Local Time Asymmetries in Jupiter's Magnetodisc Currents

We present an investigation into the currents within the Jovian magnetodisc using all available spacecraft magnetometer data up until 28 July 2018. Using automated data analysis processes as well as the most recent intrinsic field and current disk geometry models, a full local time coverage of the magnetodisc currents using 7,382 lobe traversals over 39 years is constructed. Our study demonstrates clear local time asymmetries in both the radial and azimuthal height‐integrated current densities throughout the current disk. Asymmetries persist within 30 R J where most models assume axisymmetry. Inward radial currents are found in the previously unmapped dusk and noon sectors. Azimuthal currents are found to be weaker in the dayside magnetosphere than the nightside, in agreement with global magnetohydrodynamic simulations. The divergence of the azimuthal and radial currents indicates that downward field‐aligned currents exist within the outer dayside magnetosphere. The presence of azimuthal currents is shown to highly influence the location of the field‐aligned currents, which emphasizes the importance of the azimuthal currents in future magnetosphere‐ionosphere coupling models. Integrating the divergence of the height‐integrated current densities, we find that 1.87 MA R J−2 of return current density required for system closure is absent.


Introduction
The existence of a current disk at Jupiter has been well established since the flybys of the Pioneer probes in the 1970s (Hill & Michel, 1976;Smith et al., 1974). This current disk is a consequence of the strong rotationally driven dynamics that dominate the Jovian magnetosphere. Unlike at Earth where the current sheet is present only in the tail region, Jupiter's current disk is present throughout all local times. A plasma disk is formed from plasma known to originate primarily from the volcanic moon Io, comprising of mostly atomic sulfur and oxygen dissociated from SO 2 . Iogenic neutrals are ejected into the local space environment and ionized. Once ionized, Lorentz forces accelerate the plasma toward corotation with the planet (see review by Khurana et al., 2004;Thomas et al., 2004). Radial diffusion of the centrifugally confined plasma via flux tube interchange events and hot plasma injections produces the plasma disk (Hill & Michel, 1976;Krupp et al., 2004;Mauk et al., 1999).
The magnetic field geometry of Jupiter's magnetosphere is heavily influenced by the presence of the plasma disk and associated current disk. In order to conserve angular momentum, plasma flowing radially outward begins to lag corotation. As a consequence the frozen-in field is drawn into a bent back configuration. A ⃗ × ⃗ B force, by means of a radial current, is set up to accelerate the plasma back toward corotation (Hill, 1979). The flux tube coupling the lagging magnetosphere plasma to the ionosphere will enforce a velocity differential in the ionospheric plasma. Subsequent ion-neutral collisions in the ionosphere exert a frictional torque, balanced by a ⃗ × ⃗ B force, transferring angular momentum from the planet to the magnetosphere. It is these corotation enforcement currents that drive the main auroral emission at Jupiter, with the associated electrons precipitating into the planet's atmosphere (Cowley & Bunce, 2001;Hill, 2001;Khurana et al., 2004;Ray et al., 2010;Southwood & Kivelson, 2001). Radial stretching of the intrinsic field occurs again due to a ⃗ × ⃗ B force associated with radial stress balance (Caudal, 1986;Vasyliunas, 1983).
The three-dimensional structure of the current disk is complex and time dependent. Arridge et al. (2015) provided a review of studies that demonstrated asymmetries in the magnetic field configuration, plasma flow, and thickness of the current sheet. The review discussed these complex asymmetries arising due to both internal rotational stresses and external solar wind forcing on the system. The main auroral emission signifies a steady-state coupling between the magnetosphere and the ionosphere; hence, understanding the asymmetries in the magnetodisc is fundamental to understanding the magnetosphere-ionosphere (M-I) coupled system, which drives this emission (see review by . Certain features in the main emission are known to be fixed in local time (LT), such as discontinuities in the emission near noon and bright dawn storms (Chané et al., 2018;Gustin et al., 2006;Radioti et al., 2008). Ray et al. (2014) demonstrated such variations in these currents by applying a 1-D M-I coupling model at 1-hr LT intervals throughout the Jovian magnetosphere. They showed that the auroral currents were stronger in the dawn sector than the dusk or noon sector by an order of magnitude. The authors emphasized that this approach did not consider azimuthal currents or the azimuthal bend-back in the magnetic field and explicitly called for their consideration in future studies.
LT asymmetries have been observed in the UV auroral emissions. Using 1,663 FUV Hubble space telescope images, Bonfond et al. (2015) showed that 93% of southern and 54% of northern hemispheric images suggested a larger emitted power in the dusk sector than the dawn sector. The southern dusk sector was approximately three times brighter than its dawn counterpart, while northern sectors displayed a relatively similar brightness. The authors attributed this difference to magnetic field variations between hemispheres, arguing also that the southern values are a better representation of the field-aligned current (FAC) system associated with the main emission, as they lack the superimposed uncertainty of the northern magnetic anomaly.
New insights into the Jovian system are being made with the Juno spacecraft, and in order to incorporate these findings into M-I coupling models, the need to move away from symmetric descriptions is apparent. A more complete understanding of the current system and associated magnetic field within the magnetodisc has the potential to alleviate the discrepancies between model predictions and observations. Khurana (2001) determined the radial and azimuthal height-integrated current densities (HICDs) within the magnetodisc using all spacecraft magnetometer data available up until 31 May 2000 in regions where there was a well defined current sheet. Their findings showed clear LT asymmetries within the system. The divergence of the currents in the magnetodisc indicated the presence of Region 2 currents; Khurana (2001) argued this was due to solar wind forcing on the magnetosphere. From the magnetic field data they were able to quantify the extent of field bend-back over LTs, which was included in a later current sheet geometry model (Khurana & Schwarzl, 2005). Khurana (2001) was limited by the lack of data coverage in the dusk and noon sector of the magnetosphere. Hence, the structure of the radial and azimuthal currents in the noon-dusk magnetosphere could not be fully determined. Furthermore, limited insight into the location and strength of return currents was available. As a consequence of this restricted data set, simulations have been unable to make comparisons within the dayside magnetosphere (Walker & Ogino, 2003). Now, with updated magnetic field and current sheet geometry models, and by applying automated processes where previous work relied on visual techniques, we build upon this study to provide a full LT coverage of the currents within the Jovian magnetodisc.
In this paper, section 2 covers the methodology behind extracting the lobe magnetic field values from the magnetometer data, the calculation of the radial and azimuthal HICDs, and how we subsequently deduced the location of the FACs. Our results are displayed in section 3, and discussion of the results, including their implications for the magnetospheric plasma, is undertaken in section 4. We conclude with a summary of our findings in section 5.

Measurements
We utilize magnetometer data from all Jovian missions and flybys, up to and including 28 July 2018. Cassini magnetometer data were not included in this study as the spacecraft did not traverse the current disk. We adopt the same time resolutions as Khurana (2001) for comparison and coherence: 1-min resolutions were used for Pioneer 10 and 11, Ulysses, and Juno; 48-s resolution for Voyager 1 and 2; and 24-s resolution for Galileo Real Time Survey mode. Where finer cadence data were unavailable, we used 32-min averages. To preserve similar temporal resolution between spacecraft, 1-min averaged data were used from Juno. Figure 1 shows an equatorial projection of the spacecraft trajectories to encounter Jupiter. We impose a compressed magnetopause configuration at 62 R J from Joy et al. (2002) in our analysis. This is illustrated by the solid black lines. Data employed in this study were constrained to regions within the compressed magnetopause with a fixed stand off distance.

Analysing Magnetometer Data
The observed magnetic field recorded by the spacecraft magnetometers is a summation of an internal dynamo field and an external perturbation field. Hence, to ascertain the magnetic field contribution from the magnetodisc, an internal field model must be subtracted from the observed magnetometer data. The JRM09 internal field model (Connerney et al., 2018) was used in this study. JRM09 is a 10th-order spherical harmonic expansion of Jupiter's magnetic field with coefficients derived from Juno perijove data (PJ01 through PJ09). Previous models were constructed using data from Pioneer and Voyager flybys and constrained to the Io auroral footprint (Connerney et al., 1998;Hess et al., 2011) or had an additional dipole superimposed to agree with Hubble space telescope observations . The JRM09 model is the ideal candidate for our study, as the model exploits low-altitude measurements of the magnetic field and so contamination by external fields is negligible. Stallard et al. (2018) gave evidence supporting an immutable intrinsic field between the Galileo era and the Juno era. As most of our data are from this time frame, we apply JRM09 throughout our study.
Once the internal field is subtracted, the remaining magnetic field is associated with the current disk, magnetopause, and tail currents. As we limit our study to distances inside a compressed magnetopause boundary (Joy et al., 2002), we neglect contributions from magnetopause currents in our analysis. We assume this is a good approximation as current disk effects dominate in the middle magnetosphere.
It is crucial to work in a current disk reference frame in order to isolate the magnetic perturbation from the current disk. Our new reference frame is a rotating cylindrical reference frame, centered on the planet with pointing radially outward, locally tangential to the current disk surface; z is normal to the current disk; and completes the right handed system. The angles of rotation are found by determining the normal to a model current disk surface given by Khurana and Schwarzl (2005), where x H is the hinging distance of the current disk, set to be −47 R J ; x and are the Jupiter-Sun-Orbital positions of the spacecraft, where ⃗ x points toward the Sun and ⃗ points antiparallel to Jupiter's orbital velocity; CS is the tilt angle of the current disk; is the west longitude of the spacecraft; ′ is the prime meridian of the current disk; and sun is the angle between the Sun-Jupiter line and the Jovigraphic equator. The model incorporates hinging of the current disk due to solar wind forcing and information delay as a function of radial distance due to wave travel time and field geometry. For further information on this model we refer the reader to Khurana and Schwarzl (2005) and Khurana (1992).

Calculation of Magnetodisc Currents and Their Divergences
Taking Ampere's law in cylindrical coordinates and integrating over the current disk thickness, the radial and azimuthal HICD, J ′ and J ′ , may be given as where B , B , and B z are the differenced radial, azimuthal, and normal field strengths in the lobe regions, respectively; w is the half thickness of the current disk, assumed to be 2.5 R J to align with other studies  (Connerney, 1981;Khurana & Kivelson, 1993). Within the current disk, azimuthal variations in B z are negligible and hence are not considered in the determination of the radial HICD.
It is important to note that at Jupiter, the lobe region refers to the magnetic field in regions above and below the current disk. Varying the current disk half thickness between 2 R J and 10 R J does not produce a significant difference in the HICDs outside of 60 R J . However, we find variations in the azimuthal HICD, up to 20% within 50 R J , and a variance of up to 100% localized at 50 R J . B z was determined by fitting a polynomial of the form B z ( ) = a + b 2 + c 3 , bounded between 6 and 100 R J , to the differenced z component of the magnetic field. Best fits for the coefficients were found to be a = −1.825 × 10 2 nT R J , b = 1.893 × 10 4 nT R 2 J , and c = −8.441 × 10 4 nT R 3 J . B measurements are seen to reverse periodically due to the ∼10.3 • tilt of the Jovian dipole with respect to the spin axis (Connerney et al., 2018). Periods where spacecraft traversed the lobe region are identified by applying an algorithm that retrieved values at times where the B component reaches a plateau. Plateaus are defined as regions where consecutive values of B do not deviate by more than ±7.5% for a period of 30 min or more. A variation of ±7.5% is applied as this offers a juste milieu by allowing for small fluctuations in the field while ignoring the larger variations associated with the traversal of lobes.
This study adopts the modal value of B , B , and B z measured in the determined lobe regions. The mode gives a more accurate value of the lobe field strength, as opposed to the mean that is often skewed by the slowly varying field signatures recorded while still within the current disk. Figure 2 demonstrates how the nonlobe field skews the mean, while the modal value lies within a more reasonable estimate of lobe value. For lobes with no mode we adopt the median value. The latter half of the Ulysses flyby and intervals from the Juno data set were excluded due to their large deviations from the equatorial plane. Data outside of the fixed magnetopause boundary were also excluded. In total 7,382 valid intervals were retrieved, we believe this number to be biased. For example, a sharp fluctuations in the lobe field will result in two readings being returned, one prior to and one after the fluctuation. Both of these recordings are returned and used in this study. Hence, the number of true lobes recorded will be less than what were retrieved.

Height-Integrated Current Density
The radial and azimuthal HICDs were calculated using the modal values for B , B , and B z in the lobe regions. We illustrate this in Figure 3, which shows the radial HICD (Figure 3a), the azimuthal HICD (Figure 3c), and their binned averages (Figures 3b and 3d, respectively). For the radial HICDs warmer colors indicate outward radial currents, while cooler colors indicate planetward radial currents. The azimuthal currents all flow in the direction of corotation and are shown using a natural log scale.
Initially, we see a clear asymmetry in the averaged radial currents. Strong outward radial currents dominate the midnight through dawn magnetosphere, while weaker inward currents exist from noon to dusk region. Within 40 R J strong outward radial currents are present, but weaken at noon. Within distances of 20 R J , radial currents appear to be weaker in the postdusk sector than at other LTs. The azimuthal HICDs, shown in Figures 3c and 3d, decay with radial distance. The azimuthal HICDs are larger in the midnight through dawn sectors than in noon through dusk.
These asymmetries are better seen in Figure 4, which shows the variation in the HICDs over local time at fixed radial distances. Here, data are binned in 5 R J × 3hr bins, centered at 5 R J intervals. In both the radial and azimuthal HICDs, we see local time asymmetries begin to develop with radial distance from 15 R J . For the radial currents, a maxima develops around 6 LT with a minima at noon. This minima shifts to around  18 UT at larger radial distances. The azimuthal currents exhibit a noon-midnight asymmetry. Azimuthal currents are weakest in the around noon and largest at midnight. The azimuthal currents fall off rapidly with increasing distance, becoming comparable to the radial currents further from the planet.
In order to highlight the variation in radial and azimuthal currents with local time, we bin our results in 6-hr LT regions, centered on midnight, dawn, dusk, and noon. This is shown in Figures 5 and 6 for the radial and azimuthal HICDs, respectively. Black dots represent the HICD calculated from lobe values, and the red line represents the mean binned every 5 R J . Radial currents are seen to peak around 30 R J and then steadily decrease. The azimuthal currents decrease in a 1/ fashion. For both the radial and azimuthal components, currents in the noon sector are weaker than the other sectors.
The errors in Figures 4, 5, and 6 are given by the standard error of the mean, , calculated as = √ n , where is the standard deviation of the averaged data and n is the number of data points within the bin. The mean was chosen over the median to represent our results, as the associated error was considerably less than the median absolute error obtained from the median average. Figure 7 shows the divergence of the radial (Figure 7a), azimuthal (Figure 7b), and perpendicular (Figure 7c) HICD. From current continuity,

Divergence of the HICD
Hence, we relate the divergence of our radial and azimuthal components, ∇J ′ and ∇J ′ to the divergence of the parallel currents. Warmer (cooler) colors indicate current being added to (removed from) the denoted component. Radial currents are enhanced within the inner middle magnetosphere, while depleted through noon and dusk. Azimuthal currents are fed in the dusk magnetosphere and removed at dawn, consistent with the analysis of Khurana (2001).
From current continuity, the perpendicular and parallel divergence must equal 0. Therefore, the sum of the radial and azimuthal current divergence can be used to reveal the location of upward and downward FACs, where again warmer (cooler) colors indicate upward (downward) FACs. Our deduced FAC locations agree with those from Khurana (2001); however, there is now complete coverage in the dusk and noon sector due to the additional data.

Discussion
We identify 7,382 traversals using data spanning nearly 5 decades. Over this time, the solar wind conditions at Jupiter would have varied widely. Though the orientation of the interplanetary magnetic field has little effect on asymmetries in the system, the dynamic pressure of the solar wind does (Cowley & Bunce, 2001). For example, a compression of the magnetosphere due to an increase in solar wind dynamic pressure forces plasma radially inward. This increases its angular velocity and consequently decreases the corotation enforcement currents, altering the magnetic field geometry. By binning and averaging over all data available, we do not consider temporal fluctuations, such as those associated with variations in the solar wind dynamic pressure or perturbations in the current disk. Therefore, this analysis presents an average view of Jupiter's current disk where the variance in the data, captured in our error analysis, reflects some of these natural fluctuations.
The radial currents in Figures 3a and 3b are associated with corotation enforcement. Outward radial currents act to accelerate plasma toward corotation, while inward radial currents decelerate the plasma. Peaks occurring around 30 R J in Figure 5 are consistent with the location of corotation breakdown inferred from auroral observations and numerous models (Clarke et al., 2004;Nichols & Cowley, 2004;Ray et al., 2010;Figure 7. The divergence of the HICD for the (a) radial (b) azimuthal and (c) perpendicular components, in a similar format to Figure 3. The divergence of the perpendicular components is analogous to the parallel current density.

10.1029/2019JA027455
2012). Azimuthal currents in Figures 6c and 6d are associated with the ⃗ × ⃗ B forces acting to balance radial stresses. The presence of weaker radial and azimuthal currents at noon is in part due to the influence of solar wind on the Jovian magnetosphere. As plasma rotates through dawn into the dayside magnetosphere, it is constrained by the magnetopause and forced closer to the planet. As a consequence its velocity increases, and the ⃗ × ⃗ B force required to keep the plasma in corotation is decreased (Chané et al., 2017;Kivelson & Southwood, 2005;Walker & Ogino, 2003). Additionally, the solar wind dynamic pressure acts to balance the outward radial stresses, and so weaker azimuthal currents are present in the dayside magnetosphere.
Asymmetries are prevalent throughout the magnetosphere, demonstrating the the influence of the solar wind on the Jovian system. While these are strongest in the middle-to-outer magnetosphere, the inner magnetosphere is also affected. At 20 R J , the nightside radial and azimuthal currents exceed the dayside by ∼30% as shown in Figures 5 and 6. The asymmetries increase with radial distance. Outside 40 R J the dawn-dusk asymmetry in the HICD is stark. Strong outward currents in the dawn sector transition into inward currents through noon into the dusk sector. These currents act to accelerate plasma through dawn into the dayside magnetosphere and decelerate it through the noon to postdusk sector. Radial currents in the dawn sector have little variation with radial distance but decrease steadily in the other three LT sectors, with the noon and dusk currents falling off more rapidly than the night sector. It should be noted that around 50 R J the currents are highly dependent on the choice of current disk thickness and so our results may vary by ±100%. Figure 4. Transitioning from smaller to larger radial distances, we see a fairly uniform radial HICD, growing increasingly with respect to . The weakest radial currents are observed between 12:00 and 18:00, agreeing with findings by Ray et al. (2014), who showed a region of weaker current density in the post noon sector. Azimuthal currents are seen to fall off with distance. As in Khurana (2001), a peak begins to develop in the radial and azimuthal HICD around 06:00 and 00:00 LT, respectively, but now, we see the weakest currents are present at noon. A similar development has also been reported at Saturn by Martin and Arridge (2019) where a minimum occurs in the HICD through the noon-dusk sector.

The development of LT asymmetries is highlighted in
The asymmetries determined in the HICDs are consistent with those observed in other data sets. Variations in the plasma flow velocity have been observed in Galileo energetic particle data by Krupp et al. (2001). They showed pronounced LT asymmetry in plasma velocities within 50 R J , with dawn-noon velocities being greater than noon-dusk velocities. However, plasma data are limited outside 50 R J in the dusk sector. Within the inner-to-middle magnetosphere, plasma flows derived by Bagenal et al. (2016) were slightly larger in the dawn sector than the dusk sector; however, their study only extended to 30 R J . Results by Bunce and Cowley (2001a) showed that the current disk field falls off more rapidly in the dayside than at similar distances in the nightside. Walker and Ogino (2003) applied a global magnetohydrodynamic model to investigate the influence of the solar wind on the structure of currents within the Jovian magnetosphere. In their study they compared their simulated currents with the findings of Khurana (2001); however, a system-wide comparison could not be made as a result of limited Galileo orbiter data. Our study has now revealed the structure of currents within these regions. Though the simulated current densities are overall much weaker than our observed HICDs, the asymmetries present in the simulation are in qualitative agreement with these observations. Outward radial currents are predicted in the prenoon sector and inward radial currents in the postnoon sector. This is consistent with the observations of a transition from outward to inward radial currents within the noon magnetosphere. However, the inward radial currents predicted in the midnight sector are not present in this study. Similarly, using a 3-D global magnetohydrodynamic simulation, Chané et al. (2018) investigated the cause of localized peaks in auroral emissions. They showed flux tubes being accelerated through dawn into noon before decelerating and moving in toward the planet. The presence of strong outward radial currents within the dawn sector in our results agrees with the results from this simulation.
Examining the divergences of the perpendicular currents, Figure 7a illustrates positive radial divergences present throughout all LTs within 30 R J and up to 70 R J in the dawn sector suggesting an increasing radial current, while current is being removed within the postnoon to duskside magnetosphere. For the azimuthal component in Figure 7b, we find the same dawn-dusk asymmetry reported by Khurana (2001), indicating an azimuthal current loss in the dawn magnetosphere and a gain in the dusk sector. A strong downward current region can be seen in the dayside magnetosphere. It is possible that this could be related to the auroral discontinuity observed by Radioti et al. (2008); however, we have not mapped the signature to confirm this. Radial HICDs play a key role in determining the location of the FACs. As can be seen in Figure 7, the divergence of perpendicular currents are largely similar to the divergence of radial currents, with some variation in the inner regions due to a strong azimuthal divergence in the currents. This highlights the importance of considering both radial and azimuthal currents when describing the MI coupling system responsible for Jupiter's auroral emissions.
A prominent feature in the tail region is an alternation between positive and negative divergences in the radial component, and subsequently the perpendicular current divergence. This affect, referred to as "striping" by Martin and Arridge (2019), is an artifact of the differencing method. Small variations between adjacent bins, containing low counts, result in the appearance of a larger divergence. This feature is more pronounced in the radial divergence due to the smaller magnitude values. We find this affect can be mitigated by increasing the bin size to encompass more data points with the drawback of decreased resolution, or by bootstrapping data within the bins. Our bin choice produced an agreeable trade-off between the conservation of fine structures and minimizing the striping effect. Figure 8 presents a comparison of the radial and azimuthal divergences. Regions where the divergence of the azimuthal (radial) currents are greater than the radial (azimuthal) currents are colored red (white). In white regions the divergence of perpendicular currents is determined largely by the radial currents. In all LT sectors, with the exception of the predawn sector, the divergence of the azimuthal currents influences the presence of FACs to a similar degree as the divergence of the radial currents. When utilizing M-I coupling models to describe the Jovian current system, it is therefore important to consider not only the effect on FACs by the azimuthal currents but also the effect of LT asymmetries in determining their location and magnitude. Prevailing discrepancies between 1-D M-I coupling models and observations could be a consequence of neglecting the effects of azimuthal currents in the Jovian system. Future M-I coupling models should strive to amalgamate both the influence of radial and azimuthal currents in order to obtain a more realistic description of the system. This could be done through an empirical description of the HICDs; however, we leave this for future work. At Saturn, the divergence of radial currents is much smaller in magnitude than the divergence of azimuthal currents. As such, the divergence of azimuthal currents largely determines the location of the FACs at Saturn (Martin & Arridge, 2019) and should be strongly considered.
We note that upward FACs dominate the inner (40 R J ) region of the magnetodisc; however, there are much stronger positive divergences between 16:30 and 01:30 LT and weaker, sometimes negative divergences at approximately 07:30-13:30 LT. Radioti et al. (2008) suggested that these return currents within the dayside magnetosphere would correspond to discontinuities observed in the main auroral emission. The strong upward FACs found in the dawn sector could be attributed to the shearing motion of flux tubes described by the Chané et al. (2018) simulation. As this is a common feature, we would expect it to appear in our time-averaged results and would act to enhance FACs in the region. Again, predictions made by Walker and Ogino (2003) are consistent with our findings. In the Khurana (2001) study, return FACs were not identified in the dusk-noon region due to the lack of available magnetometer data. With the increased coverage used in this study, we are able to reveal evidence of current closure in outer dusk-noon magnetosphere, radially adjacent to regions of strong upward FACs.
By summing over the divergence of the perpendicular currents throughout the system, an overall positive divergence of 1.87 MA R −2 J is calculated. As current continuity must be maintained, the missing return currents must exist in the unmapped regions of the system, that is, dayside and magnetopause regions. This value assumes that the effect of magnetopause currents is negligible in the current disk. As mentioned previously, by working within the boundary of a compressed magnetosphere, we begin to limit the influence of these currents on our results. Furthermore, the magnetopause currents are external to the current disk region and take the form of a Laplacian field inside the magnetosphere. For a Laplacian field, the two terms of equation (3) exactly cancel, providing no contribution to the local azimuthal currents. Variations in the current disk thickness would influence the strength of the azimuthal currents and alter this value. This 10.1029/2019JA027455 demonstrates the need for a description of the spatial variation of current disk thickness, which could help to provide a more accurate representation of the azimuthal currents within 50 R J . This would further constrain the contribution of azimuthal currents to the location and magnitude of FACs.

Summary
We have presented an analysis of the current structure within the Jovian magnetodisc using all magnetometer data available until 28 July 2018. We build upon previous work by Khurana (2001) using the latest internal field model, current disk geometry model, and an automated lobe finding process. In doing so, we are able to provide a high-resolution, full LT coverage of the radial and azimuthal HICDs in the Jovian current disk. Our conclusions are as follows: 1. Asymmetries in both radial and azimuthal HICDs exist within the inner portion of the middle magnetosphere, some manifesting within 20 R J . 2. Both radial and azimuthal currents are weakest in the dayside magnetosphere. 3. Azimuthal currents are shown to play a key role in determining the location of FACs. 4. By summing over all known perpendicular divergences, a net positive divergence of 1.87 MA R −2 J is found. We postulate this to be balanced along the magnetopause and/or in the tail region.
We therefore suggest that future M-I coupling models should take into account not only the presence of radial currents but also azimuthal currents and the asymmetries found in both. Furthermore, when utilizing and constructing models of the current disk, it is paramount that the asymmetries be taken into consideration. Future work aims to produce an empirical description of these asymmetries, such that they can be readily integrated into M-I coupling models, as well as producing a full spatial description of the variation in current disk thickness.