How Bi‐Modal Are Jupiter's Main Aurora Zones?

Using Juno‐measured >30 keV electrons, three regions with substantial ultraviolet emissions were identified previously for Jupiter's main aurora (excluding the polar cap): low‐latitude diffuse aurora, mid‐latitude Zone I of downward acceleration, and higher latitude Zone II of bi‐directional acceleration. Zone I, associated with upward magnetic field‐aligned currents, was represented as bimodal: sometimes supporting coherent downward electron electrostatic acceleration and sometimes downward electron broadband acceleration, with broadband acceleration usually delivering the most intense electron energy flux at Juno. Recent observations of up‐going ion beams within Zone I represent a challenge as to whether coherent electrostatic acceleration invariably accompanies broadband acceleration. Is this region strictly bi‐modal, or is there a continuum between these two modes? We address these questions by combining multiple ion and electron data sources to diagnose electrostatic potentials both above and below the spacecraft. We find: (a) During Zone I downward electron broadband events, there are examples where evidence of downward electron electrostatic acceleration completely disappears and examples where it endures at some level. (b) Most often, evidence of downward electron electrostatic acceleration is strongly suppressed with strong downward electron broadband acceleration. Residual potentials most often (not always) have values small (<10 kV) compared to the electron characteristic energies of 100–400 keV. (c) Care must be exercised in these studies because plasmasheet electron precipitation spectra can mimic broadband acceleration spectra. At least for weaker auroral broadband accelerations, there is likely to be a continuum of electrostatic and broadband participation. Why either process is favored at any one time is unknown.

2 of 22 of energies. The bimodal characteristic of the two different Zones is captured in Figure 1a by asserting that the electrostatic potentials shown are intermittent in character, and also by labeling the yellow, one-directional arrow for Zone I as representing either downward electron electrostatic or downward electron broadband acceleration. For Zone (I) of upward electric currents, contrary to Earth auroral observations, the downward broadband electron acceleration mode usually delivers the more intense energy flux to the aurora than does the downward electron electrostatic acceleration mode, and also has higher characteristic energies within those most intense regions. Similarly, Zone II sometimes reveals coherent downward ion electrostatic acceleration (in the form of downward ion inverted-Vs) and sometimes does not show such a feature at energies >50 keV. The word "coherent" in these discussions means that the evidence of electrostatic acceleration is not disrupted by stochastic processes.
The word "broadband" describes only the observational characteristics of the measured electron distributions; it is not itself an acceleration mechanism. Such characteristics are thought to be generated by wave-particle acceleration applied in a randomizing stochastic fashion. Whistler waves have been identified as contributing to some auroral features at Jupiter (Elliott et al., 2020;Kurth et al., 2018), and nonlinear solitary structures are thought to play a role in some regions of Earth's aurora (Ergun et al., 1998). However, Alfvén waves are considered the favorite for causing the stochastic accelerations, based on Earth-derived auroral physics (Sulaiman et al., 2022; see Lysak et al., 2021, for some recent theoretical underpinnings). Such Alfvén waves have not been observed directly within the jovian auroral acceleration regions. Rather, the absence of such waves in the auroral acceleration region, when such waves are observed elsewhere near the equator (Saur et al., 2003) and in the more equatorward diffuse auroral regions (Sulaiman et al., 2022), has been interpreted as resulting from the absorption of such waves as they propagate from higher to lower altitude regions. Alfvén waves have been observed at mid-latitude over nominal auroral field lines (Lorch et al., 2022). But to date those observed waves provided Poynting Flux powers to the ionosphere (∼0.8-20 mW/m 2 ) that are at most an order of magnitude less than needed to power modestly strong jovian auroral emissions. Although many researchers take the primacy of Alfvén wave acceleration for granted, more evidence will be required to verify that conclusion for broadband electrons of Jupiter's main aurora.
Aspects of the statistical probabilities of the occurrences of main aurora characteristics have been performed previously. For the first four Juno science orbits, Mauk et al. (2017b) reported the occurrences of at least brief observations of coherent downward electron electrostatic acceleration for 50% of the 8, low altitude auroral zone crossings examined. A much more complete examination was performed by Salveter et al. (2022) who found that for the time spent within the main auroral regions (a different metric than used by much more limited study of Mauk et al., 2017b), the downward broadband electron acceleration mode clearly dominated at 93% over 7% for coherent downward electron electrostatic acceleration (with ±3.8% uncertainty for each). This result was a surprise to many given pre-Juno expectations (e.g., Ray et al., 2010 and references therein) but not to all (Saur et al., 2003).
Based on the observations of upward ion beams by Szalay et al. (2021) in association with Zone I, it has been posited as a part of ongoing discussions that the downward electron electrostatic acceleration may be generally a part of Zone I acceleration even when downward electron broadband acceleration dominates the auroral energy fluxes. In this sense, the difference between the two different modes of Zone I would be more quantitative than qualitative. Here we address the question of whether, in Zone I, downward electron electrostatic acceleration invariably accompanies the downward electron broadband mode, or whether the evidence of electrostatic acceleration truly disappears. Despite the apparent statistical sparsity of electrostatically accelerated electron beams as found by Salveter et al. (2022), are there electrostatically acceleration processes hidden within the broadband regions?
One reason the answers to these questions remain uncertain is because the very high quality plasma ion measurements obtained by the Juno Jovian Auroral Distributions Experiment (JADE) instrument (McComas et al., 2017) can obtain full angular distributions only at a cadence of 15 s (1/2 of a spacecraft spin) or 30 s (one spacecraft spin), depending on the orientation of the magnetic field. That cadence is often insufficient to characterize the stronger auroral features that can come and go within the Juno data set over times scales of 1 to several seconds.
Here we address the questions at hand, and introduce a heretofore underutilized data set from the Juno Jupiter Energetic Particle Detector Instrument (JEDI) (Mauk, Haggerty, Jaskulek, et al., 2017). It is called the "time-of-flight × pulse-height" (TOF × PH) data set, providing proton distribution measurements with energies >5 keV (although >10 keV in practice). This data set has significant deficiencies when compared with the JADE 10.1029/2022JA031237 3 of 22 ion data, as described herein. Its key advantage is that it measures proton angular distributions at the same cadence as that used for the electron measurements taken by JEDI (as low as ∼0.5 s) when the magnetic field is aligned with the JEDI view plane. The result of this present investigation is to further constrain the characteristics of the Zone I portion of Jupiter's main aurora.
Our metric for deciding whether or not a particle distribution reveals the presence of electrostatic potentials is whether or not the slope of the PSD as a function of energy is greater than zero. That metric is well established in the study of Earth's aurora (as reviewed in the book by Paschmann et al., 2002, andrepublished as Amm et al., 2002). Figure 1b shows a comparison between a JEDI-measured electron Intensity spectrum and the corresponding PSD. The portion of the PSD spectrum showing the positive slope is labeled. PSD is derived by converting the Intensity from 1/keV to 1/erg (for cgs) and then dividing Intensity by the square of the particle's momentum (p 2 ; Ukhorskiy & Sitnov, 2013). In the non-relativistic limit p 2 is, of course, proportional to energy (E). Absent the conversion of the intensity to PSD, one may estimate (in the non-relativistic limit) whether or not the PSD has a positive slope by comparing the slope of the Intensity spectrum with a line that has the slope of E 1 , as illustrated with the green line in Figure 1b. In Earth's aurora it is noted that positive slopes in PSD tend to become unstable to wave-particle interactions, causing positive slopes to be reduced to observed slopes of near zero. That is why here we allow slopes of just greater than, to much greater than, zero to be considered at least candidates for evidence of electrostatic acceleration. However, in a novel environment like Jupiter's we must acknowledge that our metric may not be perfect. If wave particle interactions and electrostatic acceleration are acting independently on particle distribution in a comparable fashion in terms of energy, as they may be on ions that are being accelerated upward from below the spacecraft, some modest electrostatic potentials may be masked.
We note, finally, that the upward-going ion beams are often narrower in angle than are the downward-going electron distributions. Because of limitations in angular sampling, the result is that fewer auroral crossings can be fully characterized with regard to the up-going ions than can be characterized with regard to down-going electrons. Thus, we are able only to present a handful of event studies, and are not yet able to provide a statistical ensemble of events.

Juno Measurements
Juno orbits Jupiter with a highly elliptical polar orbit (Bolton et al., 2017). Its very low periapsis (∼1.1 RJ) allows it to fly through, and sometimes below auroral acceleration regions. Of interest to the present study are its instruments measuring energetic particles (JEDI: Mauk, Haggerty, Jaskulek, et al., 2017), plasmas (JADE;McComas et al., 2017), magnetic fields , and UV auroral images (Ultraviolet Spectrograph (UVS); Gladstone et al., 2017). The Waves (Kurth et al., 2017) and IR imager (Jovian Infrared Auroral Mapper; Adriani et al., 2017) are of great interest for studies of aurora, but are not discussed here. Figure 2 shows JEDI data for a characteristic crossing of Jupiter's main aurora. The UVS image associated with that crossing is shown in Figure 3A, compiled from a series of UVS scans taken over the time frame of about 1346 to about 1426, beginning about 8 min after the main auroral crossing of interest here (See Gladstone et al., 2017, for a discussion as to how this process is carried out). This Juno auroral crossing occurred at low enough radial distances (R ∼ 1.7 RJ; Figure 2 position labels) such that the particle detectors are able to resolve the loss cones (LC: ∼27°, Figure 1. (a) Schematic of the different regions and "Zones" (ZI and ZII) of Jupiter's main aurora, modified from a similar schematic in Mauk et al. (2020). This schematic does not address auroral processes in the Polar Caps. The image of Jupiter with the Ultraviolet Spectrograph image overlay was generated by Bertrand Bonfond for the Mauk et al. (2020) schematic (see, Bonfond et al., 2017, for more information on the character of the auroral images). (b) Example electron Phase Space Density (PSD with units s 3 cm −6 g −3 ) showing the key feature of electrostatic acceleration; a positive slope in the PSD just below a peak. Here the Intensity (I) spectra is the most direct measurement from the Jupiter Energetic Particle Detector instrument, and the PSD is I/p 2 , where "p" is the electron momentum, after "I" has been converted from the units of 1/keV to 1/erg. In the non-relativistic limit, PSD is proportional to I/E, where E is particle energy (not valid for the higher electron energies shown in the plot). Absent the PSD conversion, the presence of a positive slope may be estimated by comparing the slope of the Intensity spectra with the green line with the slope of E 1 . a rough estimate derived using the expression in Mauk et al., 2017b, which simply assumes that the magnetic field strength varies as 1/R 3 ). Panel b of Figure 2 shows energy-channel-averaged electron intensity as a function of pitch angle for electrons with energies 30-1,200 keV. Just above that panel are labels identifying two different regions of the aurora, Zone I and Zone II. With this figure, as in other figures in this report, we focus on the Zone I region and show only portions of the Zone II regions; see Mauk et al. (2020) for more information about this event. As originally defined, Zone I is where the greatest intensities reside within the downward loss cone, and specifically with intensities greater than those observed within the trapped populations outside of the loss cones. And for Zone (I), the upward loss cone is comparably empty. Zone II is where we generally see intensities within both loss cones greater than the trapped intensities, and most often with the upward intensities greater than the downward intensities. However, in Zone II the downward energy fluxes can sometimes be comparable to or greater than the downward fluxes seen in Zone I. In Figure 2, Panel d, the downward energy fluxes in Zone II for the time period shown are less than the peak Zone I energy fluxes, but equivalent to most of the Zone I downward energy fluxes. Panel d of Figure 2 shows how the differences between Zone I and Zone II are reflected in the relative values of the upward and (B)) and selected proton spectra associated with PJ4-S (Panels (a)-(d)). The UVS images were taken by the Juno Ultraviolet Spectrograph . These images are multispectral with red, blue, and green interpreted as corresponding to energy depositions of high, medium, and low energy electrons. White corresponds to a mix of energies. The Juno trajectory is magnetically projected onto Jupiter using the field models from Connerney et al. (2018) plus Connerney et al. (1981). The proton spectra (a), (b), (c), and (d) are vertical slices from Panel (g) in Figure 2, with sample times shown in Panel (f) The blue symbols show proton intensity (cm −2 s −1 sr −1 keV −1 ), and the purple points show the number of counts that went into each of the intensity points. The <50 keV data product is subject to so-called "accidental" contamination at the lower energies. The "universal" accidentals profile overlays the spectra shown in Panel (a) and (d) showing which portions of the spectra are in fact contaminated (See Text S9 in Supporting Information S1). Clear evidence of electrostatic acceleration occurs when there are positive slopes to the intensity spectra with slopes that rise faster than E 1 . The green lines in the spectral plots are included for helping to make that determination. and downward electron energy fluxes. In Zone II upward energy fluxes are greater than or equal to the downward energy fluxes (when the magnetic field is oriented such that the pitch angles [PA] are fully sampled by JEDI).
The UVS auroral image shown in Figure 3A shows a common ordering for auroral displays. These are multispectral images where red colors indicate the precipitation of the higher energy electrons (because of atmospheric absorption of some spectral bands) and blues, greens, and whites (mixtures of colors) indicate precipitation of lower energy electrons . Zone I and Zone II often order themselves as shown in this panel with higher-latitude Zone II (with often higher characteristic energies) showing up as red, and lower latitude Zone I (often with lower characteristic energy) showing up as white. The characteristic electron energies shown in Panel e of Figure 2 shows that characterization is not strictly true as portions of Zone I (near the vertical dashed line at about 1339:10) are seen to compete with Zone II with regard to characteristic energy. The characteristic energies shown in these plots is the ratio of downward integral energy intensity and downward integral number intensity for electrons with energies >30 keV (Note that in this paper we use the word "intensity" and "flux" for parameters that have in their denominators (cm 2 ⋅ s ⋅ steradian), and (cm 2 ⋅ s), respectively.) The UVS image in Figure 3A was not taken at the same time as was the particle data, and in that image it is not possible to determine exactly when the transition from Zone I to Zone II took place.
Of substantial interest to the present study is the character of the downward electron energy distributions (Panel c in Figure 2). Between about 1338:45 and 1339:05 we see a classic example of downward electron electrostatic acceleration in the form of "inverted-Vs." Panel a spectra 1 and 2 show examples from that region (at the times noted in Panel c). However, still within Zone I (relatively empty upward loss cones), the spectra become broadband between about 1339:06 and 1339:12. The most intense broadband spectra (spectrum 4 in Panel a) lasts for just about 1 s, and is associated with the highest energy fluxes throughout (Panel d) and among the higher characteristic energies (Panel e). Spectrum 3 in Panel a shows that there seems to be a transition between the downward electron electrostatic and broadband acceleration regimes.

Ion Measurements
Panel g of Figure 2 shows the new data that we bring to bear on determining the character of Zone I acceleration. This panel comprises two different JEDI data products: "time-of-flight × energy" (TOF × E) for protons with energies >50 keV (published previously), and "time-of-flight × pulse height" (TOF × PH) for energies <50 keV. Note that both of these products are discriminated with respect to mass composition such that the ions displayed in Panel g are known to be very light ions (protons) for >50 keV, and light ions (and specifically not O or S) for <50 keV (Mauk, Haggerty, Jaskulek, et al., 2017). Panel f of Figure 2 shows the angular distribution of the (>50 keV) protons that is often characteristic of Zone I; upward ion distributions narrowly confined to the magnetic field direction.
The TOF × PH data has several "features" that must be understood before it can be properly interpreted. First, the efficiency of proton detection falls very rapidly with decreasing energy. That feature produces dark regions at the lowest energies not necessarily because of the absence of intensity, but rather because of the absence of counts. This feature is apparent when we show individual spectra below. The TOF × PH technique in principal measures protons down to ∼5 keV, but in practice the efficiency of detection does not really allow measurements of proton foreground below about 10 keV.
A second feature of the TOF × PH data product is that it is susceptible to "accidental" contamination (see Text S9 in Supporting Information S1). This data product is obtained using a two-signal coincidence system that can be fooled by particle populations that are so intense that they can simultaneously (within a narrow time window) stimulate the two coincident elements. The TOF × E data product utilizes a coincidence between three different signals and is much less susceptible to accidental contamination. Fortunately, accidental contamination is very recognizable in the TOF × PH product; it has a "universal" spectral shape that can be compared with the measurements (again see Text S9 in Supporting Information S1), as we do so when showing individual spectra below. Note that accidentals, often caused by electrons, can stimulate the channels below 10 keV much more readily than can foreground protons. Help in identifying periods of accidental contamination is the reason that the spectrogram displays go down in energy to 5 keV. An additional feature of the TOF × PH data product is that the energy resolution degrades rapidly with lowering energy. As shown in Figure 34  and 72%, mostly driven by time dispersion and inaccuracies in the time-of-flight measurements (but also including the ion charge redistribution in the foils followed by a 2.5-3 kV post acceleration).
Finally, because solid-state-detectors no longer define the geometry of detection, the angular resolution of the TOF × PH observations (9° × 30°) is broader than that of TOF × E (9° × 17°). The JEDI 12 ion sensors (for both TOF × E and TOE × PH) view within a plane roughly perpendicular to the Juno spin axis, and spacecraft rotation moves these fields-of-views along their long axes. Hence, it is fairly easy for these FOVs to miss looking exactly along the magnetic field. Because of that factor, and because the full fields-of-view of the JEDI sensors are tilted and twisted somewhat (10°) to avoid looking at Juno's huge solar panels, we often see at least some spin modulation in the measurements of ions moving along the field lines.

Ion Characteristics
In Figure Figure 3 it should be noted that the sample times for the ion spectra in that figure do not all match the sample times for the electron spectra in Figure 2a (only the sample times of electron spectra 1 and 4 approximately match the sample times of ion spectra a and d). Sample times were chosen to highlight distinctive features in each species. Also, in Figure 3, where appropriate, the region where "accidental" contamination is apparent is identified by overlaying the universal accidentals spectrum. Where such an overlay is shown, that region of the spectrum should be considered noise and ignored.
In this paper, as discussed in the Introduction, in order to be considered a candidate for coherent electrostatic acceleration, the number intensity spectrum must rise with energy at a rate greater than to E 1 below a local peak such that when the intensity is converted to a PSD, its slope must be no less than flat as a function of energy (note that for non-relativistic energies, PSD ∼ Differential-Number-Intensity/Energy). The slanted green lines in Figure 3 panels are there for the purpose of testing that condition. We note that just having a peak in the differential number intensity spectra is not evidence of electrostatic acceleration. For example, the intensity spectrum obtained from a pure, unaccelerated Maxwellian has a peak at Energy = kT (the temperature of the Maxwellian), and rises as a function of energy as E 1 at energies substantially below the peak. As acknowledged in the Introduction, our absolute criterion for identifying electrostatic acceleration may not always identify regions of modest electrostatic acceleration where there is a mixture of comparable levels of independent wave-particle and electrostatic acceleration.
There is a tendency for the evidence of upward proton electrostatic acceleration to be less persistent than the evidence for downward electron electrostatic acceleration. Spectrum (b) in Figure 3 is an example where such electrostatic proton acceleration seems to disappear (at least above the 10 keV minimum measurement) while the electron electrostatic acceleration is near its peak. Apparently the electrostatic acceleration region had risen above the spacecraft. For proton spectra (a) and (c), the combined electrostatic potentials along these auroral field lines (adding the values above the spacecraft and below the spacecraft) are, respectively, ∼70 + ∼60 = ∼130 kV and ∼100 + ∼100 = ∼200 kV.
Of central importance here is what happens within the downward electron broadband acceleration regions of Zone I. Spectrum (d) in Figure 3 shows that within the brief most intense downward electron broadband acceleration event, there appears to be a relatively robust upward proton population that shows no evidence of upward ion electrostatic acceleration above the 10 keV measurement limit. Just prior to that most intense electron broadband acceleration, within the transition between downward electron electrostatic and downward electron broadband acceleration (e.g., electron spectrum 3 in Figure 2), the upward ions in Panel g of Figure 2 show either a flat ion intensity (no evidence of electrostatic acceleration) or the complete disappearance of ion populations.
We have indicated at the bottom of Panel e of Figure 2, with little purple rectangles, where in time the JADE instrument made measurements of the field-aligned ion populations. The purple numbers associated with the purple rectangles (and other features) reveal the approximate energy where JADE sees a local maximum in their differential energy intensity measurements. Note that the differential energy intensity is distinct from the differential number intensities used to plot JEDI data. Differential energy intensity and differential number intensity are favored by JADE and JEDI, respectively, because in each case the chosen parameter is closest to being proportional to the instrument count rates. As with differential number intensity, such a peak within the differential energy intensity does not necessarily indicate an accelerated beam, since a pure unaccelerated Maxwellian will give a peak at an energy of twice the temperature (kT), and have a distribution that rises at energies substantially below the peak as E 2 . For the time period just addressed (the second vertical dashed line in Figure 2) JADE was not in a configuration to measure parallel ion populations. We will return to a discussion of the JADE data below.
The time indicated with the very first dashed line in Figure 2 (1338:35.5) also shows a downward electron broadband acceleration within Zone I with reasonably high downward electron energy fluxes (200 mW/m 2 ) and modest downward electron characteristic energies (120 keV). We emphasize for later purposes that this region is clearly identified as Zone I by the pitch angle distribution (Panel b in Figure 2) showing the highest electron intensities within the downward loss cone and the relative absence of intensities within the upward loss cone. For this event no >10 keV proton signatures were associated with it. And so the JEDI TOF × PH data cannot alone be used to diagnose the presence of electrostatic potentials below the spacecraft, since the absence of protons could mean either that no upward ion electrostatic acceleration occurred or that there were no protons available to accelerate. However, we are fortunate in this case that JADE did happen to obtain complete pitch angle distributions at approximately the correct time (purple rectangle at the bottom of Panel e in Figure 2). The JADE measurements are shown in Figure 4, plotted over exactly the same time period as used for the JEDI data in   (d)) and electron data (Panels (e) and (f)). The parameters in Panel (a) are angle-averaged proton differential energy intensity spectra, and those in Panel (c) are energy-averaged proton energy-intensity pitch angle distributions. Panels (b) and (d) show, respectively, upward and downward proton differential energy intensities, sampled from the loss cone regions shown in Panel (c) with the horizontal dashed lines. Panel (f) shows the energy-averaged pitch angle distributions of electron differential energy intensities, and Panel (e) shows the electron differential electron energy intensity spectra sampled just within the upward loss cone shown in Panel (f). The purple rectangles above Panel (a)  period in Figure 4, Panel a shows the ∼2 keV population referred to with the purple "2" above Panel a (and in Figure 2). But, Panels b and d (anti-parallel and parallel distributions) make it clear that that 2 keV population is not associated with upward acceleration but more likely with some combination of a more intense downward ion population (Panel d) and perpendicular population. Panel b (upward distribution) reveals no evidence of upward electrostatic acceleration. Rather, Panel d (downward distribution) shows evidence of downward ion electrostatic acceleration. That conclusion is reinforced with Panels e and f, where we see evidence of upward, electrostatically accelerated electrons in the 5 keV range.
And so in this case, rather than seeing the upward ion electrostatic acceleration below the spacecraft associated with a downward electron broadband acceleration observed above the spacecraft, we see just the opposite, a downward beam of protons. And here a modest upward electron electrostatic acceleration in one region (below the spacecraft) is associated with a strong downward electron broadband acceleration observed in another region (above the spacecraft). Somehow in this case the downward electron broadband acceleration is decoupled from the generation of modest electrostatic potentials below the spacecraft. We presume, but do not demonstrate, that upward magnetic field-aligned electric currents in this region, supported and bolstered by the more energetic (∼120 keV) electron broadband processes occurring above the spacecraft, are somehow partially resisted in regions below the spacecraft. We speculate without verification that the downward electron broadband acceleration populated the very low altitude regions with electrons (but still above the ionosphere) causing a low altitude electrostatic potential to redistribute lower energy ions or electrons to enforce quasi-neutrality.
For the second (right hand) dashed-line event in Figures 2 and 4, we see in Figure 4 just prior to the downward electron broadband observation, upward peaked proton distributions with characteristic energies in the 0.3-3 keV range. This region is the one of transition between downward electron electrostatic acceleration and downward electron broadband acceleration (Spectrum 3 in Figure 2a). While the first (∼3 keV) peak in Panel b of Figure 4 has too few counts to interpret confidently, Figure 4 may be showing that some modest upward ion electrostatic acceleration (kV-level) is occurring below the spacecraft during this transition.
In all of these discussions it is worth paying attention to the downward electron characteristic energies (Panel e in Figure 2) and the hints of upward ion electrostatic acceleration occurring below the spacecraft. We need to understand how relevant an electrostatic potential that is less than, say, 10 kV is to the physics of energy flux (and electric current) transport when the downward electron broadband characteristic energies are in the 100's of keV.
A final comment about Figure 4 concerns the region of upward electron electrostatic acceleration occurring on the right-hand-side of Panel e as the spacecraft transitions to Zone II (as noted in Figure 2 above Panel b). We will return to that subject in the Discussion Section. Figure 5 shows an example of Zone I auroral acceleration (from PJ7-N) that has no real precedent within Earth's auroral region  has compared some jovian auroral features with Earth features as articulated in Amm et al., 2002;Paschmann et al., 2002; see also Kurth et al., 2018). Panel B of Figure 3 shows the corresponding UVS image, compiled from UVS scans from about 0030 to 0110 UT, ending about 6 min prior to the Juno crossing of the region of interest here. We see in Panel c of Figure 5, beginning at about 1115:43, the start of a rising region of downward electron "Inverted-V" electrostatic acceleration. Between about 0115:50.3 and 0115:53.6 (between the two vertical dashed lines) the downward electron distribution becomes broadband with higher downward energy fluxes and characteristic energies (Panels d and e). Following that broadband region, the distribution then suddenly reverts back to downward electron electrostatic acceleration. This sudden transition between different modes is one reasons that the downward electron electrostatic and broadband acceleration are both categorized within a single Zone, Zone I. In the UVS image in Figure 3B, the red to white transition fairly accurately reflects the time of the changeover from Zone II to Zone I. Note that most of Zone II had higher typical characteristic energies prior to the time period shown in Figure 5  Both prior to and then following the region of downward electron broadband acceleration we see strong upward ion electrostatic acceleration. Panel a-left of Figure 5 shows just one of these distributions, the one just following the region of downward electron broadband acceleration (see Panel g for the time location of this spectrum). Comparing Panels c (downward electrons) and g (upward protons) just following the region of downward electron broadband acceleration, we see 230 kV of upward electrostatic potential above the spacecraft and 130 kV below the spacecraft, for a total of about 360 kV along this auroral field line. Just prior to the region of electron broadband acceleration, that total is roughly 230 kV + 70 ∼ 300 kV. In Panel a-left, we see the clear evidence of accidental contamination at the lower energies, and we see that such contamination it is clearly distinguishable from the natural accelerated protons.

Perijove 7 North (PJ7-N)
Within the region of broadband acceleration in Figure 5, we see that the ion populations are strongly suppressed, and that any clear ion evidence of upward ion electrostatic acceleration must be below the potentials of ∼10 kV; see Panel a-right, where only the accidental contamination survives. The purple rectangles in Panel e of Figure 5 Figure 2. Here the range of centroid pitch angles used for the electron and the proton spectra are 155°-180°, and 0°-22°, respectively. The two proton spectra in Panel (a) were sampled at the times shown in Panel (g) Again, the universal "accidentals" spectrum is overlaid to show which portions of the spectra are contaminated. See the figure caption to Figure 3 and the main text for more information.
show that JADE was not in a position to measure parallel ion distributions in association with the broadband acceleration. However, the purple arrows labeled "2" in Panel e show that there was a weak non-parallel (trapped) proton population at this designated time (See Figure A1 in Appendix A; we put some figures in an Appendix to declutter the discussions somewhat). That observation may be important because it indicates that protons were present (both up going and down going) that could have revealed strong (>10 kV) electrostatic acceleration in the JEDI data if such acceleration were operating.
Within the broadband acceleration region of Figure 5 (between the dashed lines), there is some evidence of a conic distribution in Panel f, with a peak in intensity near 45°. Within that conic distribution, examination of the energy spectra shows no evidence of electrostatic acceleration as defined in this paper. A similar feature is seen in the example discussed in the next section (PJ10-N), and a hint to that feature is seen in Panel f in Figure 2 for PJ4-S near 1339:10 in association with that brief broadband region (identified with the red arrow in Panel f of Figure 2). One thought is that in the electrostatic acceleration mode of Zone I, conic-like accelerations below the spacecraft are subsequently electrostatically accelerated to look like field-aligned beams, whereas in the region of broadband acceleration, conics can retain their conic configuration. However, such a feature is not consistently observed (e.g., see the first dashed line in Figure 2), and this thought remains an undemonstrated hypothesis. We have tested the conic feature to determine that it is free of accidentals contamination, which can occur even in the TOF × E data when the electrons are unusually intense (See Text S9 in Supporting Information S1). See Clark et al. (2020) for a brief discussion of other conic observations at Jupiter made by Juno. Figure 6 shows another example that is similar to the example shown in Figure 5. The UVS image for this PJ10-North auroral crossing is shown in Figure 7A, compiled with UVS scans from about 1646 to 1726 UT, overlapping the time of the Juno crossing of the main aurora. The identification of the character of the northern edge of the auroral feature (labeled "Mix) is based on the quick transiting back and forth between Zones at the northern edge of the aurora. In the JEDI data in Figure 6 there is once again (Panel b) the rising downward electrostatic "Inverted-V" acceleration of electrons beginning at ∼1654:00, a transition to broadband at about 1654:20, and a transition back to electrostatic at about 1654:40. And once again the downward electron energy fluxes and characteristic energies peak within the broadband regions (Panels c and d). A feature of the downward electron broadband spectra seen in Panel b is a broad peak near 160 keV. As described in Text S1 in Supporting Information S1, this is a "minimum ionizing" peak associated with foreground electrons with energies in the MeV range that fully penetrate the detectors. That peak is easily distinguished from electrostatic acceleration as discussed in Text S3 in Supporting Information S1. The moment calculations shown in Figure 6 are lower limits for such conditions as discussed in Text S2 in Supporting Information S1. That minimum ionizing peak was not seen in Figure 5 because there the JEDI detectors were near to saturation, in the fashion described in Text S4 in Supporting Information S1.

Perijove 10, North (PJ10-N)
The upward going proton intensities shown in Panels e and f of Figure 6 (PA near zero) are sporadic, probably in part because of the high altitude of these measurements, and the incomplete viewing at just zero-degree pitch angle. But once again we do see clear examples of upward ion electrostatic acceleration of the protons both before (Spectrum (a) in Figure 7) and after (Spectrum (d) in Figure 7; see Panel f in Figure 6 for the time locations of these spectra). Within the downward electron broadband region, the protons show only accidental counts on the left side of the region (Spectrum (b)). On the right side of the region, and still within the downward electron broadband region, there is an ion population that is peaked, but not enough so to be interpreted as upward ion electrostatic acceleration according to our metric (Spectrum (c) in Figure 7).
The upward proton electrostatic accelerations observed both before and after the downward electron broadband region are unusual in being doubled-peaked. We have no definitive explanation for that unusual characteristic. In the time period just prior to the broadband region, that characteristic occurs just for the single, 1-s sampling shown in Spectrum (a) in Figure 7 at 1654:19 UT. Following the broadband region, that double-peaked characteristic occurs from 1654:45 to 1654:53 UT, with hints of the double peak extending to about 1655:03 UT. One possibility for explaining the double peak is that that some charged molecular hydrogen (H 2 + or H 3 + ) may be mixed in with the protons. Figure A2 in Appendix A shows that at the lower energies in the surrounding regions, H 3 + is indeed present near the times of the double-peaked electrostatic acceleration observations. Following an electrostatic potential acceleration, each proton within a H 3 + molecule would gain only (3) −0.5 as much energy as a single H + . The H 3 + molecule would likely dissociate in the front JEDI foil, and the scattering and energy-dependent energy losses there and in subsequent foils, may in some cases create an additional "proton" population. In the double-peaked ion spectra, the ratios of the higher peak energy to the lower peak energy is not consistent between Spectrum (a) and (d) in Figure 7, and it is not obvious that the process described above can give enough energy separation. But, the response of JEDI to H 3 + has not been studied, and is likely to be very complex. It is beyond the scope of the present study to further address the possibility that H 3 + is mixed into the accelerated ion distributions for this rare observation. Alternatively, could there have been, perhaps, two spatially separated regions of electrostatic acceleration with angle scattering in-between? These possibilities will have to be studied at a later time.
As shown with the purple rectangles in Panel d of Figure 6, JADE did make two measurements of parallel protons at the very beginning and end of the broadband region. The numbers of counts associated with these observations were too low to be interpreted with confidence (Panel b in Figure A3 in Appendix A). But in Panel d of Figure 6 we have labeled the two dim upward parallel peaks for the first period with 20 and 3 keV. If interpreted in this way, these two measurements could correspond to the ramp down of upward electrostatic potentials at the transition between downward electron electrostatic and downward electron broadband accelerations, down to perhaps a nominal 3 kV within the downward electron broadband region. That possible 3 kV electrostatic potential value needs to be compared with the downward electron characteristic energies of ∼350 keV (Panel d of Figure 6). The second parallel measurement by JADE on the right side of the electron broadband region shows only what looks to be a few counts near 30 keV (again see Figure A3). And, to the extent that those few counts can be interpreted, they look consistent with Spectrum (c) in Figure 7 (peaking near 30 keV), taken at about the same time.   N) shows in Panel e that the conic proton distribution peaking near 40° within the broadband acceleration region, is even more definitive than that seen in Figure 5 (PJ7-N). Again, examination of the energy distributions within this conic distribution (pitch angle >25°) shows none of the evidence of >10 kV electrostatic acceleration that we are looking for here.

Perijove 9, South (PJ9-S)
Figure 8 is significant as it is the event that stimulated the discussion that led to the questions addressed in this paper. Here we have combined the JEDI data in the top 6 panels with the JADE data in the bottom 4 panels (shown also in Szalay et al., 2021). The UVS auroral image associated with this event is shown in Figure 7B, where the boundary between Zone I and Zone II is very roughly at the position of the black dot positioned along the (interpolated) trajectory. The UVS scans used to compile this image spanned from about 1836 to 1916 UT, beginning at about the time when Juno crossed that black dot position. Of great interest in Figure 8 is the series of upward ion beams observed by the JADE as revealed in Panel g. The red arrows between Panels f and g highlight their locations as JADE samples the parallel directions about every 15 s; this timing has to do with spacecraft rotation and not a temporal characteristic of the beam. It has been demonstrated that the phase space densities of these beams have positive slopes as a function of energy at least through the bright beam at 1834:25, the beam with the white arrow pointing to it within the Panel g. In Panel b we see that the downward >30 keV electron distributions appear to be broadband.
But, we need to be more careful. The energetic electron pitch angle distributions (Panel a) show that the region at least through 1834:15 does not strongly fit the definition of a Zone I region in that the downward intensities in the loss cone are only comparable to the trapped intensities outside of the loss cones (with parallel intensities usually slightly below but occasionally reaching slightly above the trapped intensities). This situation should be compared with Figure 2, Panel c where in the distinct regions of Zone I, downward electron intensities are an order of magnitude larger than trapped intensities. We acknowledge that this metric for the identification of parallel acceleration is not a perfect one given that it averages over all energies measured by the sensor. Our best  Figure 4, with the exception that Panel (i) shows angle-averaged electron spectra rather than upward electron spectra. The centroid pitch angles used for JEDI electrons and protons were 0°-25° and 165°-180°, respectively. guess is that in this region (labeled "precipitation?" in Panel d of Figure 8) JEDI largely observed the precipitation of plasmasheet electrons in the fashion of a DifA. As those observed electrons carried only ∼10-15 mW/m 2 of downward energy flux (Panel c), we presume that the strong auroral acceleration leading to intense auroral emissions in Figure 7B is occurring below the spacecraft. Such parameters as the downward energy flux and the characteristic energies do not uniquely identify a region, but none-the-less those parameters in this region labeled "precipitation?" in Panel d of Figure 8 are in fact unexceptional as compared to the diffuse auroras regions identified by Mauk et al. (2020). While a statistical ensemble of measurements is not available, the diffuse auroras in that paper had an average energy flux of 23 mW/m 2 with a range of 1-100 mW/m 2 . The characteristic energy had an average of 194 keV with a range of 120-300 keV. These values are not very constraining, but in the region in Figure 8 labeled "precipitation?" the observed parameters are unexceptional as compared to those typical precipitation parameters. They are exceptional only in having a smaller characteristic energy for a portion of the time period, not a characteristic that suggests substantial additional broadband acceleration. While there may be some downward electron broadband acceleration going on within this "precipitation?" region, the observations are mostly consistent with a simple precipitation of plasmasheet populations down onto an auroral acceleration region that is primarily below the location of the spacecraft.
Things look different in the region between about 1834:18 and 1835:10, the region labeled "BB accel?" in Panel d of Figure 8. While the data in Panel a shows that the intensities within the downward loss cone are still only comparable to the trapped intensities, the downward energy fluxes and the characteristic energies have risen suddenly and substantially (Panels c and d). Also, the energy spectra (Panel b) show a modest minimum ionizing feature (identified in the panel with a double arrow) that suggests that MeV class electrons are involved (see Text S1 in Supporting Information S1). Here we believe that modest downward electron broadband acceleration is taking place, modest because the pitch angle plot (Panel a) shows that downward intensities are only comparable to the trapped intensities, not substantially larger as seen in Figure 2. However, the spacecraft is at a fairly high radial position (∼2 RJ), and it is possible that much more dramatic downward electron broadband acceleration was going on below the spacecraft.
The JADE ion beam labeled "upward ES accel." in Panel g of Figure 8 occurred squarely within the region ("BB accel.?" in Panel d) that we have identified as a region of downward broadband electron acceleration. And so, here is a "smoking gun" that robust (>40 kV) upward ion electrostatic acceleration can persist within a region of downward electron broadband acceleration within Zone I, at least for this possibly modest (as far as we can tell) broadband acceleration region.
The JEDI ion data (Panel f of Figure 8) shows additional but somewhat puzzling details. These ion data for the entire region of interest are strongly spin modulated because the JEDI sensors are missing the exact direction of the magnetic field line during portions of the spacecraft spin. For this reason, the JEDI ion data (Panel f) does not show what JADE shows at just the time of the critical JADE ion beam labeled with the white arrow in Panel g. However, good field-aligned viewing was achieved just following that time. The puzzling aspect of these ion data (still within the region identified as downward electron broadband acceleration) is that the characteristic bright rim at the higher energies, normally associated with electrostatic acceleration, disappears. More details are shown in Figure 9.
Spectrum (a) in Figure 9 reveals that when both JADE and JEDI have good viewing, they both see the electrostatically accelerated beam configuration. Spectra (b) shows the "rimless" protons spectra mentioned above. That spectrum drops precipitously at just above 100 keV in a fashion that one might expect with electrostatic acceleration. But the peak is not there. In spectrum (c) we have picked out one of the more peaked "beamlike" proton spectrum within the region clearly identified as broadband. These spectra should be compared with Spectra (a) and (d) where clear electrostatic acceleration is occurring. We suspect that electrostatic acceleration is playing a role in this puzzling region, but we have no explanation for why the distributions have taken on their puzzling characteristics. Could this distribution be the result of some combination of electrostatic and wave-particle acceleration?
There appears to be a second region of broadband acceleration between about 1835:22 and 1835:45 (again labeled "BB accel?" in Panel d and based mostly on Panel b in Figure 8). Here JEDI observes a clear upward accelerated (∼50 keV) proton beam at the very beginning of that period (a white arrow points to the particular observation in question in Panel f). While Spectra (b) and (c) in Figure 9 may (or may not, given their peculiar characteristics) indicate evidence electrostatic acceleration, this proton spectrum at 1835:22 is a second clear "smoking gun" for the persistence of robust (50 kV) electrostatic potentials during modest (again as far as we know) downward electron broadband acceleration. It is of interest that the electrostatic potential in question is lower than values just prior to this beam observation, and that the electrostatic potentials do not seem to persist into the rest of the broadband acceleration period.

Still Other Events
There were three other auroral crossing that were characterized by Mauk et al. (2020) with respect to Zone and mode of acceleration. And in Szalay et al. (2021), details of the ion beaming for three other events not addressed here were presented. We have not highlighted those events because the JEDI TOF × PH samplings along the field lines were no less sporadic than were the JADE measurements. However, we have examined these events with considerable care and have found general consistency with the findings of the four periods discussed in detail here. In just one of those six events (PJ6-S just after about 0653 UT) we did find a brief period of modestly large (20-30 keV) upward ion electrostatic acceleration during a Zone I electron broadband regions, with characteristic downward electron energy of 120 keV. This event was similar to PJ9-S in that this acceleration did not persist throughout the electron broadband acceleration region.

Summary of Observations
We have presented here four crossings of Zone I main auroral regions, and have examined six others for consistency. For PJ4-S we examined two different downward electron broadband acceleration periods. In one case there was no discernible evidence of upward electrostatic potential within an observed proton population, and in fact there appeared to be a modest (5 kV) upward electron electrostatic acceleration below the spacecraft, just the opposite of any expectations for such an upward electric current region. In the other case, evidence of upward ion  Figure 3 and the text for more about these spectra. electrostatic acceleration was not observed within a robust proton population, and any upward ion electrostatic acceleration had to be below the 10 kV limit of our observations as compared to the downward electron broadband characteristic energies of 200 keV. For PJ7-N any upward electrostatic potentials had decreased from ∼300 to 350 kV in surrounding regions to values, again, no greater than 10 kV as compared to the electron characteristic energies of 350 keV. For PJ10N the case was similar to that of PJ7-N where the electrostatic potentials decreased below 10 kV, with perhaps a possible brief direct JADE measurement of a 3 kV potential. Again these potentials are to be compared to downward characteristic electron energies of 350 keV. For PJ9-S, we finally found clear evidence of robust upward electrostatic potentials (40-50 kV) in association with what we think is a region of modest downward broadband acceleration (with characteristic energies of 180-200 keV). We use the word modest because the region only marginally met the criterion of Zone I broadband acceleration of having the downward intensities within the loss cone be substantially greater than the trapped intensities. We acknowledge that the broadband acceleration could be more robust at lower altitudes, that is below the 1.9-2 RJ radial observation position.
Based on the several cases examined here, we can say that evidence of electrostatic potentials is often clearly suppressed during strong downward electron broadband acceleration. Sometimes the evidence for the electrostatic potentials is completely eliminated but at other times very low values of electrostatic potentials (much lower than the downward electron characteristic energies) can persist. And occasionally higher values of electrostatic potentials can persist within (as far as we can tell) modest levels of downward electron broadband acceleration.

Discussion
We do not have enough studied events with sufficient angular sampling to provide comprehensive statements about when and where Zone I takes on its various characteristics. Our premise, based on the observations that we have, is that strong downward electron broadband acceleration can substantially suppress evidence of downward electron electrostatic acceleration (observed as upward ion electrostatic acceleration here) to levels that are perhaps irrelevant to the downward transport of energy flux (and possibly electric currents) associated with bright aurora. That is, for this condition, any remaining electric potentials (positive or negative) are dramatically less than the characteristic energies of the down going broadband electrons. An aspect of this premise is that H+ ions are likely accelerated at low altitudes by a combination of wave-particle interactions and electrostatic acceleration. When the upward ion electrostatic acceleration is suppressed, the wave-particle nature of the acceleration can sometimes be revealed in the form of conics. Electromagnetic Ion Cyclotron (EMIC) waves are one possible agent for accelerating the ions associated with conics (e.g., Clark et al., 2020 using the theory of Chang et al., 1986), and such waves have been observed in the Jupiter's auroral regions by Sulaiman et al. (2022). Much lower frequency, but highly dispersive, Alfvén waves have also been identified at Earth as another possible agent (Chaston et al., 2004(Chaston et al., , 2016. The situation of conic distributions revealing themselves during the apparent suppression of electrostatic potentials is not always observed, and there may be more than one wave-particle mechanism involved with the proton accelerations, given the at least two possibilities mentioned above. And, it is acknowledged here that our formal definition of the occurrence of electrostatic acceleration, positive slopes in the PSD of the up going ion distributions, may not be completely adequate when multiple independent acceleration processes are involved, particularly if the two processes are acting at comparable levels in energy. In cases where downward electron broadband acceleration is not taking place at any functional level, coherent downward electron electrostatic acceleration can play a relatively solo role in accelerating the electrons to a level that stimulates relatively bright aurora. One might conclude from the statistical work of Salveter et al. (2022) that that occurrence is relatively rare using the metric of the fraction of the main auroral zone where downward electron electrostatic acceleration prevails. But, there also appear to exist in-between states where both downward broadband electron acceleration and downward electron electrostatic acceleration (revealed as upward ion beams) play roles in the generation of the auroral brightness. And in the cases examined here, the downward electron electrostatic acceleration that would have added to the observed downward electron broadband acceleration only occurred below the spacecraft such that the consequences to electron acceleration were not directly measurable. Our premise (not proven) is that this in-between state exists where broadband acceleration is playing a role but where that process is acting only at a modest level.
We must leave open the possibility that electric potentials exist above the spacecraft even as the evidence for those potentials disappears. Is it possible that the electrostatic potentials remain even as the broadband acceleration processes wipe out any evidence of those potentials within the electron populations? Alternatively, as discussed in the next paragraph, the need for the electrostatic potentials may disappear as the broadband processes become more effective in transporting electrons. Given that the potentials below the spacecraft appear to be suppressed with strong downward broadband electrons, it would seem unexpected for electrostatic potentials to survive above the spacecraft as the evidence for those potentials disappear below the spacecraft. But, we note that the altitude location of the region of maximum electrostatic potential is expected to be quite variable, likely varying according to the parameter B/n, where B is magnetic field strength and "n" is electron density (Sulaiman et al., 2022). Given the high variability of the parameter "n", the altitude of the electrostatic potentials could shift dramatically with conditions, and possibly correlate with the conditions that cause shifts in the mode of acceleration.
Presumably the global magnetosphere-ionosphere system is insisting that the electrons on auroral field lines carry certain levels of electric current. Given insufficient electron densities and/or electron mobility, the current carrying ability of those electrons can be enhanced by downward electron broadband wave energization (increasing the mobility of the electrons within the loss cone) or by downward electron electrostatic acceleration. Sulaiman et al. (2022) have demonstrated that electron densities are indeed depleted in the Zone I regions, hence the apparent need for enhanced electron mobility for carrying the requisite currents. The two mechanisms of downward electron broadband and downward electron electrostatic acceleration might act alone in some situations or might act in concert with each other in other situations. Why the system chooses one mechanism over another at any one time or position is unknown at this time. There is a hint in the PJ7-N and PJ10-N cases (Figures 5 and 6) that the electrostatic potential solution to the current transport problem can become unstable; that when some parameter (current?) becomes too high, downward broadband electron acceleration becomes the solution that is available.
While there are regions of the main aurora where keV electrons play a persistent role, Jupiter's brighter aurora are dominated by electrons with energies in the 10's of keV and above . The regions Zone I and Zone II were defined on the bases of these greater-than-10's of keV energetic electrons. Given that strong downward electron broadband acceleration in Zone I seems often to suppress electrostatic potentials to levels that are at most below 10 kV and perhaps often much lower, the identification of the different modes of Zone I with the energetic electrons measured by JEDI has the appearance of being strictly bi-modal. But, what the present study has revealed is that the lower energy particles have a world of their own that may not be as well ordered as the Zone and Zone-Mode identifications would imply. We pointed out the feature in Figure 4 where the upward electron electrostatic acceleration at ∼5 keV energies seems to be in the opposite direction as the downward electron broadband acceleration with characteristic energies of 120 keV. That observation suggests that there is a decoupling of the behaviors of these two energy ranges. PJ10-N in Figure A3 shows a similar phenomenon between about 1653:30 and 1654:00. Here, in the midst of Zone I activities on either side (Figure 6), and where there are essentially no observed energetic electrons, there is a region of upward electron electrostatic acceleration (downward electric fields with a potential of 10-20 kV). Is this a low-energy Zone II region buried between two regions of energetic electron Zone I, or is something else going on?
The Zone II of Jupiter's main aurora was also represented as being bi-modal in Mauk et al. (2020). Specifically, as observed in energetic particles, this Zone sometimes shows downward electrostatic potentials revealed with downward proton Inverted-Vs with energies >50 keV, and sometimes does not show such a feature. And so, Zone II also needs to be examined with regard to its bi-modal character. Such an examination is beyond the scope of the present paper. It is specifically of interest that an upward electron electrostatic acceleration is seen in the right-hand-side of Figure 4, identified as Zone II, and also in Figure A3 at about 1655:30, also identified as a region of Zone II acceleration (in Figure 6). While downward electrostatic potentials above the spacecraft have been identified for some Zone II regions based on downward proton electrostatic acceleration , it has not been reported that those potentials below the spacecraft (in Zone II) are revealed with electrons. The potentials at the level shown in Figure A3 would have been missed by the JEDI measurements. The lesson here and above is there is much work to be done to integrate understanding of these different auroral Zones and modes across the wide range of particle energies and other parameters available to Juno.
As a final topic of interest, we note the findings of Szalay et al. (2021) regarding the role of auroral acceleration processes in extracting hydrogen ions from the upper atmosphere and ionosphere and populating the more distant reaches of Jupiter's magnetosphere. That study reported the extraction of 3 ± 2 kg/s of hydrogen utilizing strictly the occurrence of accelerated magnetic field-aligned beams of protons. That study may not have fully captured the extractions that arise from wave-particle interactions that may not manifest themselves as parallel beams. Future studies may result in higher efficiencies of auroral mechanism of populating Jupiter's magnetosphere with hydrogen ions.   Figure 6) and includes averages over all angles.

Data Availability Statement
The particle data presented here are available from the Planetary Plasma Interactions Node of NASA's Planetary Data System (PDS/PPI: https://pds-ppi.igpp.ucla.edu/), submitted under the authority of the Lead Investigators, Barry H. Mauk, of the Juno JEDI instrument (data located at: Mauk, 2021, https://doi. org/10.17189/1519713), and Frederic Allegrini of the Juno JADE instrument (data located at: Allegrini et al., 2021, https://doi.org/10.17189/1519715). The UVS data are at the atmosphere node of the Planetary Data System (https://pds-atmospheres.nmsu.edu/data_and_services/atmospheres_data/JUNO/uvs.html), submitted under the authority of the Juno UVS Lead Investigator, G. Randy Gladstone (data located at: Gladstone, 2017, https://doi.org/10.17189/1518951). All of these data are submitted under the direction of the Juno Principal Investigator, Scott Bolton. All of the data providing individuals listed here are co-authors on this paper. There are no restrictions on the use and publication of this data, but citations to the PDS data are requested, as documented on the PDS/PPI web site at https://pds-ppi.igpp.ucla.edu/citations_policy.jsp. Also, ASCII dumps with header documentation has been performed for each panel of the JEDI data displayed in this paper and is freely accessible without restrictions at Zenodo (Mauk, 2023, https://10.5281/zenodo.7675147). While any number of commercial software packages (including Excel) may be used to extract and display the ASCII JEDI data used here with descriptive header information, display software used for this publication is available online and can be accessed by contacting the lead author. A one-hour teleconference tutorial provided by the lead author or his designate is generally sufficient for a user to have sufficient expertise to proceed.