASAPlanetary Data System ( https://pds.nasa.gov ). Tables of correction fac-tors to the IES geometric factor and the list of events used in our statisticalstudy are included as Supplementary materials.
References
Anderson, B.R., Skoug, R.M., Steinberg, J.T., McComas, D.J., 2012. Vari-ability of the solar wind suprathermal electron strahl: VARIABILITY OFELECTRON STRAHL. Journal of Geophysical Research: Space Physics117, n/a–n/a, doi: .Auster, H.U., Apathy, I., Berghofer, G., Fornacon, K.H., Remizov, A., Carr,C., Guttler, C., Haerendel, G., Heinisch, P., Hercik, D., Hilchenbach, M.,Kuhrt, E., Magnes, W., Motschmann, U., Richter, I., Russell, C.T., Przyk-lenk, A., Schwingenschuh, K., Sierks, H., Glassmeier, K.H., 2015. Thenonmagnetic nucleus of comet 67p/Churyumov-Gerasimenko. Science 349,aaa5102–aaa5102, doi: .Balsiger, H., Altwegg, K., Bochsler, P., Eberhardt, P., Fischer, J., Graf,S., Jckel, A., Kopp, E., Langer, U., Mildner, M., Mller, J., Riesen, T.,Rubin, M., Scherer, S., Wurz, P., Wthrich, S., Arijs, E., Delanoye, S.,Keyser, J.D., Neefs, E., Nevejans, D., Rme, H., Aoustin, C., Mazelle,C., Mdale, J.L., Sauvaud, J.A., Berthelier, J.J., Bertaux, J.L., Duvet,L., Illiano, J.M., Fuselier, S.A., Ghielmetti, A.G., Magoncelli, T., Shelley,E.G., Korth, A., Heerlein, K., Lauche, H., Livi, S., Loose, A., Mall, U.,Wilken, B., Gliem, F., Fiethe, B., Gombosi, T.I., Block, B., Carignan,G.R., Fisk, L.A., Waite, J.H., Young, D.T., Wollnik, H., 2007. RosinaRosetta Orbiter Spectrometer for Ion and Neutral Analysis. Space ScienceReviews 128, 745–801, doi: .26ingham, R., Shapiro, V.D., Tsytovich, V.N., de Angelis, U., Gilman, M.,Shevchenko, V.I., 1991. Theory of wave activity occurring in the AMPTEartificial comet. Physics of Fluids B: Plasma Physics 3, 1728–1738, doi: .Broiles, T.W., Livadiotis, G., Burch, J.L., Chae, K., Clark, G., Cravens,T.E., Davidson, R., Eriksson, A., Frahm, R.A., Fuselier, S.A., Goldstein,J., Goldstein, R., Henri, P., Madanian, H., Mandt, K., Mokashi, P., Pol-lock, C., Rahmati, A., Samara, M., Schwartz, S.J., 2016. Characteriz-ing cometary electrons with kappa distributions: CHARACTERIZINGCOMETARY ELECTRONS. Journal of Geophysical Research: SpacePhysics 121, 7407–7422, doi: .Burch, J.L., Goldstein, R., Cravens, T.E., Gibson, W.C., Lundin, R.N., Pol-lock, C.J., Winningham, J.D., Young, D.T., 2007. RPC-IES: The Ion andElectron Sensor of the Rosetta Plasma Consortium. Space Science Reviews128, 697–712, doi: .Bchner, J., Zelenyi, L.M., 1989. Regular and chaotic charged particle motionin magnetotaillike field reversals: 1. Basic theory of trapped motion. Jour-nal of Geophysical Research 94, 11821, doi: .Carr, C., Cupido, E., Lee, C.G.Y., Balogh, A., Beek, T., Burch, J.L.,Dunford, C.N., Eriksson, A.I., Gill, R., Glassmeier, K.H., Goldstein, R.,Lagoutte, D., Lundin, R., Lundin, K., Lybekk, B., Michau, J.L., Mus-mann, G., Nilsson, H., Pollock, C., Richter, I., Trotignon, J.G., 2007. RPC:The Rosetta Plasma Consortium. Space Science Reviews 128, 629–647,doi: .Cravens, T., Gombosi, T., 2004. Cometary magnetospheres: a tutorial. Ad-vances in Space Research 33, 1968–1976, doi: .Cravens, T.E., 1986. The Physics of the Cometary Contact Surface. Eur.Space Agency Spec. Publ. 244.Eriksson, A.I., Bostrm, R., Gill, R., hln, L., Jansson, S.E., Wahlund,J.E., Andr, M., Mlkki, A., Holtet, J.A., Lybekk, B., Pedersen, A.,Blomberg, L.G., The LAP Team, 2007. RPC-LAP: The Rosetta Lang-muir Probe Instrument. Space Science Reviews 128, 729–744, doi: . 27riksson, A.I., Engelhardt, I.A., Andr, M., Bostrm, R., Edberg, N.J., Johans-son, F.L., Odelstad, E., Vigren, E., Wahlund, J.E., Henri, P., 2017. Coldand warm electrons at comet 67p/Churyumov-Gerasimenko. Astronomy& Astrophysics 605, A15.Feldman, W.C., Asbridge, J.R., Bame, S.J., Montgomery, M.D., Gary, S.P.,1975. Solar wind electrons. Journal of Geophysical Research 80, 4181–4196, doi: .Fuselier, S.A., Feldman, W.C., Bame, S.J., Smith, E.J., Scarf, F.L., 1986.Heat flux observations and the location of the transition region boundaryof Giacobini-Zinner. Geophysical Research Letters 13, 247–250, doi: .Galand, M., Hritier, K.L., Odelstad, E., Henri, P., Broiles, T.W., Allen, A.J.,Altwegg, K., Beth, A., Burch, J.L., Carr, C.M., Cupido, E., Eriksson, A.I.,Glassmeier, K.H., Johansson, F.L., Lebreton, J.P., Mandt, K.E., Nilsson,H., Richter, I., Rubin, M., Sagnires, L.B.M., Schwartz, S.J., Smon, T.,Tzou, C.Y., Vallires, X., Vigren, E., Wurz, P., 2016. Ionospheric plasmaof comet 67p probed by
Rosetta at 3 au from the Sun. Monthly Noticesof the Royal Astronomical Society 462, S331–S351, doi: .Gasc, S., Altwegg, K., Fiethe, B., Jckel, A., Korth, A., Le Roy, L., Mall, U.,Rme, H., Rubin, M., Hunter Waite, J., Wurz, P., 2017. Sensitivity andfragmentation calibration of the time-of-flight mass spectrometer RTOFon board ESA’s Rosetta mission. Planetary and Space Science 135, 64–73,doi: .Glassmeier, K.H., 2017. Interaction of the solar wind with comets: aRosetta perspective. Philosophical Transactions of the Royal SocietyA: Mathematical, Physical and Engineering Sciences 375, 20160256,doi: .Glassmeier, K.H., Richter, I., Diedrich, A., Musmann, G., Auster, U.,Motschmann, U., Balogh, A., Carr, C., Cupido, E., Coates, A., Rother, M.,Schwingenschuh, K., Szeg, K., Tsurutani, B., 2007. RPC-MAG The Flux-gate Magnetometer in the ROSETTA Plasma Consortium. Space ScienceReviews 128, 649–670, doi: .28oetz, C., Koenders, C., Hansen, K.C., Burch, J., Carr, C., Eriksson,A., Frhauff, D., Gttler, C., Henri, P., Nilsson, H., Richter, I., Ru-bin, M., Sierks, H., Tsurutani, B., Volwerk, M., Glassmeier, K.H.,2016a. Structure and evolution of the diamagnetic cavity at comet67p/ChuryumovGerasimenko. Monthly Notices of the Royal Astronom-ical Society 462, S459–S467, doi: .Goetz, C., Koenders, C., Richter, I., Altwegg, K., Burch, J., Carr, C., Cu-pido, E., Eriksson, A., Gttler, C., Henri, P., Mokashi, P., Nemeth, Z.,Nilsson, H., Rubin, M., Sierks, H., Tsurutani, B., Vallat, C., Volwerk,M., Glassmeier, K.H., 2016b. First detection of a diamagnetic cavity atcomet 67p/Churyumov-Gerasimenko. Astronomy & Astrophysics 588,A24, doi: .Gombosi, T.I., 2015. Physics of cometary magnetospheres. Magnetotails inthe Solar System 207, 169–188, doi: .Graham, G.A., Rae, I.J., Owen, C.J., Walsh, A.P., Arridge, C.S., Gilbert, L.,Lewis, G.R., Jones, G.H., Forsyth, C., Coates, A.J., Waite, J.H., 2017. Theevolution of solar wind strahl with heliospheric distance: HELIOSPHERICSTRAHL EVOLUTION. Journal of Geophysical Research: Space Physics122, 3858–3874, doi: .Gurnett, D.A., Anderson, R.R., Ma, T.Z., Haerendel, G., Paschmann, G.,Bauer, O.H., Treumann, R.A., Koons, H.C., Holzworth, R.H., Lhr, H.,1986. Waves and electric fields associated with the first AMPTE arti-ficial comet. Journal of Geophysical Research 91, 10013, doi: .Haerendel, G., Paschmann, G., Baumjohann, W., Carlson, C.W., 1986. Dy-namics of the AMPTE artificial comet. Nature 320, 720–723, doi: .Hajra, R., Henri, P., Vallires, X., Mor, J., Gilet, N., Wattieaux, G., Goetz,C., Richter, I., Tsurutani, B.T., Gunell, H., Nilsson, H., Eriksson, A.I.,Nemeth, Z., Burch, J.L., Rubin, M., 2018. Dynamic unmagnetized plasmain the diamagnetic cavity around comet 67p/ChuryumovGerasimenko.Monthly Notices of the Royal Astronomical Society 475, 4140–4147,doi: . 29alekas, J.S., Brain, D.A., Ruhunusiri, S., McFadden, J.P., Mitchell, D.L.,Mazelle, C., Connerney, J.E.P., Harada, Y., Hara, T., Espley, J.R., Di-Braccio, G., Jakosky, B.M., 2016. Plasma clouds and snowplows: Bulkplasma escape from Mars observed by MAVEN: Plasma Clouds at Mars.Geophysical Research Letters 43, 1426–1434, doi: .Hansen, K.C., Altwegg, K., Berthelier, J.J., Bieler, A., Biver, N., Bockele-Morvan, D., Calmonte, U., Capaccioni, F., Combi, M.R., Keyser, J.D.,Fiethe, B., Fougere, N., Fuselier, S.A., Gasc, S., Gombosi, T.I., Huang,Z., LeRoy, L., Lee, S., Nilsson, H., Rubin, M., Shou, Y., Snodgrass, C.,Tenishev, V., Toth, G., Tzou, C.Y., Wedlund, C.S., the ROSINA team,2016. Evolution of water production of 67p/Churyumov-Gerasimenko: Anempirical model and a multi-instrument study. Monthly Notices of theRoyal Astronomical Society , stw2413, doi: .Haser, L., 1957. Distribution d’intensit dans la tte d’une comte. Bulletin dela Classe des Sciences de l’Acadmie Royale de Belgique 43, 740–750.Hassig, M., Altwegg, K., Balsiger, H., Bar-Nun, A., Berthelier, J.J., Bieler,A., Bochsler, P., Briois, C., Calmonte, U., Combi, M., De Keyser, J.,Eberhardt, P., Fiethe, B., Fuselier, S.A., Galand, M., Gasc, S., Gom-bosi, T.I., Hansen, K.C., Jackel, A., Keller, H.U., Kopp, E., Korth,A., Kuhrt, E., Le Roy, L., Mall, U., Marty, B., Mousis, O., Neefs, E.,Owen, T., Reme, H., Rubin, M., Semon, T., Tornow, C., Tzou, C.Y.,Waite, J.H., Wurz, P., 2015. Time variability and heterogeneity in thecoma of 67p/Churyumov-Gerasimenko. Science 347, aaa0276–aaa0276,doi: .Henri, P., Vallires, X., Hajra, R., Goetz, C., Richter, I., Glassmeier, K.H.,Galand, M., Rubin, M., Eriksson, A.I., Nemeth, Z., Vigren, E., Beth, A.,Burch, J., Carr, C., Nilsson, H., Tsurutani, B., Wattieaux, G., 2017. Dia-magnetic region(s): structure of the unmagnetized plasma around Comet67p/CG. Monthly Notices of the Royal Astronomical Society 469, S372–S379, doi: .Heritier, K.L., Altwegg, K., Balsiger, H., Berthelier, J.J., Beth, A., Bieler,A., Biver, N., Calmonte, U., Combi, M.R., De Keyser, J., Eriksson, A.I.,Fiethe, B., Fougere, N., Fuselier, S.A., Galand, M., Gasc, S., Gombosi,T.I., Hansen, K.C., Hassig, M., Kopp, E., Odelstad, E., Rubin, M., Tzou,30.Y., Vigren, E., Vuitton, V., 2017. Ion composition at comet 67p nearperihelion: Rosetta observations and model-based interpretation. MonthlyNotices of the Royal Astronomical Society 469, S427–S442, doi: .Heritier, K.L., Galand, M., Henri, P., Johansson, F.L., Beth, A., Eriksson,A.I., Vallires, X., Altwegg, K., Burch, J.L., Carr, C., Ducrot, E., Hajra, R.,Rubin, M., 2018. Plasma source and loss at comet 67p during the Rosettamission. Astronomy & Astrophysics 618, A77, doi: .Huang, Z., Tth, G., Gombosi, T.I., Bieler, A., Combi, M.R., Hansen, K.C.,Jia, X., Fougere, N., Shou, Y., Cravens, T.E., Tenishev, V., Altwegg, K.,Rubin, M., 2016. A possible mechanism for the formation of magnetic fielddropouts in the coma of 67p/ChuryumovGerasimenko. Monthly Noticesof the Royal Astronomical Society 462, S468–S475, doi: .Ip, W.H., Axford, W.I., 1987. The formation of a magnetic-field-free cavityat comet Halley. Nature 325, 418–419, doi: .Kajdi, P., Lavraud, B., Zaslavsky, A., Blanco-Cano, X., Sauvaud, J.A.,Opitz, A., Jian, L.K., Maksimovic, M., Luhmann, J.G., 2014. Ninety de-grees pitch angle enhancements of suprathermal electrons associated withinterplanetary shocks. Journal of Geophysical Research: Space Physics119, 7038–7060, doi: .Keller, H.U., Mottola, S., Davidsson, B., Schrder, S.E., Skorov, Y., Khrt,E., Groussin, O., Pajola, M., Hviid, S.F., Preusker, F., Scholten, F.,AHearn, M.F., Sierks, H., Barbieri, C., Lamy, P., Rodrigo, R., Koschny,D., Rickman, H., Barucci, M.A., Bertaux, J.L., Bertini, I., Cremonese, G.,Da Deppo, V., Debei, S., De Cecco, M., Fornasier, S., Fulle, M., Gutirrez,P.J., Ip, W.H., Jorda, L., Knollenberg, J., Kramm, J.R., Kppers, M., Lara,L.M., Lazzarin, M., Lopez Moreno, J.J., Marzari, F., Michalik, H., Naletto,G., Sabau, L., Thomas, N., Vincent, J.B., Wenzel, K.P., Agarwal, J., Gt-tler, C., Oklay, N., Tubiana, C., 2015. Insolation, erosion, and morphologyof comet 67p/Churyumov-Gerasimenko. Astronomy & Astrophysics 583,A34, doi: .31oenders, C., Glassmeier, K.H., Richter, I., Ranocha, H., Motschmann, U.,2015. Dynamical features and spatial structures of the plasma interactionregion of 67p/ChuryumovGerasimenko and the solar wind. Planetary andSpace Science 105, 101–116, doi: .Larson, D.E., Anderson, K.A., Lin, R.P., Carlson, C.W., Rme, H., Glass-meier, K.H., Neubauer, F.M., 1992. Electron distributions upstream of thecomet Halley bow shock: Evidence for adiabatic heating. Journal of Geo-physical Research: Space Physics 97, 2907–2916, doi: .Lindgren, C.J., Cravens, T.E., Ledvina, S.A., 1997. Magnetohydrodynamicprocesses in the inner coma of comet Halley. Journal of Geophysical Re-search: Space Physics 102, 17395–17406, doi: .Luter, M., Kramer, T., Rubin, M., Altwegg, K., 2018. Surface local-ization of gas sources on comet 67p/Churyumov-Gerasimenko based onDFMS/COPS data. Monthly Notices of the Royal Astronomical Society ,doi: .Madanian, H., Cravens, T.E., Burch, J., Goldstein, R., Rubin, M., Nemeth,Z., Goetz, C., Koenders, C., Altwegg, K., 2016a. PLASMA EN-VIRONMENT AROUND COMET 67p/CHURYUMOVGERASIMENKOAT PERIHELION: MODEL COMPARISON WITH
ROSETTA
DATA.The Astronomical Journal 153, 30, doi: .Madanian, H., Cravens, T.E., Rahmati, A., Goldstein, R., Burch, J., Eriks-son, A.I., Edberg, N.J.T., Henri, P., Mandt, K., Clark, G., Rubin, M.,Broiles, T., Reedy, N.L., 2016b. Suprathermal electrons near the nu-cleus of comet 67p/Churyumov-Gerasimenko at 3 AU: Model comparisonswith Rosetta data: Electrons Near Comet 67p: Rosetta Data. Journalof Geophysical Research: Space Physics 121, 5815–5836, doi: .Mandt, K.E., Eriksson, A., Edberg, N.J., Koenders, C., Broiles, T., Fuselier,S.A., Henri, P., Nemeth, Z., Alho, M., Biver, N., 2016. RPC observa-tion of the development and evolution of plasma interaction boundaries at67p/Churyumov-Gerasimenko. Monthly Notices of the Royal Astronomi-cal Society 462, S9–S22. 32emeth, Z., Burch, J., Goetz, C., Goldstein, R., Henri, P., Koen-ders, C., Madanian, H., Mandt, K., Mokashi, P., Richter, I.,2016. Charged particle signatures of the diamagnetic cavity of comet67p/ChuryumovGerasimenko. Monthly Notices of the Royal Astronom-ical Society 462, S415–S421.Neubauer, F.M., Glassmeier, K.H., Pohl, M., Raeder, J., Acuna, M.H.,Burlaga, L.F., Ness, N.F., Musmann, G., Mariani, F., Wallis, M.K.,Ungstrup, E., Schmidt, H.U., 1986. First results from the Giotto mag-netometer experiment at comet Halley. Nature 321, 352–355, doi: .Nilsson, H., Lundin, R., Lundin, K., Barabash, S., Borg, H., Norberg,O., Fedorov, A., Sauvaud, J.A., Koskinen, H., Kallio, E., Riihel, P.,Burch, J.L., 2007. RPC-ICA: The Ion Composition Analyzer of theRosetta Plasma Consortium. Space Science Reviews 128, 671–695,doi: .Nilsson, H., Wieser, G.S., Behar, E., Gunell, H., Wieser, M., Galand, M., Si-mon Wedlund, C., Alho, M., Goetz, C., Yamauchi, M., Henri, P., Odelstad,E., Vigren, E., 2017. Evolution of the ion environment of comet 67p duringthe Rosetta mission as seen by RPC-ICA. Monthly Notices of the RoyalAstronomical Society 469, S252–S261, doi: .Plaschke, F., Karlsson, T., Gtz, C., Mstl, C., Richter, I., Volwerk, M., Eriks-son, A., Behar, E., Goldstein, R., 2018. First observations of magneticholes deep within the coma of a comet. Astronomy & Astrophysics 618,A114, doi: .Rubin, M., Koenders, C., Altwegg, K., Combi, M., Glassmeier, K.H., Gom-bosi, T., Hansen, K., Motschmann, U., Richter, I., Tenishev, V., Tth,G., 2014. Plasma environment of a weak comet Predictions for Comet67p/ChuryumovGerasimenko from multifluid-MHD and Hybrid models.Icarus 242, 38–49, doi: .Scime, E.E., Phillips, J.L., Bame, S.J., 1994. Effects of spacecraft potentialon three-dimensional electron measurements in the solar wind. Journal ofGeophysical Research 99, 14769, doi: .33imar, A., Nemeth, Z., Szego, K., Dosa, M., Opitz, A., Madanian, H.,Goetz, C., Richter, I., 2017. Modelling the size of the very dynamic dia-magnetic cavity of comet 67p/ChuryumovGerasimenko. Monthly Noticesof the Royal Astronomical Society 469, S723–S730, doi: .Trotignon, J.G., Michau, J.L., Lagoutte, D., Chabassire, M., Chalumeau,G., Colin, F., Dcrau, P.M.E., Geiswiller, J., Gille, P., Grard, R., Ha-chemi, T., Hamelin, M., Eriksson, A., Laakso, H., Lebreton, J.P.,Mazelle, C., Randriamboarison, O., Schmidt, W., Smit, A., Telljohann,U., Zamora, P., 2007. RPC-MIP: the Mutual Impedance Probe of theRosetta Plasma Consortium. Space Science Reviews 128, 713–728,doi: .Vigren, E., Altwegg, K., Edberg, N.J.T., Eriksson, A.I., Galand, M., Henri,P., Johansson, F., Odelstad, E., Tzou, C.Y., Vallires, X., 2016. MODEL-OBSERVATION COMPARISONS OF ELECTRON NUMBER DENSI-TIES IN THE COMA OF 67p/CHURYUMOVGERASIMENKO DUR-ING 2015 JANUARY. The Astronomical Journal 152, 59, doi: .Vigren, E., Eriksson, A.I., 2017. A 1d Model of Radial Ion Motion Interruptedby IonNeutral Interactionsin a Cometary Coma. The Astronomical Journal153, 150, doi: .Wang, R., Lu, Q., Huang, C., Wang, S., 2010. Multispacecraft observationof electron pitch angle distributions in magnetotail reconnection: ELEC-TRON DISTRIBUTIONS IN RECONNECTION. Journal of GeophysicalResearch: Space Physics 115, n/a–n/a, doi: .Wang, Z., Fu, H.S., Liu, C.M., Liu, Y.Y., Cozzani, G., Giles, B.L., Hwang,K.J., Burch, J.L., 2019. Electron Distribution Functions Around a Re-connection X-Line Resolved by the FOTE Method. Geophysical ResearchLetters 46, 1195–1204, doi: .Young, S.L., Denton, R.E., Anderson, B.J., Hudson, M.K., 2008. Magneticfield line curvature induced pitch angle diffusion in the inner magneto-sphere: FLC PITCH ANGLE DIFFUSION. Journal of Geophysical Re-search: Space Physics 113, n/a–n/a, doi: .34hang, Y.C., Shen, C., Marchaudon, A., Rong, Z.J., Lavraud, B., Fazakerley,A., Yao, Z., Mihaljcic, B., Ji, Y., Ma, Y.H., Liu, Z.X., 2016. First insitu evidence of electron pitch angle scattering due to magnetic field linecurvature in the Ion diffusion region: Pitch Angle Scattering. Journalof Geophysical Research: Space Physics 121, 4103–4110, doi: . Appendix A. Comparing IES Fluxes with Different GeometricalFactors
Anode E l e v a t i on s eV str] -5
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Anode
Figure A.1: Comparison of differential flux of 173 eV electrons in the IES FOV using theupdated geometric factors (left) and the initial geometric factor of the instrument paper(right). Energy (eV) I E S e l ec t r on f l ux ( / c m s e V ) Constat GUpdated Gs